Child PIP
Version 2.0
Saving lives
through death auditing
CHILD HEALTHCARE PROBLEM IDENTIFICATION PROGRAMME
Saving lives through death auditing
Saving Children 2009
FiveYearsof Data
A sixth survey of child healthcare in South Africa
Compiled by CR Stephen, LJ Bamford and ME Patrick, and the MRC Unit for Maternal and Infant
Health Care Strategies
Technical editing by DF Wittenberg
ISBN No: 978-0-620-50443-0
Pretoria 2011
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
This publication was supported by Cooperative Agreement Number 1U2GPS001053-02 from Centers
for Disease Control and Prevention (CDC). Its contents are solely the responsibility of the authors
and do not necessarily represent the official views of CDC.
The report should be cited as follows:
Stephen CR, Bamford LJ, Patrick ME, Wittenberg DF eds. Saving Children 2009: Five Years of Data
A sixth survey of child healthcare in South Africa. Pretoria: Tshepesa Press, MRC, CDC; 2011
The individual chapters in the report should be cited as follows:
Chapter authors. Chapter title. In Stephen CR, Bamford LJ, Patrick ME, Wittenberg DF eds. Saving
Children 2009: Five Years of Data A sixth survey of child healthcare in South Africa. Pretoria:
Tshepesa Press, MRC, CDC; 2011, pp
Child PIP Project Leaders
ME Patrick, CR Stephen, A Chiba, LJ Bamford
Phone 033 8973146 • Fax 033 8973409
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Foreword
How do you know when you have a successful innovation or enterprise? It
might be because your franchise has increased by over 400% in five years or
that your product is attracting international interest and is showing spectacular
success in performing whatever it’s designed to do. It may be that your
sponsors are keen to continue to invest in the enterprise’s growth. By all these
criteria, the Child Healthcare Problem Identification Programme (Child PIP) is
a best-seller and a raging success. While it most definitely is not a for-profit
business endeavour, Child PIP extended its coverage from 19 South African
hospitals in 2005 to 98 in 2009, is recognised by the WHO as a leading
innovator in the field, and has continued to attract funding from the United
States Centers for Disease Control (CDC) despite the global financial recession.
I’ve had the pleasure of observing the growth of Child PIP from the time its
progenitors, Bob Pattinson and Angelika Krug, first proposed the relatively
simple (but mostly untested) concept of using the information gathered from
careful mortality review of child deaths to identify remediable or preventable
elements, and use it to improve the quality of care sick children receive in the
health system. That was in the early 2000’s, and I must admit that I initially
considered the model somewhat unwieldy and possibly difficult to administer
reliably with over 140 modifiable factors to consider. However, the model has
undergone multiple refinements since then and under the stewardship of a
highly dedicated team led by Mark Patrick and Cindy Stephen has become a
reliable, valid and highly valued source of child hospital mortality information.
In 2010, South Africa exulted in the glory of a remarkably successful hosting of
the FIFA World Cup. Many of the key determinants of this triumph are the
same ingredients driving the success of Child PIP. Health professionals
involved with Child PIP are displaying leadership at national and provincial
level as well as at individual hospitals and championing change rather than
embracing inertia and the widely prevalent culture of mediocrity. An audit
system demands accountability – no child death can be left unexplained. It
influences attitudes and behaviours, either by restoring the health professionals’
sense of responsibility and yearning for excellence or by discouraging the
“indifferent, impatient, uncaring and lazy” behaviours often assigned by the
public to many health workers. Although there are still no national norms or
standards for health resources or practice, the Child PIP audit system at least
ii
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
requires these be considered at the local level to judge unit performance.
Finally, the system offers hospitals and individuals the possibility of
empowerment by promoting local ownership of actions in transforming an
ailing or failing health system, within the security of a larger family of
individuals motivating for similar changes.
I have attended the past three annual Child PIP conferences, where delegates
from across the country meet to present and discuss the latest developments.
It’s definitely one of my favourite meetings, and I inevitably leave inspired by
the innovation, enthusiasm, commitment and optimism shown by Child PIP
practitioners operating in a health service that I universally experience as
gloomy, disgruntled and dispassionate. Appreciating that ordinary folk can
make a difference is a powerful motivator for me to continue my own work. Of
course, it is easy to be deluded by all the positive energy circulating at the
meeting into concluding that Child PIP must be good. The kind of evidence
generated is sobering, and most, if not all practitioners, recognise that change in
complex healthcare systems cannot be brought about simply by the analysis of
data that indicate that care might be less than perfect. The management of
change is often more challenging than the clinical issues addressed by audit.
And so, as Child PIP heads forward to a brighter future, these are three key
challenges I’d like to see it surmount. Firstly, to convince authorities and
practitioners at every one of the 238 non-participating hospitals nationally, that
they too should be part of the auditing system. Secondly, the programme must
extend its efforts into systematically generating evidence that child mortality
audit can make a difference, and is cost effective. Finally, the data generated
must not remain only in the domain of those of the already converted
(including those who read this report!). The solutions to reducing child
mortality require policy makers, health managers and, particularly, families and
communities, to know that it is their problem too and we all have to act. If
there is an accelerated impetus to respond to the very achievable actions
suggested in this report, then the lives of thousands of South African children
will undoubtedly be saved.
Haroon Saloojee
Personal Professor and Head
Division of Community Paediatrics
Department of Paediatrics and Child Health
University of the Witwatersrand
iii
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Table of Contents
Foreword
ii
Prof Haroon Salojee
Acknowledgements
vi
Executive Summary
vii
PART ONE
Chapter One
THE CHILD HEALTHCARE SURVEY:
2005-2009
An overview of five years of Child PIP data
1
LJ Bamford
Chapter Two
Call to Action
20
Chapter authors
PART TWO
Chapter Three
PRIORITY CONDITIONS AND
OPPORTUNITIES FOR CHANGE
HIV
24
N McKerrow
Chapter Four
Malnutrition
34
T de Maayer, A Chiba
Chapter Five
Acute Respiratory Infections
47
K Harper
Chapter Six
Diarrhoeal Disease
61
A Westwood
Chapter Seven
Sepsis
76
ME Patrick
iv
S A V I N G
Chapter Eight
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
Tuberculosis
O F
D A T A
86
BL Dhada
PART THREE
Chapter Nine
CHILDREN AT SPECIAL RISK
The very young child: neonates (0-28 days)
102
CR Stephen
Chapter Ten
The critically ill child: deaths within 24 hours
119
ME Patrick
List of Abbreviations and Definitions
133
Appendices
137
Appendix A
Appendix B
Data tables for Chapter 1
Provincial Data
Eastern Cape
Free State
Gauteng
KwaZulu-Natal
Limpopo
Mpumalanga
North West
Northern Cape
Western Cape
Appendix C
Child PIP Data Capture Sheets & Code Lists
Monthly Tally
Death Data Capture Sheet
Cause of Death
Modifiable Factors
Appendix D
Additional Tools
Child PIP Mortality Review Process Guideline
v
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Acknowledgements
The Child PIP Group would like to thank the following, whose ongoing
contributions make a difference to the quality of child healthcare in South
Africa:
Professor Haroon Salojee, a legendary champion for child health,
for graciously writing the Foreword
Professor Dankwart Wittenberg, for his excellent and astute
content editing which has added greatly to the quality of this report
The National Department of Health, Child and Youth Health
Directorate, for ongoing interest and support
Provincial Maternal, Child and Women’s Health units, for their
enthusiastic involvement
Centers for Disease Control and Prevention, for ongoing support
and funding; and to Helen Savva for her expert editing skills
Professor Bob Pattinson, for his leadership, wisdom and support
Mrs Cathy Bezuidenhout, for her efficient and skilled organisation
Perlcom cc, for expert software programming and ongoing
assistance
The Child PIP Users, both those whose work appears in this
report, as well as current and future users
This report is dedicated to the those leaders who care, dream, plan and motivate for
change to improve the quality of healthcare children receive in South Africa
vi
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Executive Summary
Introduction
As with earlier editions, this Saving Children Report presents the findings from
South African hospitals that used paediatric mortality auditing to assess their
quality of child health care. The Child Healthcare Problem Identification
Programme (Child PIP), a South African mortality auditing tool and system was
used for this purpose.
Objectives
This survey has aims to:
Continue to collect demographic, social, nutrition, HIV, cause
of death and modifiable factors data on children who die in South
African hospitals to assess the quality of care,
Analyse demographic and quality of care data for each of the leading
causes of death, and
Reinforce and update earlier recommendations for improving care and
reducing child deaths.
Setting
One hundred and one hospitals in all nine South African provinces contributed
data to the report. This represents slightly less than 30% of all hospitals in
South Africa. District hospitals accounted for two thirds of the contributions,
although a higher proportion of regional and provincial tertiary hospitals
contributed data (43% and 36% respectively).
Methods
Following earlier data collection methodology, all sites used Child PIP to
structure the mortality review process. Each site integrated collection of the
data with their ongoing local audit process and used the data to compile a site
report. Data from the sites were amalgamated into a national database.
Survey Period
1 January 2005 to 31 December 2009.
vii
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Survey Population
All children, from birth to eighteen years, who were admitted to children’s
wards in the participating hospitals.
Findings
A total of 19 295 deaths of 343 408 admissions were audited in detail during the
five-year period. For these deaths, 53 326 modifiable factors were identified
with an average of 2.8 modifiable factors per death.
Most children who died (63.2%) were under 1 year of age with children
between 1 and 5 years accounting for a further quarter (25.2%) of deaths. The
majority of deaths in children were due to a limited number of conditions with
five conditions accounting for more than three quarters (77%) of all audited
deaths, and 79% of deaths in children younger than five years of age. These
conditions are acute respiratory infections (ARI) (29%), diarrhoea (21%),
septicaemia or possible serious bacterial infection (16%), tuberculosis (TB)
(7%) and meningitis (7%).
Slightly more than 50% of children who died were known to be HIV-infected
or -exposed, with many missed opportunities for prevention of HIV infection
being documented. Similarly, 34% of all children who died were classified as
having severe malnutrition, whilst a further 30% were underweight-for-age.
Thirty-four percent of child deaths occurred within 24 hours after admission.
Approximately one-quarter (26%) of deaths were considered to be avoidable.
Two thirds of modifiable factors occurred in the health system, while home or
community level modifiable factors accounted for the remaining third.
Approximately half (54.4%) of modifiable factors were related to clinical
personnel.
Call to Action
The Child PIP process not only provides valuable data regarding child deaths,
but also identifies how gaps can be addressed. The recommendations target
different key functions within the health sector and are thus divided into four
sections, namely policy, management and administration, clinical practice and
education.
Analysis by Cause of Death
The second part of the report focuses on each of the leading causes of death.
Newborn infants and children requiring emergency care have been identified as
viii
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
groups that require particular attention and chapters on these groups are
included in Part Three. Each chapter contains Child PIP data regarding the
condition or group of children. In addition, the standards of care for the
condition, gaps with regard to knowledge, and treatment guidelines and
recommendations regarding key steps for reducing mortality and morbidity
associated with the condition are presented.
The Appendices contain data from the Child PIP national database, provincial
summaries, the Child PIP data capture sheets and code lists, as well as an
outline of the mortality review process.
ix
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
PART ONE
THE CHILD HEALTHCARE SURVEY:
2005-2009
Y E A R S
O F
D A T A
Chapter
1
An overview of five years of
Child PIP data
Lesley J Bamford
MBChB, BSoc Sci (Hons), FCPead(SA), DrPH
Child and Youth Health Directorate, National Department of Health, South Africa, and
School of Health Systems and Public Health, University of Pretoria
Introduction
On the 26th October 2010, the Honourable Minister of Health, Dr
Aaron Motsoaledi, together with eight other ministers and the nine
provincial MECs for Health, signed the Negotiated Service Delivery
Agreement (NSDA) on behalf of the health sector. The NSDA, which
is a contract or “charter, that reflects the commitment of key sectoral
and intersectoral partners linked to the delivery of identified outputs”,
outlines four key strategic outputs that the health sector must achieve
within the next five years. These are:
Output 1: Increasing life expectancy
Output 2: Decreasing maternal and child mortality
Output 3: Combating HIV and AIDS and decreasing the
burden of disease from tuberculosis
Output 4: Strengthening Health Systems effectiveness.
The NSDA therefore places child mortality and the need to reduce the
number of children who die in South Africa firmly on the country’s
agenda. The scale of the problem should not be underestimated.
Approximately 75 000 children die before their fifth birthday in South
1
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Africa each year, and the United Nations provide under-five mortality
rates of 69 per 1000 live births in 2008.1 Despite some variation in
estimates of child mortality rates from different sources, there is no
doubt that these levels are far higher than those in other comparable
middle-income countries. Even though poverty and inequity contribute
to many deaths, ensuring that all children have access to a package of
simple preventive and curative healthcare services has been shown to
reduce child mortality substantially.2
Since 2004, the Child Healthcare Problem Identification Programme
(Child PIP) has contributed to improving knowledge regarding child
deaths and to reducing levels of these deaths. Child PIP is a voluntary
audit process that is designed to ascertain the quality of care that
children receive in the South African health system. It provides
structure and tools for conducting mortality reviews or audits of inhospital deaths of children by:
Ensuring that all deaths are identified;
Determining the social, nutritional and HIV context of each
child who dies;
Assigning a cause to each death; and
Determining modifiable factors, which identify instances
where failure to meet specific standards of care contributed or
may have contributed to the child’s death.
The strength of Child PIP lies in the fact that it encourages teams of
healthcare workers to reflect on the quality of health and other services
that children in their care receive. Child PIP also encourages them to
identify gaps or deficiencies in this care, and to find solutions that will
improve care in the future.
1
Every Death Counts Writing Group. Every Death Counts: saving the lives of mothers, babies and children
in South Africa. National Department of Health, Medical Research Council, University of Pretoria,
UNICEF, 2008.
2
WHO and UNICEF (2010) Countdown to 2015 Decade Report: Taking Stock of Maternal, Newborn and
Child Survival. Geneva.
2
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
The Saving Children reports that have been published since 2004
contain information on child deaths collated from data submitted by
participating hospitals. Although all sites follow a similar methodology
when auditing deaths, data are collected from multiple sites by different
reviewers and inevitably represent subjective judgment by the reviewers
(especially with regard to identification of modifiable factors). Data
were also submitted by different hospitals in different years, thus
comparisons between years and between facilities should be interpreted
with care. These reports, however, are based on detailed and careful
analysis of many child deaths providing a useful description of child
deaths and the quality of care that these (and other children) receive,
but also identifying and highlighting areas for change and
improvements in the health system.
This report covers the period January 2005 until December 2009.
Overall 101 hospitals in all nine provinces out of a total of 339
hospitals contributed data to the report. Just less than 30% of all
hospitals submitted data: district hospitals accounted for two-thirds of
the contributions, although there was a higher proportion of regional
(43%) and provincial tertiary (36%) hospitals. This discrepancy is due
to the historical development of Child PIP which was developed and
initially implemented primarily by paediatricians (who are usually based
at regional or tertiary levels), but has since been taken up by general
doctors and professional nurses working in district hospitals.
Table 1. Number and percentage of hospitals submitting Child PIP data: 2009
Level of Hospital
Total no.
No. submitting data
% of hospitals
District
Regional
Provincial Tertiary
National Central
Total
259
53
14
13
339
68
23
5
2
98
26
43
36
15
29
This chapter aims to provide an overview of data on all deaths audited
during this period. In addition, improvements in the Child PIP
software now allow for more detailed analysis of data using various
parameters including cause of death. For the first time it is possible to
present and analyse demographic and quality of care data for each
cause of death. An overview on cause of death is presented in this
chapter, while subsequent chapters provide more detailed data on each
of the leading causes of death.
3
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Information about the children who died
Baseline data
ADMISSIONS AND
IN HOSPITAL
MORTALITY RATE
AUDITED DEATHS
Data collected through the monthly tally sheets that record all
admissions and deaths in participating hospitals are shown in Table 2.
As the number of participating hospitals increased from 19 in 2005 to
98 in 2009, the total number of annual admissions increased from 23
653 to 108 852 with a corresponding increase in the number of
recorded deaths. The In Hospital Mortality Rate (IHMR) has been
stable at 5% of admissions over the last few years. The higher mortality
rates recorded during the first two years of reporting may have resulted
from the relatively small numbers of participating hospitals, but may
also indicate an encouraging downward trend. Even so, one in twenty
children still dies during the admission to hospital.
A total of 19 295 deaths of 343 408 admissions were audited in detail
during the five-year period. For these deaths, 53 326 modifiable factors
were identified with an average of 2.8 modifiable factors per death. The
number of modifiable factors identified per death appears to be
increasing over time, as seen in Table 2. This may reflect a decline in
quality of care, but conversely, may be indicative of a greater awareness
of standards of care and the need to strive towards meeting these
standards.
Table 2. Core data (all ages): 2005-2009
Total Admissions
Total Deaths
In Hospital Mortality Rate
Audited deaths (all)
Total Modifiable Factors
Modifiable Factors per death
* From monthly tally sheets
†
3
2005
2006
2007
2008
2009
23 653
1 543
6.5%
1 537
3 757
2.4
40 665
2 393
5.9%
2 871
5 539
1.9
63 378
3 190
5.0%
3 828
7 986
2.1
106 860
5 379
5.0%
5 539
16 773
3.0
108 852
5 398
5.0%
5 520
19 271
3.5
Individual audited deaths
3
The relatively small increase in the number of admissions between 2008 and 2009 despite a significant
increase in the number of hospitals is due to a number of factors. These include the fact that a number of
large hospitals did not submit data in 2009, whilst many of the additional hospitals were small district
hospitals with low numbers of admissions.
4
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Cause of death
The majority of deaths in children are due to a small number of
conditions with five conditions accounting for 77.2% of all audited
deaths and 79% of deaths in children younger than five years of age
(Appendix A, Table A8). These conditions are acute respiratory
infections (ARI) (28.9%), diarrhoea (20.7%), septicaemia or possible
serious bacterial infection (16.2% of deaths), tuberculosis (TB) (7.1%)
and meningitis (6.6%). These figures are consistent with data from
other sources, although it should be noted that deaths occurring in
nurseries and newborn units are not included in the Child PIP audit
process. Thus newborn deaths are underrepresented and the data do
not highlight the important contribution of deaths during the newborn
period to overall under-five mortality rates.
Figure 1. Main cause of death in children under 5 years: 2005-2009
HIV AND
NUTRITIONAL
STATUS
Information regarding the nutritional and HIV status of all children
who die is collected. Slightly more than 50% of children who died were
known to be HIV-infected or -exposed (Appendix A, Table A10).
Similarly, 34.3% of all children who died were classified as having
severe malnutrition, whilst a further 29.5% were underweight-for-age
(Appendix A, Table A14). It is of concern that the weight of almost
8% of children who died was unknown.
Demographics
Boys accounted for slightly more deaths than girls (52.2% compared
with 46.6%). Most children who died (63.2%) were under 1 year of age
with children between 1 and 5 years accounting for a further quarter
5
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
(25.2%) of deaths. Thus 88.4% of deaths occurred in children under
five years of age (Appendix A, Table A2).
Figure 2. Proportion of deaths in children under five years, by age: 2005-2009
Sixty-five percent of these deaths occurred in the category “1 month to
1 year”, whilst 6% of deaths in children younger than 5 years of age
occurred in the newborn period (0-28 days). The remaining 28% of
under-5 deaths occurred in the age group 1-5 years (Appendix A, Table
A2). For the most part Child PIP only collects data on children who
are admitted to paediatric wards. It is of concern that so many
newborns are admitted (and die) in paediatric wards which are generally
not designed or adequately staffed to provide specialised care for ill
newborns. Furthermore, it is increasingly recognised that all ill
newborns should be cared for in nurseries or neonatal units.
LEADING CAUSES
OF DEATH BY AGE
CATEGORY
As outlined above, five causes of death accounted for 77.2% of all
deaths. The contribution of these conditions was highest in infants
between one month and one year of age and in children between 1 and
5 years of age. In these respective age groups 82.6% and 74.2% of all
deaths were due to these conditions.
6
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Figure 3. Age distribution of leading causes of deaths: 2005-2009
The contribution of each cause of death varied across the age
categories. Septicaemia was the leading cause of death in the newborn
period accounting for 27.7% of deaths in this group of infants. ARI
was the leading cause of death in children between 1 month and 1 year
of age (37.4% of deaths), whilst diarrhoea was the most common cause
of death in children between 1 and 5 years (22.5% of deaths). TB,
which accounted for 20% of deaths in children between 5 and 18 years,
was the leading cause of death in older children (Appendix A, Table
A10).
Nutritional status
Just less than 35.0% of children who died had severe malnutrition,
whilst a further 29.5% were classified as being underweight-for-age.
Less than 30% of deaths occurred in children with normal or above
average weight (Appendix A, Table A6).
7
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Figure 4. Nutritional status of children who died: 2005-2009
The most common causes of death in children with severe
malnutrition who died were diarrhoea (23.6% of deaths), ARI (23.6%)
and septicaemia (23.4%), whilst the proportion of children with severe
malnutrition was particularly high amongst children who died from
diarrhoea (39.2%), septicaemia (49.4%) and TB (49.8%), as shown in
Figure 5. More than 80% of HIV-infected children who died had
severe malnutrition or were underweight-for-age (Appendix A, Table
A14).
Figure 5. Nutritional status and leading causes of death: 2005-2009
8
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
The contribution of severe malnutrition was most prominent in
children 1 to 5 years, being present in 53.1% of children in this age
group who died.
HIV and AIDS
CONTRIBUTION TO
MORTALITY
HIV testing indicated that more than half of children who died had
evidence of HIV infection or exposure (in the absence of a positive or
negative confirmatory test). The HIV status of 34.9% of children who
died was not known, whilst only 14.1% of children who died tested
negative (Appendix A, Table A7). Amongst HIV-infected children
who died, 19.7% were assessed as having Stage III disease and 67.5%
Stage IV disease. Only 3.1% of these children had Stage I or Stage II
disease whilst the clinical stage was unknown in 9.8% (Appendix A,
Table A15).
These figures confirm the huge burden of HIV infection in children
and highlight the need for ongoing strengthening of efforts to prevent
HIV infection in children, and to identify and treat those who are
infected.
Figure 6. HIV status and cause of death: 2005-2009
HIV status and cause of death are shown in Figure 6. Unsurprisingly
children dying from TB and with severe malnutrition were most likely
to be HIV-infected (54.7% and 43.5% respectively). Fifty-nine percent
of children who died of ARIs were HIV-infected or exposed,
9
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
compared with 47.3% of children who died from diarrhoeal disease
(Appendix A, Table A16).
The three leading causes of death amongst children who were known
to be HIV-infected were ARI (28.7%), septicaemia (17.7%) and
diarrhoeal disease (17.7%). The five leading causes of death accounted
for 81.2% of deaths in HIV-infected children, which highlights the
association between HIV infection and common childhood conditions
in children.
Infants accounted for 45.3% of deaths amongst HIV-infected children
and 63.4% of deaths amongst HIV-infected or exposed children
(Appendix A, Table A11).
IDENTIFICATION
AND MANAGEMENT
OF HIV INFECTION
IN CHILDREN
Guidelines for identifying and treating children with HIV infection
have changed substantially during the period 2005 to 2009. Every
healthcare encounter with a child should be used as an opportunity to
assess the child’s HIV status. Early identification of HIV-infected
children and provision of appropriate treatment (including
antiretroviral therapy) is critical to saving lives. It is therefore of
concern that the HIV status of 35% of children was not known. The
HIV status of more than half of newborns was not known, as
compared with less than 20% in children older than five years of age
who died (Appendix A, Table A18).
Data on testing rates in children for each year are presented in Figure 7.
The percentage of children who were HIV-infected or exposed has
remained relatively constant with approximately half of the children
who died falling into this category. The percentage of children whose
HIV status was not known decreased slightly, whilst the percentage of
children who died and HIV-negative has increased from 8.7% to
17.6%. This may indicate an improving testing rate for HIV among
hospitalised children (Appendix A, Table A17).
Only 15.2% of children with documented Stage III or IV HIV disease
who died had received antiretroviral treatment (Appendix A, Table
A20).
It should be noted that these data only reflect HIV prevalence in
children who died and not the prevalence of HIV infection in the
10
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
general population. The data highlight opportunities where care which
is known to improve the health of children was not provided.
Figure 7. HIV status of children who died: 2005-2009
PREVENTION OF
HIV INFECTION
THROUGH PMTCT
Efforts to reduce perinatal transmission of HIV infection through
prevention of mother-to-child transmission (PMTCT) programmes
have also expanded during this period. Data on uptake of prophylactic
antiretrovirals during the perinatal period amongst children who died
are shown in Figure 8. In 53.5% of child deaths, no data were available
regarding access or uptake of PMTCT. The mother was known to be
HIV-negative in 15.1%, whilst prophylaxis was given in 17.3% of cases
and not given in 14.1% of cases (Appendix A, Table A19). Whilst the
trends shown in Figure 8 are encouraging with more mothers testing
negative and prophylaxis being given more often than previously, the
data once more highlight the many missed opportunities for
prevention of HIV infection in children who subsequently died.
11
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Figure 8. Uptake of perinatal ARVs amongst children who died: 2005-2009
Contact with the health system
REFERRAL
Although Child PIP focuses on deaths that occur in hospital,
information on other parts of the health system is also collected to
identify and address gaps at all levels of care. Data on whether children
who die were referred are shown in Appendix A, Table A12. Half of
children (49.6%) who died were not referred, whilst 42.5% were
referred. These data suggest that caregivers of ill children often bypass
primary health care (PHC) facilities. Referrals came from PHC facilities
(24.6%), another hospital (10.4%) and private practitioners (6.9%)
(Appendix A, Table A3).
READMISSION
Approximately one fifth (20.1%) of children who died had been
previously admitted, whilst 64.3% had not. A higher proportion of
children who died from TB and chronic diarrhoea had previously been
admitted (29.8% and 29.4% respectively). This may reflect the chronic
nature of these conditions, but also suggests that the conditions had
not been adequately managed during the previous admission
(Appendix A, Table A12).
Thirty-four percent of child deaths occurred within 24 hours after
admission. A significantly higher percentage of deaths due to diarrhoea
(44.2%) occurred within this period, with the figure for deaths due to
acute diarrhoeal disease being particularly high (49.6%). These deaths
account for 26.8% of all deaths that occurred within 24 hours of
admission. The vast majority of these deaths result from dehydration
12
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
and highlight the ongoing need for strengthening interventions to
prevent and treat dehydration in children with diarrhoeal disease at
home, at PHC facilities and within hospital settings (Appendix A,
Table A13).
Social context
PRIMARY
CAREGIVER
The mother was the primary caregiver of 72.9% of children who died.
In the remaining child deaths, the grandmother was the caregiver in
10.9% of cases and the father in just less than 1% of cases. The
primary caregiver was unknown in 11.7% of child deaths (Appendix A,
Table A5).
MOTHERS’
Child PIP collects data on the mother’s well-being. The mother was
recorded as being alive and well in 72.6% of child deaths, as having
died in 6.2% and as being sick in 8.1%. The mother’s health was not
recorded in 13.1% of deaths (Appendix A, Table A5).
WELL-BEING
FATHERS’
WELL-BEING
Information regarding the father’s well-being was not known in 62.1%
of cases. The fathers of 32.6% of children who died were alive and
well, whilst 3.6% had died and 1.6% were recorded as being ill
(Appendix A, Table A5).
Information about the quality of healthcare which
children received
Quality of records
A folder with adequate records and notes in was available in 55.1% of
deaths, whereas notes were found to be incomplete in 42.5% of cases.
Folders were not available, or the quality of records was unknown in
just over seven percent of audited deaths (Appendix A, Table A25).
Was the death avoidable?
Approximately one-quarter (26%) of deaths were considered to be
avoidable. A further quarter (24.4%) of the deaths were considered
unavoidable, whilst in 34.8% of deaths the reviewer/s were unsure.
This pattern was consistent across all of the leading causes of death
with the exception of deaths due to diarrhoeal disease, where a higher
percentage of deaths (34.9%) were assessed as being avoidable
(Appendix A, Table A21).
13
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Modifiable factors
A total of 53 328 modifiable factors were identified during audit of the
19 295 deaths. On average almost three (2.8) modifiable factors could
be identified for every child who died. The average number of
modifiable factors per death has increased over time, and as previously
noted this may reflect declining quality of care, but may also reflect
greater awareness of standards of care and the need to strive towards
meeting these standards. The average number of modifiable factors
was higher for deaths due to diarrhoeal disease with an average of 3.6
modifiable factors per death being identified.
Modifiable factors can be categorised in three ways. They can be
categorised according to (i) the key health service activity involved; (ii)
where the modifiable factor occurred; and (iii) who was responsible.
Overall findings related to modifiable factors are presented here, whilst
more detail regarding modifiable factors related to each of the leading
causes of death will be described in the following chapters.
Key health service activities
Modifiable factors related to care-seeking and compliance accounted
for 21.4% of all modifiable factors. Modifiable factors related to clinical
management accounted for 11.1%, assessment for 10% and
monitoring for 9.9% (Appendix A, Table A24)
Where the modifiable factors occurred
Approximately one quarter of modifiable factors occurred in the wards
(27.2%) and in the accidents and emergency (A&E or casualty units
(23.6%). Modifiable factors occurring at clinic or out-patient
department (OPD) levels accounted for 14.4% of the total, with a
further 1.5% associated with transit (Appendix A, Table A26).
Modifiable factors thus occurred in the health system in two thirds of
cases, while home or community level modifiable factors accounted for
33.2% of the total. A similar distribution with regard to the place at
which modifiable factors occurred was evident for each of the leading
causes of death(Appendix A, Table A23).
WARD
The five most frequently listed modifiable factors that occurred in the
ward are shown below in ranked order. Most were related to lack of
facilities (especially high care and intensive care units (ICUs) and staff
14
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
(professional nurses and experienced doctors) or gaps in clinical care.
ADMISSION AND
EMERGENCY CARE
Table 3. Leading modifiable factors occurring in Wards: 2005-2009
Ranked modifiable factors in ward
% of deaths
Lack of high care and/or ICU facilities for children
Inadequate investigations in ward
Lack of professional nurse in children's ward 24 hours a day
Inadequate monitoring of blood glucose in ward
Inadequate monitoring of respiratory rate &/or oxygen saturation in ward
4.2%
3.8%
3.4%
2.8%
2.7%
The initial care and assessment of ill children usually takes place in a
designated area, known as casualty or A&E. Almost one-quarter of
modifiable factors occurred during this phase of care, which highlights
an important gap in the care provided to children. Children usually
spend a relatively short time in these units, yet the care that they do
receive impacts on their outcome, especially in the case of critically ill
children. Most of the modifiable factors related to clinical personnel,
and there is an urgent need to improve the quality of care provided for
children (often by generalist doctors and nurses) in these units. The
top five modifiable factors identified in A&E are listed below.
Table 4. Leading modifiable factors occurring in A&E: 2005-2009
Ranked modifiable factors in A&E
Inadequate notes on clinical care (assessment, management, monitoring)
Not classified as critically ill despite presence of danger signs at A&E
Inadequate treatment for shock (fluid type, amount, rate; intraosseus line)
Blood glucose not monitored in child with danger signs at A&E
Lack of Intensive and High Care beds in own, or higher level hospital
TRANSIT OR AT
REFERRAL FACILITY
% of deaths
7.0%
6.4%
4.8%
3.9%
3.0%
This set of modifiable factors has recently been added to Child PIP to
identify gaps in the care which children receive with regard to referral
and during transit. Although the numbers are relatively small, they
highlight gaps and problems with regard to movement of children
from one facility (or health provider) to another facility.
Table 5. Leading modifiable factors occurring during transit: 2005-2009
Ranked modifiable factors during transit
% of deaths
No or delayed referral to higher level
Inappropriate care or late referral from private sector/general practitioner
Inadequate ambulance service from health facility to receiving hospital
Severity of child’s condition incorrectly assessed at referring facility
Inadequate referral letter from referring facility
15
0.8%
0.7%
0.6%
0.4%
0.2%
S A V I N G
CLINIC OR OPD
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
The most common modifiable factors identified in relation to care
which children received at primary health care clinics or other
outpatient settings are shown below in ranked order. Although the
number of modifiable factors is relatively low, this may hide significant
problems given the fact that many children who subsequently die have
bypassed PHC facilities and reported directly to hospital.
As in other settings, the most common modifiable factors relate to
failure to follow well-recognised standards of care such as the
Integrated Management of Childhood Illness (IMCI) guidelines. Failure
to identify and refer children with growth faltering or other nutritional
problems is of particular concern. Almost all PHC facilities have
professional nurses who have been trained in IMCI. These modifiable
factors highlight the ongoing need for ensuring that the IMCI case
management process is correctly and consistently used in the
assessment and treatment of children.
Table 6. Leading modifiable factors occurring in clinics or OPD: 2005-2009
Ranked modifiable factors in clinics or OPD
% of deaths
IMCI not used for patient assessment at clinic/OPD
Inadequate notes on clinical care (assess, classify, treat) at clinic
IMCI not used for case management at clinic/OPD
Delayed referral for severe malnutrition, weight loss, or growth faltering
Child's growth problem inadequately identified or classified
HOME AND
COMMUNITY
3.8%
3.1%
3.0%
2.6%
2.4%
Identification of modifiable factors that occur at home contribute to
understanding child deaths; however it is important to emphasise that
caregivers are not being blamed for these deaths. It should also be
noted that many home and community modifiable factors are broader
in scope than health system factors, they are listed far more frequently
than those associated with health system factors. For example, the
modifiable factor “delay in seeking care” encompasses a wide range of
factors and may indicate difficulties in overcoming a range of
geographic, transport, financial and other barriers to accessing care.
However, delays in seeking care and caregivers not recognising the
severity of illness are identified as modifiable factors in a high
percentage of child deaths, and this highlights the need for better
community-based care and empowerment of caregivers. The third
most common modifiable factors relates to children not being
provided with adequate food at home places the spotlight on social
16
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
determinants of illnesses, such as poverty, lack of information, and
poor resources. It is of concern that inappropriate and harmful home
or traditional treatments are identified as modifiable factors in a high
proportion of child deaths.
Table 7. Leading modifiable factors occurring at home & in community: 2005-2009
Ranked modifiable factor at home
% of deaths
Caregiver delayed seeking care
Caregiver did not recognise danger signs/severity of illness
Child not provided with adequate (quality and/or quantity) food at home
Inappropriate treatment given at home with negative effect on the child
Insufficient notes on home circumstances or child's health history
26.1%
16.5%
13.8%
6.4%
5.3%
Who was responsible (Appendix A, Table A22)
CLINICAL
PERSONNEL
Approximately half (54.4%) of modifiable factors were related to
clinical personnel. This finding was consistent for all causes of death
with clinical personnel being responsible for at least half of modifiable
factors for each of the leading causes of death. The most common
personnel related modifiable factors are shown below.
Table 8. Leading modifiable factors related to clinical personnel: 2005-2009
Modifiable factor
% of deaths
Inadequate history taken at A&E
Inadequate investigations (blood, x-ray, other) at A&E
Insufficient notes on home circumstances or child's health history
Inadequate physical examination at A&E
Appropriate antibiotics not prescribed at A&E
7.0%
6.4%
5.3%
4.8%
3.9%
It is noteworthy that at least four of the five leading personnel-related
modifiable factors occurred in the emergency assessment and care
setting, suggesting that improved training and supervision of personnel
in this area is a key intervention for improving the care that children
receive within the health system.
ADMINISTRATORS
AND MANAGERS
Managers or administrators were responsible for 13.6% of modifiable
factors. The top five modifiable factors all related to the availability of
infrastructure and staffing resources. This underscores the need for
having more clearly defined norms and standards for provision of
healthcare services and for ensuring an equitable distribution of
available resources.
17
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Table 9. Leading modifiable factors related to administrators & managers: 20052009
Modifiable factor
% of deaths
Lack of high care and/or ICU facilities for children
Lack of professional nurse in children's ward 24 hours a day
Lack of experienced doctors (post-community service), for children's ward
Inadequate resuscitation area and/or trolley in ward
Lack of intensive and high care beds in own or higher level hospital
CAREGIVERS
4.2%
3.4%
2.6%
2.1%
1.8%
Caregivers were thought to be responsible for approximately one third
(32.0%) of modifiable factors.
Table 10. Leading modifiable factors related to caregivers: 2005-2009
Modifiable factor
% of deaths
Caregiver delayed seeking care
Caregiver did not recognise danger signs/severity of illness
Child not provided with adequate (quality and/or quantity) food at home
Inappropriate treatment given at home with negative effect on the child
Caregiver took child to clinic infrequently
26.1%
16.5%
13.8%
6.4%
4.4%
Conclusion
Pneumonia and diarrhoeal disease remain the leading causes of death
in South Africa, accounting for half of all deaths in children. A high
proportion of deaths from these conditions are preventable. Preventive
and promotive interventions can prevent many cases, and case fatality
rates can be reduced if children receive treatment based on simple yet
effective case management guidelines.
Ensuring that all children have access to a simple package of preventive
and curative interventions is likely to result in a substantial decrease in
the number of deaths. Data on avoidable deaths and modifiable factors
suggest that an especially high proportion of deaths from acute
diarrhoea are avoidable, and interventions to prevent these deaths
should be prioritised. Introduction of rotavirus vaccine into the routine
immunisation programme should play a role in reducing these deaths.
However, efforts to improve water and sanitation, and management of
children with diarrhoea (especially prevention and treatment of
dehydration), at home, at PHC and hospital levels are also required.
Guidelines for assessing, managing and monitoring of children with
dehydration are available, and health service managers must ensure that
staff in their facilities competent in the use of these guidelines.
18
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
HIV infection continues to contribute significantly to childhood
deaths, and substantial reductions in child mortality are unlikely to be
seen without a reduction in the prevalence of HIV infection in children
The Child PIP data demonstrates some improvements in testing rates
and uptake of PMTCT in patients who died, but many missed
opportunities persist. Whilst the scale-up of PMTCT and antiretroviral
therapy programmes has prevented many child deaths, ongoing
strengthening is required to ensure that all mothers and children
benefit from these programmes. Likewise undernutrition remains an
important contributing factor in many childhood deaths, and it is of
great concern that food insecurity remains one of the leading
community or home modifiable factors.
In addition to HIV-infected and undernourished children, the data
highlight two vulnerable groups of children. Newborn infants account
for more than 5% of children admitted to paediatric wards, yet these
wards are seldom specifically equipped or staffed to care for such
young babies. The data also identify children requiring emergency care
as a vulnerable group. As described above, more than one third of
deaths occur within 24 hours of admission, and a disproportionate
number of modifiable factors occur in the emergency care setting.
Assessment and management of children in this stage appeared to be a
particular problem, and highlights the need for introduction of a
standard approach to provision of emergency care. The Emergency
Triage, Assessment and Triage approach, which was developed by the
World Health Organisation and adapted for use in South Africa,
provides an excellent system for ensuring adequate care during this
period.
Analysis of modifiable factors continues to identify many such factors
at all levels of the health system, and highlight the need to strengthen
the overall health system. Specific actions that are required are
addressed in the next chapter.
The number of hospitals that use the Child PIP process has continued
to expand, and it is hoped that this number will continue to grow as
efforts to reduce child deaths in South Africa expand and gain
momentum.
19
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Chapter
2
Call to Action
Lesley Bamford, with Barnesh Dhada; Ajay Chiba; Tim de Maayer; Kim Harper; Neil McKerrow;
Mark Patrick; Cindy Stephen and Anthony Westwood
The Child PIP process not only provides us with valuable data
regarding child deaths, but also points us in the direction of how
identified gaps can be addressed. Whilst the strength of audit processes
such as Child PIP lies in their potential to inspire teams of healthcare
workers at local levels to develop local solutions to identified problems,
analysis of the data also highlights areas which require particular
attention at a national level.
Recommendations
Below are a set of recommendations that have arisen from analysis of
the overall Child PIP data and of each of the main causes of mortality
in young children. The recommendations are targeted at different key
functions within the health sector and are thus divided into four
sections, namely policy, management and administration, clinical
practice and education.
Policy
Develop a co-ordinated and intersectoral response to child
poverty and ill health, with special emphasis on improving
water supply and sanitation; improving food security through
the creation of employment; addressing barriers to accessing
social grants; and providing supplementary feeding
programmes.
Develop a national plan for improving the quality of
emergency care for critically ill children in South Africa. This
20
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
should target all levels of care within the District Health
System to build capacity for health workers, improve systems
for provision of emergency care within health; and ensure that
adequate resources are allocated.
Define and implement a package of perinatal and post-natal
care of newborn babies (including home post-natal visits and
appropriate hospital care for sick newborns), as well as
strengthening of primary maternal and obstetric care.
Clarify the policy regarding infant feeding and provision of
formula at healthcare facilities and communicate this clearly to
all role players.
Review and improve the paediatric section of the National TB
Guidelines and ensure that healthcare workers have access to
clear and practical guidance with regard to assessment,
diagnosis and management of children with suspected or
proven TB.
Develop and implement a national guideline for improving
the quality of care for children with severe sepsis and septic
shock. The complex clinical challenges posed by the major comorbidities of malnutrition, HIV infection and tuberculosis
(TB) should be considered when devising the strategy.
Management and Administration
Appoint district/regional specialist teams to support maternal
and child health service delivery at district level.
Ensure the availability of infrastructure, equipment, staffing,
and record keeping and monitoring systems for delivering
emergency care, based on agreed country standards.
Ensure high care facilities for severely ill children which are
adequately equipped and staffed.
Provide hospital facilities that are adequately staffed and
equipped to provide care for newborn and small babies, into
21
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
which all newborns (including those admitted from home)
must be admitted, instead of to general paediatric wards.
Ensure that paediatric Essential Drug List drugs are available
at healthcare facilities. Ensuring availability of paediatric
antiretroviral therapy is particularly important.
Ensure that supplies used in management of severe
malnutrition (pre-mixed F75 or F100 or the combined
vitamin and mineral mix) are available at all hospitals, at all
times.
Strengthen the vaccine supply chain in order to improve
coverage with all recommended immunisations.
Clinical Practice
Scale up and strengthen paediatric nurse-initiated antiretroviral
therapy services by means of a mentorship programme.
Use each and every healthcare encounter with a child as an
opportunity to ascertain his/her HIV status.
Improve the recognition and management of children with
malnutrition through universal use of Integrated Management
of Childhood Illness guidelines at clinic level, and national
guidelines for severe malnutrition (based on WHO Ten Steps)
at hospital level.
Improve hospital care through obligatory use of available
guidelines and protocols. Issues which require particular
attention include:
Improved emergency care including better management of
children with sepsis
Better management of children with acute diarrhoeal
disease
Correct and early use of oxygen in children with ARI
22
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
The establishment and general use of Guidelines for
Neonatal Care, with special emphasis on perinatal HIV
care and emergency care for small babies.
Education
Empower community caregivers to provide home care for
children, and to recognise danger signs of serious illness that
require immediate care at health facilities.
Ensure that all health workers are aware of the importance of
exclusive breast-feeding and are able to advise mothers on
optimal feeding choices and good nutrition.
Ensure that all health professionals dealing with children are
competent to identify and manage those conditions that
account for most deaths in children, and that undergraduate
medical, nursing and allied health curricula offered in South
African institutions are relevant to the health needs of the
country’s children.
Ensure competence in emergency triage and treatment by
providing certified emergency care training, based on a
national standard, to doctors, nurses and paramedics
(ambulance personnel).
Ensure that healthcare professionals understand the
contribution of neonatal deaths to under-five mortality and
are competent in providing the specialised care required by
small babies.
We believe that implementation of these recommendations can go a
long way to reducing child mortality rates in South Africa and to
delivering on what has come to be understood as one of the key
functions of our health system and health sector.
23
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
PART TWO
PRIORITY CONDITIONS AND
OPPORTUNITIES FOR CHANGE
Chapter
3
HIV
Neil McKerrow
BA, MBChB, DCH (SA), FCPaed (SA), MMed(Paed)
Chief Specialist & Head: Paediatrics & Child Health
Pietermaritzburg Metropolitan Hospitals Complex
Department of Paediatrics, NRMSM, University of KwaZulu-Natal
Abstract
An antenatal HIV seroprevalence of 29.4% poses a major risk for
HIV infection for babies and by 2009 South Africa was home to an
estimated 280 000 HIV-positive children.
Although HIV infection remains a major factor in the health of
children of all ages, a significant proportion (57.7%) continue to die
without their HIV status being established despite known
exposure to the infection (22.9%) and a clinical picture compatible
with advanced HIV disease (41.4%). More than half (51.0%) of
childhood deaths are HIV associated with the majority of these
occurring before the age of five years even though HIV-infected
children tend to die at a later age than their non-infected
counterparts. Tuberculosis and severe acute malnutrition are the
two diseases most commonly associated with HIV.
Obstacles to the appropriate care of infected children occur at all
stages from the home to the hospital albeit with a slightly lower
frequency than in other diseases. The most significant of these is
the failure to detect infected children early thereby limiting our
opportunities for appropriate early intervention.
24
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Key Recommendations
The HIV status of all children attending the health service must be
established and documented in their clinical record and/or Road-toHealth Card.
The integration of midwifery and antiretroviral therapy (ART) services
is key to the achievement of a more effective prevention of motherto-child HIV transmission (PMTCT) programme.
Clarity is urgently required with respect to national policy regarding
infant feeding choices and the provision of free infant formula.
The integration of well-baby and HIV services is critical in order to
promote early access to ART for those children who require this.
Continuous training, ongoing supervision and mentorship
programmes are required in all districts to ensure the effective
implementation of clinical guidelines and Nurse Initiated
Management of Antiretroviral Therapy (NIMART).
Introduction: Why is HIV important?
South Africa remains the epicentre of the global HIV epidemic and the
20th Antenatal HIV Seroprevalence Survey1 estimated that in 2009 the
HIV prevalence among pregnant women was 29.4%. This projected to
a prevalence of 17.8% in the general population with an estimated 5.2
to 5.63 million HIV-infected adults and children in the country.
The Human Sciences Research Council’s 2008 household survey2
estimated that the HIV prevalence was 2.9% in children below the age
of 18 years and 3.3% in those below 5 years. This means that with an
estimated 280 000 HIV-infected children South Africa had the greatest
number of HIV-infected children in the world. Although more than 57
000 of these children were on ART by 2009 the associated burden of
1
Department of Health. National Antenatal Sentinel HIV and Syphilis Seroprevalence Survey, 2009. National
Department of Health, Pretoria. 2010.
2
B Shisana O, Simbayi LC, Rehle T, Zungu NP, Zuma K, Ngogo N, Jooste S, Pillay-Van Wyk V, Parker W,
Pezi S, Davids A, Nwanyanwu O, Dinh TH and SABSSM III Implementation Team. South African National
HIV Prevalence, Incidence, Behaviour and Communication Survey, 2008: The health of our children. Cape
Town: HSRC Press. 2010.
25
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
disease remained high and the contribution of HIV to childhood
mortality significant.
What does the data say about the health profile of
children who died of HIV?
During the period 2005 to 2009 the HIV status of the 19 295 children
whose deaths were subject to review was known for only 42.3% 28.2% were HIV-infected and 14.1% uninfected. Among the
remaining 57.7%, the HIV status was unknown even though 22.9% of
these children were known to have been born to HIV-infected
mothers.
This pattern was repeated among children under-5 years of age where
25.6% were confirmed to be HIV-positive, 14.4% were known to be
HIV-negative and the HIV status of the remaining 60.0% was
unknown although 24.6% were known to have been exposed.
For a country recognised to have the worst HIV epidemic in the world
the fact that almost 60% of critically ill children have died without an
evaluation of their HIV status is a sad indictment of the standard of
care provided to our children.
Table 1. Association of HIV status with leading causes of death
Cause of death
Tuberculosis
Severe acute malnutrition
Sepsis
Meningitis
Acute respiratory infection
Diarrhoeal diseases
Five leading causes of death
All deaths
% HIV positive
54.7
43.5
30.7
30.1
28.1
23.4
29.6
28.2
% HIV negative
13.5
12.3
16.5
18.6
8.6
11.1
11.9
14.1
% HIV unknown
31.8
44.2
52.8
51.3
63.3
65.5
58.5
57.7
Table 1 shows that the proportion of children who were HIV-infected
varied markedly between the five leading causes of death, ranging from
54.7% of children who died from tuberculosis to less than a quarter
(23.4%) in those dying from diarrhoeal diseases. The ratio of HIVnegative children was fairly similar in all groups with a spread from
8.6% to 18.6% and the status was undetermined in the majority of
children. This table suggests that when one looks for HIV infection we
26
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
are likely to find it and that the reported contribution of HIV to
childhood deaths may be grossly underestimated.
Considering both HIV exposure and confirmed infection, 51.0% of all
childhood deaths were associated with HIV with a similar proportion
in the under-5 deaths (50.5%). The proportion of HIV associated
deaths was lowest in the first 28 days of life (36.3%) and highest
between 5 and 13 years (58.0%).
The contribution of HIV infection to childhood deaths increased with
age and only 4.6% of children who died in the first 28 days of life were
HIV-positive compared with 25% in the children under-5 years and
over 50% in those over five.
As is to be expected, the opposite pattern is seen in the proportion of
children who were HIV exposed. Among children who died below one
year of age 31% were HIV exposed compared with 25.2% among the
under-5 deaths and 3.1% in those over 5 years. The number of
exposed children older than 5 years who had not had a definitive HIV
test is worrying.
The age distribution of HIV-infected deaths appears to have shifted to
the older child. Newborns accounted for only 0.9% of all HIV-positive
deaths, but for 5.5% of all deaths. A similar pattern is evident in
infancy during which 45.3% of HIV-infected deaths occurred,
compared with 63.2% of all deaths, as well in the under-5 age group
(79.0% vs 87.5%). Those older than 5 years accounted for 19.7% of
HIV-infected deaths compared to only 10.6% among all deaths.
Although only 28.2% of children who died were confirmed HIVinfected, 45.6% of them underwent clinically staging of HIV disease
and it is worrying that an obvious clinical suspicion of HIV so seldom
resulted in confirmatory testing.
Of equal concern is the observation that only 90.2% of known HIVpositive children and only 59.6% of HIV-exposed children were
staged. Among the former group 87.2% were assessed with advanced
disease, stage 3 or 4.
Table 2 shows the pattern of clinical staging in children dying from the
five leading causes of deaths. The similarity in the proportion of
27
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
children who were clinically staged and of those who were assessed
with advanced disease suggests that the clinical picture rather than the
HIV status is the deciding factor in the decision to stage a child and
begs the question why this does not also lead to an HIV test. Perhaps
consent for an HIV test was not forthcoming.
Table 2. Clinical staging and different causes of death
Disease
Tuberculosis
Severe acute malnutrition
Acute respiratory illness
Sepsis
Meningitis
Diarrhoeal diseases
Five leading causes of death
All deaths
Staged (%)
68.6
63.9
52.2
50.9
42,6
39.1
49.0
45.6
HIV positive (%)
54.7
43.5
28.1
30.7
30.1
23.4
29.6
28.2
Stage 3 or 4 (%)
67.0
62.4
47.4
46.3
38.5
34.7
44.6
41.4
Childhood deaths associated with HIV infection tend to be associated
with a more prolonged hospital stay than non-HIV associated deaths.
Only 21.3% of HIV-positive children died within the first 24 hours
following admission compared with 31.3% for all deaths. Similarly,
43.4% occurred between 24 hours and 3 days compared with 57.0%
for all deaths. In contrast, 22.3% of the HIV-infected deaths occurred
more than 14 days after admission compared with only 12.0% for all
deaths. This is similar to the proportion of late deaths seen with
tuberculosis.
Finally, there was a high rate of previous admissions among the HIVinfected children who died. The re-admission rate for HIV-infected
children who died was 34.5%, which was higher than that for other
chronic diseases such as severe acute malnutrition (28.1%), tuberculosis
(29.4%) and chronic diarrhoeal diseases (29.8%) and much higher than
that for all deaths combined (20.1%). This suggests a missed
opportunity for appropriate earlier intervention.
What does the data say about the quality of care
received by the children who died of HIV?
The single most striking feature in the care of our children was the
number who died with an unknown HIV status. Although in 3.1% of
deaths this was because the parent or caregiver refused consent for an
28
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
HIV test, there is no data on the number in whom the healthcare
worker failed to explore this possibility.
The average number of modifiable factors per HIV-infected child (2.4)
was lower than that for all deaths (2.8) as well as that for any one of the
five leading causes of death which ranged from 2.4 for meningitis to
for 3.6 for diarrhoeal diseases.
Table 3. Number, site and level of modifiable factors for various causes of death
MFs per death
% Admin
% Clinical Pers
% Caregiver
% Ward
% A&E
% Clinic
% Home
ARI
2.9
13.3
57.0
29.6
26.3
26.3
15.6
30.6
DD
3.6
11.3
57.3
31,4
28.1
26.8
11.6
32.6
Sepsis
2.6
13.7
49.6
36.7
26.2
17.7
16.5
38.1
Menin
2.4
17.0
52.9
30.1
25.7
22.6
17.3
31.9
TB
2.5
16.5
51.6
31.9
26.8
17.8
20.3
33.6
SAM
3.0
11.9
50.2
37.8
25.5
18.2
16.8
38.5
All
2.8
13.6
54.4
32.0
27.2
23.6
14.4
33.2
HIV+
2.4
13.5
50.1
36.4
26.0
17.3
18.9
36.5
While there was little difference in the proportion of modifiable factors
occurring at the different sites in the health system between HIV
infected children and other leading causes of death, the overall
impression is that in both HIV and other less “acute” diseases, such as
sepsis, tuberculosis and severe acute malnutrition (SAM), there was a
shift towards more modifiable factors in the home and ambulatory
setting than after admission to hospital. A similar grouping emerges
with respect to the person responsible for the modifiable factors with a
larger contribution by the clinician and less by the caregiver in the more
“acute” diseases.
The nature of the modifiable factors found in deaths linked to HIV
infection was similar to that for other causes of death. At the home
these included a delay in seeking care, failure to recognise danger signs
or the severity of the illness, inadequate nutrition and infrequent clinic
visits. Refusal by the caregiver to consent to an HIV test was reported
to a variable degree in all categories of death except diarrhoeal diseases
where it was not evident at all.
At the clinical or outpatient setting the more common modifiable
factors included failure to arrive on the day of referral, although this
was much less prominent than with other causes of death, delayed
29
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
referral for growth faltering and an inadequate assessment of growth
problems.
Modifiable factors occurring during transit between different levels of
care in the health sector were similar across all causes of death and
included delayed referral to a higher level of care as well as
inappropriate care or late referral from the private sector.
There was no difference in the nature of modifiable factors seen on
arrival at the hospital or in the ward between the different causes of
death. However, of concern is the proportion of deaths in HIVinfected children that were associated with an inadequate assessment
and inadequate management during a previous admission, which
suggests previous missed opportunities for effective care.
What is the South African national standard of caring for
children with HIV?
National policies and guidelines for both PMTCT and the care of HIV
infected children have been developed. Furthermore these are all
regularly revised in line with emerging knowledge and evolving
international practices.
PMTCT
3
The cornerstone of the PMTCT programme is the integration of the
following four elements into routine health services:
Primary prevention of HIV, especially among women of
childbearing age;
Preventing unintended pregnancies among women living with
HIV;
Preventing HIV transmission from a woman living with HIV
to her infant; and
Providing appropriate treatment, care, and support to women
living with HIV and their children and families.
3
National Department of Health. Clinical Guidelines: PMTCT (Prevention of Mother-to-Child
Transmission). Pretoria. 2010.
30
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
The specifics of the PMTCT programme include modified antenatal
and midwifery practices, antiretroviral prophylaxis or treatment for all
pregnant woman, antiretroviral prophylaxis for the infant immediately
after birth and for as long as the infant is breastfeeding, an informed
infant feeding choice and a routine HIV polymerase chain reaction
(PCR) test at six weeks of age.
Guidelines for the care of HIV exposed or infected children
4
These guidelines include the early identification of HIV infection and
the initiation of ART for all HIV-infected infants as well as for
symptomatic or immune incompetent children.
Are there controversies within the accepted South
African national standard for childhood HIV care?
The 2010 National Guidelines for PMTCT and those for the
management of HIV infection in children have addressed most of the
recent controversies in both areas. There are, however, two
outstanding issues with respect to PMTCT that should be addressed as
soon as possible.
The first is the inclusion of nevirapine in the regimen for lifelong highly
active antiretroviral therapy (HAART) in eligible pregnant women.
Reports to the National Committee on the Confidential Enquiry into
Maternal Deaths5 suggest that this drug is associated with severe skin
reactions and maternal deaths. However, since no severe adverse
events have been reported this matter requires further investigation
before definite changes are made.
The second is the provision of free infant formula to women who elect
to formula feed their babies. The low rate of breast-feeding, the poor
socioeconomic circumstances of many households and the high under5 mortality rate in South Africa all raise questions around the role of
free formula in infant feeding choices. In the absence of a definite
stance by the National Department of Health the KwaZulu-Natal
4
National Department of Health. Guidelines for the Management of HIV in Children. Pretoria. 2010.
5
N Moran, Obstetrician and Member, National Committee on Confidential Enquiry into Maternal Deaths.
Personal communication, March 2011.
31
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
provincial department has already taken a unilateral decision to stop
issuing free formula in an effort to encourage exclusive breast-feeding
with nevirapine cover.
What are the challenges, and what further research is
needed for the successful implementation of national
standards for caring for children with HIV?
The two focal areas in managing HIV infection in children must be the
prevention of mother-to-child transmission and when this fails, the
timely initiation of ART for those who are infected. Current policy
supports both of these initiatives although, as mentioned above, clarity
is required regarding infant feeding choices and the provision of infant
formula.
The effective implementation of this policy does however require a
more equitable share of resources between adult and paediatric
programmes than currently exists. This is especially pertinent in the roll
out of NIMART.
PMTCT programmes are not yet fully effective and District Health
Information System indicators reveal that while almost 90% of
antenatal attendees are tested for HIV and an equally high proportion
of HIV-positive women receive nevirapine, less than 60% of exposed
newborns receive the drug and even fewer have a six week HIV PCR
test. As the intention to offer ART to all HIV-infected infants requires
the early identification of eligible infants this poses a significant
obstacle to access to appropriate care. The challenge is thus to ensure
that systems for the follow up and testing of exposed babies are
strengthened.
The number of children on ART has increased from an estimated 57
000 children in 2009 to close to 130 000 at the start of 2011. It is
unlikely that this rate of enrolment can be sustained without a shift
away from the current practice of doctor-initiated treatment.
Theoretically this should be possible with the introduction of
NIMART although in many areas throughout the country trained
nurses have not yet started treating patients or where they have, this is
limited to the treatment of adult patients. An effective mentorship
programme is required to support such nurses in converting
32
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
knowledge to practice and a phased approach must be adopted,
starting with initiation in adults and the follow-up of stable children on
ART before moving on to the initiation of antiretroviral treatment for
older children and eventually for infants.
Finally it is critical to ensure that NIMART enhances the care of all
children and does not result in a shift of resources from the sick child
to HIV programmes or sick adults.
Summary and conclusion
Between 2005 and 2009, 51.0% of childhood deaths were associated
with HIV, 28.2% of children who died were HIV-infected and 22.9%
were exposed without confirmation of their actual HIV status. The
diseases with the greatest association with HIV infection were
tuberculosis and severe acute malnutrition although these were also the
diseases with the highest proportion of children in whom the HIV
status was assessed. While the HIV status was unconfirmed in a high
proportion of deaths it seems that the more one looked for HIV
infection the higher the rate of infected children.
The age group with the highest proportion of HIV associated deaths
occurred between 5 and 13 years, although the contribution of HIV
infection to death was proportional to age, rising from 4.6% in the first
month of life to more than 50% after the age of 5 years. HIV-infected
children tended to die at a later age than those not infected, with 79%
doing so before the age of five compared with 87.5% of all deaths
occurring by this age.
Clinical staging appears to be conducted according to the clinical
appearance of children rather than their HIV status, and the high
proportion of HIV-infected children with a prior admission suggests
that opportunities for earlier intervention may be being missed.
Although there are fewer modifiable factors associated with HIV
infected deaths these have a similar pattern to other less “acute”
illnesses and their nature does not differ substantially from those of
other illnesses. The failure to consider or to consent to HIV testing is
the most worrying modifiable factor as this oversight precludes the
ongoing informed management of HIV infected children.
33
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Chapter
4
Malnutrition
Ajay C Chiba
MBBCh, FCPaed (SA), Senior Specialist
Tim De Maayer
MBBCh, FCPaed (SA), MMed, Specialist
Department of Paediatrics, Rahima Moosa Mother and Child Hospital
Abstract
Malnutrition remains a major co-morbidity for children dying in
South African hospitals. Between January 2005 and December
2009 there were 19 295 audited deaths at the participating Child
PIP sites. Of these, 34 % were severely malnourished. A further
29.5% were underweight for age. The “In-Hospital Mortality Rate”
(IHMR) for “Severe Malnutrition” in children under 5 years was
17.5%. This is appreciably higher than the mortality rate of 3.2% in
children whose weight was normal.
There is a strong association between severe malnutrition and HIV
infection. Nearly half (40.2%) of children under 5 years who died
were severely malnourished and were co-infected with HIV.
Comparatively, 12.9% of children who died from severe
malnutrition were HIV-negative.
There were 19 837 modifiable factors identified in the children
who died with severe malnutrition. Of these, 62.6% were within
the healthcare system: 25.5% were in the wards; 18.3% in
emergency departments and 16.8% occurred at clinic level. Many
of these modifiable factors focussed on failure of implementation
of either Integrated Management of Childhood Illness (IMCI)
guidelines at clinic level or the South African Standard Treatment
Guidelines for severe malnutrition at hospital level.
34
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Reducing mortality from malnutrition will take a multi-faceted
approach. Preventing malnutrition in the first place will require
interventions aimed at improving food security as well as
improved parental education on childhood nutrition. Interventions
aimed at social upliftment are also warranted.
Associated diseases such as HIV and tuberculosis will also need
to be addressed.
Key Recommendations
Weigh and plot weight of all children attending a healthcare facility on
the Road-to-Health Card (RTHC).
Appropriately manage and follow-up (according to IMCI guidelines)
children identified with growth and weight faltering on the RTHC.
Treat children identified with severe malnutrition at primary
healthcare clinics according to IMCI guidelines and timeously refer to
hospitals.
Intensify IMCI training and implementation at all levels.
Assess, investigate and manage children with severe malnutrition at
hospital level. Investigation and management must be in accordance
with the Standard Treatment Guidelines for Hospital Level Paediatrics.
Improve emergency management of severe malnutrition.
Provide ongoing education regarding hospital management of
malnutrition.
Use ongoing audits such as Child PIP to assess quality of care.
Conduct more research in the management of children with severe
malnutrition as a co-morbidity of HIV and tuberculosis (TB) infection.
Ensure that ready-to-use therapeutic foods, pre-mixed F-75 and F-100
as listed in the Standard Treatment Guidelines and Essential Drug List
must be made available throughout all provinces in South Africa.
35
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Introduction
Today, approximately 129 million children younger than five years are
underweight worldwide.1 One in ten children in the developing world is
severely malnourished.1 The first Millennium Development Goal aims
to reduce the prevalence of underweight in under-fives by 50% from
1990 to 2015. While progress towards this goal is on track in many
Asian countries, Africa has made little progress in this regard; and
South Africa is one of 20 countries that has made no progress at all. 1
Recent estimates suggest that 23% of South African children between
1 and 3 years are stunted, 11% are underweight, and 5% are wasted.2
Stunting rightly receives much attention in many publications
addressing undernutrition, as it may lead to impaired learning ability,
low work productivity, and obesity and lifestyle diseases later in life.
However, children who are wasted face a greatly increased risk of death
in the near future. It is estimated that more than one third of deaths in
children under five globally are associated with maternal or child
undernutrition.3
Health profile of children that died with severe
malnutrition
Malnutrition remains a significant co-morbid finding in children who
died at the Child PIP sites. The data analysed included all childhood
deaths in the Child PIP database from January 2005 to December
2009. The malnutrition data therefore represent children who have
died and may not necessarily be representative of the hospital or
community populations. However, by looking at the extreme end of
1
UNICEF. Tracking Progress on Child and Maternal Nutrition. New York, USA, 2009.
http://www.unicef.org/publications/files/Tracking_Progress_on_Child_and_Maternal_Nutrition_EN_1103
09.pdf (Accessed 3 March 2011)
2
Labadarios D. National Food Consumption Survey- Fortification Baseline (NFCS-FB): The knowledge,
attitude, behaviour and procurement regarding fortified foods, a measure of hunger and the anthropometric
and selected micronutrient status of children aged 1- 9 years and women of child bearing age: South Africa
2005. Stellenbosch: Department of Health, 2007.
3
Black RE, Allen LH, Bhutta ZA, Caulfield LE, de Onis M, Ezzati M, et al. Maternal and Child
Undernutrition: Global and regional exposures and health consequences. The Lancet 2008; 371: 243–260.
36
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
the spectrum, it does give insight into problems arising at community
and hospital levels.
For this report, “Severe malnutrition” is defined as weight below 60%
expected weight for age, or if a child is 60% to 80% expected weight
for age with oedema due to a nutritional cause. Hence all children with
kwashiorkor, marasmus or marasmic-kwashiorkor were defined as
having severe malnutrition. Underweight was defined as a weight of
60% to 80% expected weight for age without oedema. At the time of
data collection for this report, the new WHO growth standards were
not available.
The figure below shows all deaths by age category and nutritional
category.
Figure 1. All deaths by age category and nutritional category
From the graph, it can be seen that malnutrition has a significant
impact on childhood mortality. Of all children who died, 29.5% were
underweight for age compared with national prevalence of 8.2%.4 A
further 34% had severe malnutrition (marasmus: 23.4%, kwashiorkor:
6.4%, and marasmic-kwashiorkor: 4.5%) compared with a national
survey prevalence of 2.4%.4 All age groups were affected. In the “28
days to 1 year” age group, 31.9% of children who died were
underweight for age whilst a further 28.7% were severely
4
World Health Organisation. Global Database on Child Growth and Malnutrition. South Africa.
http://www.who.int/nutgrowthdb/database/countries/who_standards/zaf.pdf (Accessed 30 March 2011)
37
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
malnourished. A staggering 53.1% of children who died in the “1 year
to 5 years” age group were severely malnourished and a further 22.7%
underweight for age.
Whilst much focus has been placed on the under-5 population,
malnutrition remains a significant co-morbidity in older children.
Children who died in the 5-13 year group (28.6%) and 13-18 year age
group (30.6%) respectively were underweight for age whilst a further
29.8% and 37.5% respectively were marasmic.
In-hospital mortality rates in children under five years
The overall IHMR among children younger than 5 years was 5.8 deaths
per 100 admissions. The mortality rate for children who were
underweight for age was 9% and those with severe malnutrition 17.5%.
These rates are significantly higher than those of children whose weight
was normal (3.2%) (Table 1). Underweight and severely malnourished
children were 3 and 6.4 times more likely to die than normally
nourished children respectively (OR 2.98; 95%CI 2.85 – 3.11;
p<0.0001 and OR 6.39; 95%CI 6.10 – 6.69; p<0.0001).
Of concern is that 26.8% of admitted children had an “unknown”
weight and of equal concern is that is that in 19% of children who died,
the weight was unknown.
Table 1: In Hospital mortality rate by weight category in children under 5 Years
Weight
Normal
Severe
Underweight
Unknown
category
weight
malnutrition
Admissions 125425 46.4% 48166 17.8% 24450 9.0% 72466 26.8%
Deaths
IHMR
4044
25.8%
3.2%
4352
27.8%
4290
9%
27.4%
17.5%
2974
19.0%
4.1%
Total
270507
100.0%
15660
100.0%
5.8%
Cause of death in children with severe malnutrition
Malnourished children often present to clinics and hospitals with
associated conditions such as HIV, TB, diarrhoeal diseases, pneumonia
and septicaemia. In many cases, it is the illness that leads to
malnutrition through inadequate energy intake coupled with increased
energy requirements. It is often difficult to determine whether it was
the disease or the malnutrition that was the initiating event.
38
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
The malnourished child on the other hand is also at a greater risk for
acquiring most of the conditions mentioned above. Irrespective of
which came first, the malnourished child with an illness is at significant
risk for mortality.
From the figure titled “U5 Deaths: Weight categories for deaths from
leading causes”, it can be seen that children with more chronic diseases
such as chronic diarrhoea or TB are more malnourished than those
with meningitis or pneumonia. In addition, the graph indicates that
malnutrition is an important accompaniment of the top five causes of
childhood deaths (i.e. acute respiratory infections, diarrhoea,
septicaemia, TB and meningitis.
Of all children under-5 dying from diarrhoea, 38.4% were severely
malnourished, but when deaths from chronic diarrhoea were
considered, this proportion rose to 61.4%. Similarly 49.6% of children
dying from septicaemia and 52.6% dying from TB were severely
malnourished.
Figure 2. Under 5 Deaths: Weight categories for deaths from leading causes
Malnutrition and HIV
HIV infection is commonly associated with malnutrition. What is
evident from the figure titled “Weight Category by HIV status” is that
a greater proportion of children dying with severe malnutrition were
39
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
HIV-infected. Of all children under-5 dying with severe malnutrition,
12.9% were HIV-negative compared to 40.2% who were HIV infected
and 21.1% HIV-exposed.
Similarly, of the underweight for age children, 24.2% were HIVinfected compared to 14.1% who were HIV-negative.
More HIV-infected children who died are malnourished than HIVuninfected children.
Of concern is the large percentage of children (23.6%) who died with
severe malnutrition whose HIV result was not known at the time of
death. These children would have either not been tested, or if they
were tested the results were not available at the time of death.
Figure 3. Under 5 Deaths: Weight category by HIV status
Social context of children who died from malnutrition
Children are amongst the most vulnerable members of society. They
are almost solely dependent on an adult to provide them with
appropriate nutrition, shelter and medical care. The mother is often
the individual who bears the brunt of this responsibility.
In children younger than 5 years who died with severe malnutrition, 6.8
% had a mother who was deceased and 13% had a mother who was ill
40
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
(Table 2). Collectively, 20% of children under five who died with
severe malnutrition were at significant nutritional risk as their mother
was unable to be the primary caregiver due to illness or death. When
compared to children with a normal weight, the figures are 1.9% and
4.4% respectively.
This implies that interventions aimed at maternal well-being may be
beneficial.
Table 2: Under 5: Mother’s wellbeing and Weight category
Weight
OWFA
Normal
UWFA
Sev maln
Unknown
Total
Mother
Alive
Dead
Sick
Unknown
No.
%
No.
%
No.
%
No.
%
No.
%
No.
%
122
86.5
4092
84.9
3982
78.4
4033
68.6
675
58.3
12904
75.6
5
3.5
92
1.9
154
3.0
398
6.8
16
1.4
665
3.9
3
2.1
214
4.4
426
8.4
763
13.0
33
2.9
1439
8.4
11
7.8
420
8.7
516
10.2
683
11.6
433
37.4
2063
12.1
Total
141
100
4818
100
5078
100
5877
100
1157
100
17071
100
Quality of care
There were 19 837 modifiable factors identified in all children who
died with severe malnutrition. The table below illustrates where the
modifiable factors were identified. The healthcare system was
associated with 62.6% of identified modifiable factors (Table 3).
Table 3: Modifiable factors for severe malnutrition by place
Modifiable Factor Place
Ward
A&E
Referring Facility & Transit
Clinic/Outpatients
Home
Total
Possible
3735
2320
158
2538
5122
Probable
1329
1301
35
794
2505
Number
5064
3621
193
3332
7627
Percent
25.5%
18.3%
1.0%
16.8%
38.4%
13873
5964
19837
100.0%
Hospital
From the modifiable factors pertaining to the ward, accidents and
emergency (A&E) and transit, it can be seen that children with severe
malnutrition are not being adequately assessed, investigated, monitored
and treated (Table 4). Simple adherence to national guidelines for the
management of severe malnutrition would go a long way in improving
41
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
quality of care.4 Healthcare professionals should be made aware of the
existence of such guidelines and adherence enforced.
Table 5 shows that among children under 5 years who died with
severe malnutrition, 23.1% died within 24 hours of admission and a
further 25.9% between 24 and 72 hours after admission. This indicates
that acute care of children with severe malnutrition needs to be
improved.
Problems such as lack of nursing staff, lack of adequately trained
medical staff, lack of high care/intensive care unit (ICU) beds, or an
inadequate ambulance service will require a concerted effort from
national and provincial departments of health.
Table 4: Top 5 modifiable factors for severe malnutrition: Ward, A&E and Transit
Ward
Lack of professional nurse in
children's ward 24 hours a day
A&E
Inadequate history taken
at A&E
Inadequate monitoring of blood
glucose in ward
Inadequate investigations
(blood, x-ray, other) at
A&E
Appropriate antibiotics
not prescribed at A&E
Lack of experienced doctors
(post-community service), for
children's ward
Inadequate investigations in ward Inadequate physical
examination at A&E
Inadequate case assessment and
management at previous
admission to ward
Blood glucose not
monitored in child with
danger signs at A&E
Transit
Inadequate ambulance service
from health facility to
receiving hospital
No or delayed referral to
higher level
Inappropriate care or late
referral from private
sector/general practitioner
Severity of child’s condition
incorrectly assessed at
referring facility
Inadequate referral letter from
referring facility
Table 5: Length of stay in children under 5 dying from severe malnutrition
DOA
No.
66
%
1.1
< 24 hrs
No.
1389
%
23.6
1-3 days
No.
1521
%
25.9
4-7 days
No.
1190
%
20.2
8-14 days
No.
856
%
14.6
> 14 days
No.
855
%
14.5
Total
No.
5877
%
100
Clinic
The modifiable factors relating to clinics highlight a need for the
widespread adoption of IMCI guidelines5 (Table 6). IMCI training
5
Standard Treatment Guidelines and Essential Drugs List. Hospital Level Paediatrics. 2nd Edition. The
National Department of Health, Pretoria, South Africa. 2006. http://www.doh.gov.za/docs/index.html
(Accessed 7 March 2011)
42
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
needs to be intensified. Clinic staff who have been IMCI trained need
to ensure that IMCI is implemented at their clinics.
Home
Food security remains a major problem in the South African context.
The solution will require a multi-faceted approach. In the short term,
solutions such as educating caregivers on the planting of vegetable
gardens may go a long way in alleviating the situation. Children
qualifying for social grants should have easy access to these resources.
Long-term solutions will require overall social upliftment. Job creation
will need to be intensified.
Table 6: Top five modifiable factors for severe malnutrition: Clinic/Outpatient
Department (OPD) and Home
Clinic/OPD
Delayed referral for severe malnutrition,
weight loss, or growth faltering from
clinic/OPD
IMCI not used for patient assessment at
clinic/OPD
Child's growth problem (severe
malnutrition, not growing well)
inadequately identified or classified
IMCI not used for case management at
clinic/OPD
Did not arrive at clinic/OPD on day of
referral/did not keep appointment
Home
Caregiver delayed seeking care
Child not provided with adequate (quality and/or
quantity) food at home
Caregiver did not recognise danger signs/severity
of illness
Caregiver took child to clinic infrequently
Other caregiver modifiable factor at home/in
community
South African national guidelines
Three sets of guidelines are frequently referred to in South Africa.
Firstly, the IMCI guidelines (which have been adapted for use in South
Africa) addresses the primary care at clinic level and appropriate
referral.6 The in-hospital management of severe malnutrition is dealt
with in the Standard Treatment Guidelines and Essential Drugs List
for Hospital Level Paediatrics,5 which is based on the WHO’s
6
KwaZulu-Natal Department of Health. IMCI guidelines, adapted by South Africa from WHO and
UNICEF guidelines of 1995. KZN Department of Health; 2002.
http://www.kznhealth.gov.za/chrp/documents/Guidelines/Guidelines%20National/IMCI/IMCI%20Chart
%20Booklet.pdf (Accessed 7 March 2011)
43
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Management of Severe Malnutrition: a Manual for Physicians and
Other Senior Health Workers published in 1999.7
Challenges and research needs
While the implementation of the WHO guidelines have elsewhere
enabled decreasing case fatality rates for severe malnutrition to below
5%,8 this has not been possible in settings like South Africa where HIV
is prevalent.9 The prevalence of HIV infection in severe malnutrition
has been shown to be as high as 51%, and it can be argued that the
WHO guidelines, designed to address problems in the pre-HIV era, are
inadequate in such situations.10 Recently, several of the principles of the
“ten steps” guidelines have been questioned. As such, the composition
of F-75 and F-100, the use of intravenous fluids in severe dehydration
and hypovolaemic shock, and the lack of adequate TB and HIV
guidelines in these children needs to be examined.11 12 13 Tuberculosis is
difficult to diagnose in children, and the resultant delay in diagnosis
further complicates the timeous initiation of antiretroviral drugs in
HIV-positive children. The timing of this initiation remains
7
World Health Organisation. Management of Severe Malnutrition: A Manual for Physicians and Other
Senior Health Workers. Geneva: WHO, 1999.
http://www.who.int/entity/nutrition/publications/severemalnutrition/en/manage_severe_malnutrition_eng
.pdf). (Accessed 7 January 2010)
8
Ashworth A. Treatment of Severe Malnutrition. J Pediatr Gastroenterol Nutr 2001; 32: 516-518
9
Ferguson P, Tomkins A. HIV prevalence and mortality among children undergoing treatment for severe
acute malnutrition in sub-Saharan Africa: a systematic review and meta-analysis. Trans R Soc Trop Med Hyg
2009; 103, 541—548. doi:10.1016/j.trstmh.2008.10.029
10
De Maayer T, Saloojee H. Arch Dis Child (2011). doi:10.1136/2 of 6 adc.2010.205039
11
Brewster D. Critical appraisal of the management of severe malnutrition: 2. Dietary management. J Paediatr
Child Health 42 (2006) 575–582. doi:10.1111/j.1440-1754.2006.00932.x
12
Brewster D. Critical appraisal of the management of severe malnutrition: 3. Complications. J Paediatr Child
Health 42 (2006) 583–593. doi:10.1111/j.1440-1754.2006.00933.x
13
Akech SO, Karisa J, Nakamya P, Boga M, Maitland K. Phase II trial of isotonic fluid resuscitation in
Kenyan children with severe malnutrition and hypovolaemia. BMC Pediatr. 2010; 10: 71.
44
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
controversial and poorly researched, and it is associated with a high
mortality rate in severely malnourished children.14
In South Africa, the implementation of the guidelines has many other
challenges. Many public hospitals have no access to pre-mixed F-75 or
F-100, and the combined mineral and vitamin mix is also frequently
not accessible. Although the management of severe malnutrition in
many African countries has been made more efficient by using ready to
use therapeutic food (RUTF), most South African provinces do not
have such food supplements available.
These issues highlight the urgent need for quality, multicentre, clinical
research on the hospital management of severe malnutrition that is
associated with co-morbidities. While the above guidelines have been
based on extensive biochemical research, randomised controlled trials
addressing the safety and efficacy have been lacking. Nevertheless,
many of the WHO 10 steps are based on sound principles, and it
would be unwise to ignore the guidelines altogether.
Several studies that examine the implementation of the “10 steps” in
South Africa provide interesting insights. These studies show that
implementation is feasible in rural hospitals in South Africa, with some
local modifications.14 Case fatality rates were reduced by this
intervention, but none of the hospitals reached the 5% target.15 16 17
While it is tempting to blame inadequate care, the prevalence of HIV
and TB certainly account for some of this excess mortality, and the
target rate appears unachievable in HIV-positive children.10 The
14
Bolton-Moore C, Mubiana-Mbewe M, Cantrell RA, Chintu N, Stringer EM, Chi BH, et al. Clinical
outcomes and CD4 cell response in children receiving antiretroviral therapy at primary health care facilities in
Zambia. JAMA 2007; 298: 1888 – 99.
15
Deen JL, Funk M, Guevara VC, Saloojee H, Doe JY, Palmer A, et al. Implementation of WHO guidelines
on management of severe malnutrition in hospitals in Africa. Bull World Health Organ 2003; 81: 237–43.
16
Ashworth A, Chopra M, McCoy D, Sanders D, Jackson D, Karaolis N, et al. WHO guidelines for
management of severe malnutrition in rural South African hospitals: effect on case fatality and the influence
of operational factors. The Lancet 2004; 363: 1110–5.
17
Puoane T, Cuming K, Sanders D, Ashworth A. Why do some hospitals achieve better care of severely
malnourished children than others? Five-year follow-up of rural hospitals in Eastern Cape, South Africa.
Health Policy Plan 2008; 1-10.
45
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
sustainability of the improved case management and mortality needs to
be addressed to ensure continuity of the improved care. Puoane and
colleagues showed that self-sustaining programmes of induction, inservice training, supervision and audit are needed. Training of senior
managers will foster leadership and teamwork, and encourage
consistent and meticulous adherence to existing protocols.14
Summary and conclusion
Despite all these challenges and controversies, it is clear from the
analysis of the modifiable factors that implementation of the available
guidelines can reduce mortality in malnourished children in South
African hospitals. At clinic level, following IMCI principles will
expedite referrals for appropriate patients, and ensure correct nutrition
advice and HIV testing for others. The realisation of in-hospital
protocols will help avoid many of the potential errors identified in
Table 4. It is clear from this audit that traditional causes of death such
as uncorrected hypoglycaemia, poor correction of dehydration, lack of
adequate antibiotic coverage, and poor training of healthcare staff
continue to cause avoidable deaths.
Ultimately, the old adage “prevention is better than cure” rings true for
malnutrition. The UNICEF framework recognises basic and
underlying causes of malnutrition, with poverty playing the central
role.18 South Africa should focus on reversing the general inequity,
deprivation and poor education of the majority of the population.
Together with improved AIDS awareness and PMTCT of HIV, this
will drastically reduce the prevalence of undernutrition. Reducing
unemployment will go a long way in ensuring food security.
18
De Maayer T, Saloojee H. Arch Dis Child (2011). doi:10.1136/2 of 6 adc.2010.205039
46
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Chapter
5
Acute Respiratory Infections
Kim Harper
MBChB (UCT), DCH (SA), FCPaed (SA)
Paediatrician, East London Hospital Complex, Eastern Cape
Provincial Child PIP Coordinator for the Eastern Cape
Abstract
Acute Respiratory Infections (ARIs) are the leading cause of
childhood deaths in South Africa, as throughout the world. In the
5-year period from 2005 to 2009, 5 572 children died from ARIs
(28.9% of the 19 295 audited deaths).
Child PIP users in South African hospitals thought that 25% of
deaths due to ARIs were avoidable. Clinical personnel and
administration were together responsible for 70% of factors that
may have modified outcomes.
Progress in reducing mortality from pneumonia in children younger
than five years of age has been relatively slow. Effective vaccines
exist to reduce the disease profile from respiratory illnesses.
There is also evidence that effective and appropriate management
of clinical cases is possible at healthcare facilities and in the
community.
There is much to do to realise this potential reduction in ARI
related deaths. Efforts to control and manage pneumonia,
specifically, are needed to reduce child mortality locally and to
meet Millennium Development Goal 4, which aims to reduce
childhood deaths throughout the world by two thirds by 2015.
47
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Key Recommendations
Large scale dissemination of community Integrated Management of
Childhood Illness (IMCI) messages on vaccination, recognition of
respiratory danger signs and when to seek help.
Implementation of national policy on availability of oxygen and
respiratory support equipment such as appropriate size nasal prongs
and masks in all primary care sites and for all ambulances.
Repeated health worker training and supervision on IMCI and
recognition of respiratory danger signs.
Implementation of national policy on availability of antibiotics.
Implementation of national policy on facilities for high-care and
intensive care of children.
Introduction
Acute Respiratory Infections (ARIs) take a heavy toll on life in
children. It is estimated by the WHO that one fifth of childhood
deaths are due to pneumonia (mainly community-acquired) in
developing countries, particularly in Africa.1 Worldwide, 1.9 million
children died from pneumonia in 2004.2 In addition, ARIs also cause
significant acute morbidity and interfere with nutritional status, placing
children at risk for future health complications.
Many ARI deaths are preventable with well-established appropriate
medical interventions. Indeed, case management according to the
WHO IMCI guidelines for childhood pneumonia that have been
adopted in more than 81 countries since 2000 has been shown to
reduce childhood pneumonia mortality.3
1
Williams BG, Gouws E, Boscho-Pinto C, et al. Estimates of world-wide distribution of child deaths from
acute respiratory infections. Lancet Infect Dis 2002;2(1):25-32.
2
Chang A. Common Respiratory Symptoms and Illnesses: A graded evidence approach. Pediatric Clinics of
North America. 2009;56(1)135-156.
3
Sazawal S, Black RE. Effect of pneumonia case management on mortality in neonates, infants, and
preschool children: a meta-analysis of community-based trials. Lancet Infect Dis 2003;3(9):547-56.
48
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Acute respiratory infections may involve the upper respiratory tract,
which are usually self-limiting, (the complications being more onerous
that the initiating illness), or extend to become lower respiratory tract
infections (LRIs), which frequently produce systemic illness.
ARIs are the most common causes of both illness and mortality in
children younger than five years of age. The severity of LRIs in
children under five is worse in developing countries, resulting in a
higher case fatality rate than in developed countries. On average, each
child may experience up to six episodes of ARI annually, many of
which are viral illnesses. While effective antibiotic therapy may be
available for bacterial infections, specific antiviral therapy is not,
limiting curative treatment. Supportive treatment including oxygen,
which ought to be thought of as an essential drug, often plays as crucial
a role in managing ARIs as medication. Robust and cost effective
guidelines on prevention and treatment have been established, are
readily available and need to be widely implemented to reduce the
heavy yoke of respiratory infections.
Health profile
The following demographics were obtained from Child PIP data for
January 2005 to December 2009. A total of 19 295 child deaths were
analysed in detail. These deaths are the focus of this section.
Figure 1. ARI deaths: 2005-2009
PCP
30%
Pneumonia
70%
Of 19 295 deaths in Child PIP hospitals in South Africa, 5 572 (28.9%)
were due to ARIs. They are the leading cause of death in hospitalised
children. Of the ARI deaths, 70% were considered to be due to
pneumonia and 30% were recorded to be due to the Pneumocystis
jirovecii syndrome (PCP) (Figure 1).
49
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Age
Table 1 shows the extreme vulnerability of children younger than one
year of age. More than 78% of ARI deaths occurred in this age group.
This applies even more to PCP, of whom more than 90% were
younger than 1 year. In fact, PCP deaths are rarely seen out of this age
group.
Table 1. ARI deaths by age group
PCP
Pneumonia,
ARI
All ARI’s
0 - 28 days
28 d - 1 yr
1 - 5 yrs
5 - 13 yrs
13 - 18 yrs
Unknown
2.2%
89.3%
6.2%
1.8%
0.0%
0.6%
3.9%
68.4%
19.3%
6.9%
0.3%
1.1%
3.4%
74.8%
15.3%
15.3%
0.3%
1.0%
HIV status
Figure 2 shows the extent to which HIV influences ARI deaths. More
than a quarter of child ARI deaths (28%) were infected with HIV and
nearly one-third (31%) were exposed to HIV, indicating mothers with
HIV infection. Only 9% were known to be unaffected by HIV, while
32% had either not been tested or their status was unknown.
Pneumonia requiring admission should be a trigger for HIV testing so
it is alarming that so many children dying from pneumonia did not
have their HIV status known.
Fifty-one percent of HIV infected children dying from PCP had stage
IV disease, compared to 23% who died from pneumonia.
Figure 2. ARI deaths and HIV status: 2005-2009
Unknow n
19%
HIV Infected
28%
Not Tested
13%
HIV Negative
9%
HIV Exposed
31%
50
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Nutritional status
One third of children who had died from ARI were underweight for
age and another quarter suffered from severe malnutrition, highlighting
the contribution of malnutrition as an important co-morbid factor.
Figure 3 shows the nutritional status breakdown for deaths from ARIs.
Only 35% were considered to have a normal nutritional status.
Figure 3. ARI deaths and nutritional status: 2005-2009
2%
6%
1%
3%
OWFA
35%
Normal
UWFA
20%
Marasmus
Kw ashiorkor
M-K
Unknow n
33%
Quality of care
Child PIP users identified more than 16 113 modifiable factors in the 5
572 children who succumbed to ARIs in South African hospitals from
January 2005 to December 2009, or 2.9 per death, indicating that there
were many opportunities for improving management. Twenty-five
percent of these deaths were thought to have been avoidable.
The top five modifiable factors relating to ARIs are ranked in Table 2.
Table 2. Top five ranked ARI modifiable factors
1
2
3
4
5
Caregivers delayed seeking care
Caregiver did not recognise danger signs/severity of illness
Child not provided with adequate (quality and/or quantity) food at home
Inadequate history taken at A&E
Inadequate investigations (blood, X-Ray, other) at A&E
In the opinion of hospital-based care-workers, 1349 child deaths due to
ARIs (24.2%) were associated with “delay in seeking care by
51
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
caregivers”. Notwithstanding any possible bias, this factor is far more
common than any other reported modifiable factor associated with
ARIs. On the other hand, in only 92 out of 5572 ARI deaths of
children (1.7%) that were referred through clinics to hospitals was the
severity of their illness deemed to be inadequately assessed. Thus
severe respiratory disease is apparently adequately recognised at
primary care level, but caregivers seem only to recognise the need for
care in children when they have severe disease.
If we are to address deaths attributable to ARIs, efforts should be focused on
community awareness of respiratory danger signs and on community careseeking or access to care.
While the individual factors associated with substandard care did not
feature as the most common identified problems, there were
nevertheless numerous instances of inadequate or inappropriate health
worker responses:
In more than 200 referred children who ultimately died from
ARIs in hospital, the IMCI guidelines were not used correctly
at clinic level for assessment or management
Worryingly, on presentation to the Accident and Emergency
Department (A&E), more than 260 children (4.7%) who died
from ARIs never had their respiratory rate recorded; they
were not noted to be in respiratory distress and did not have
oxygen saturations checked.
An inadequate examination of children presenting to A&E
and failure to give antibiotics immediately in A&E occurred in
approximately 300 deaths (5.3%).
In addition, second only to limited high-care facilities as far as
ARI ward-level modifiable factors are concerned, is that
children had substandard monitoring of their respiratory
condition whilst in the ward.
Inadequate 24-hour nursing staff was deemed to be a
significant reason.
52
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Once identified as having danger signs, nearly 300 ARI child deaths
were thought to have been potentially preventable if high-care and/or
intensive care unit (ICU) had been available. This is particularly true
for children with PCP. Twenty percent of children dying from
pneumonia were re-admissions, indicating failure of prior treatment.
The theme of inadequate investigations ranks highly in the analysis of
modifiable factors associated with ARI deaths in the Child PIP
program. Even though the evidence as to whether investigations per se
such as X-rays can actually reduce mortality due to ARIs is limited, it
remains of concern. It is, however, not clear whether this factor may
indicate a lack of awareness on the part of health practitioners as to
when or which investigations to perform or an inability to investigate
due to logistical reasons. This needs greater exploration.
There were 1 696 deaths related to PCP. Doctors on the ward felt that
these children would be more appropriately managed in a high-care
facility. The second ranking modifiable factor listed for PCP deaths
was lack of professional nurses, even at ward level, to monitor the
respiratory status of these critically ill children. However, the
overarching specific thought in relation to avoiding PCP deaths at
hospital level was that “appropriate antibiotics”, presumably “high dose
co-trimoxazole”, were not prescribed early on in their admission. A
lack of experienced doctors (post-community service), specifically was
identified as being of concern to more junior doctors managing these
very ill patients.
Child PIP data shows that health sites (A&E, wards and clinics/outpatient departments) had nearly one avoidable health worker factor for
every death. There is thus an urgent need to provide health worker
supervision and training to enable implementation of national policies
and guidelines pertaining to ARIs.
South African national guidelines for ARI care
Assessing severity of pneumonia for hospital admission
The South African national guidelines set forth diagnostic criteria with
regard to the degree of severity of pneumonias in the Standard
53
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Treatment Guidelines and Essential Drugs List (STG & EDL).4
Pneumonia can be classified as very severe, severe or non-severe using
clinical signs:
1 Non-Severe Pneumonia
Cough with tachypnoea.
Tachypnoea is defined as:
Age
Respiratory rate
< 60 days
2-12 months
1-5 years
> 60/minute
> 50/minute
> 40/minute
2 Severe Pneumonia
Cough, tachypnoea and one from the list below:
- Indrawing of lower chest wall
- Presence of abnormal auscultatory findings
- Dullness to percussion
3 Very Severe Pneumonia
Severe Pneumonia plus one from the list below:
-
Central cyanosis
Inability to feed
Convulsions, decreased level of consciousness or lethargy
Grunting
Nasal flaring
Younger than 60 days (All infants <60 days with pneumonia
are automatically considered to have very severe pneumonia)
(HIV-infected children and pneumonia due to Staph. Aureus are
deemed as having very severe pneumonias.)
Severe and Very Severe Pneumonias should be admitted to hospital.
4
Standard Treatment Guidelines and Essential Drugs List Hospital Level Paediatrics 2006. The National
Department of Health, Pretoria, South Africa, 2006.
54
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
In practical terms, there are concerns over the need to provide oxygen
or intravenous antibiotics dictate admission, as well as the feasibility of
caring safely for the child at home.
The British Thoracic Society includes an oxygen saturation of <92%
and intermittent apnoea as additional reasons for admission.5
The detection of rapid breathing and chest wall indrawing are two of
the most important signs to be used for detecting pneumonias in
children. Tachypnoea is reported to detect 85% of children with
potentially fatal pneumonias. If WHO guidelines are used
“… more than 80 percent of children with potentially fatal pneumonia
are probably successfully identified and treated.”6 The use of indrawing
of the chest wall has been well studied. Conclusions are that indrawing
of the lower chest wall is a more specific sign for serious lower ARIs,
hence the specificity of the recommendation above. Using a criterion
of “any chest indrawing” may result in an overwhelming number of
children being referred to in-patient health services.
Simoes and Cherian et al 6 point out that while antibiotic treatment of
tachypnoeic children has been shown to reduce mortality, the low
specificity of the criterion means that 70% to 80% of children who
may not need antibiotics will receive them. They argue reasonably,
however, that, for primary care workers, rapid breathing is clearly the
most useful clinical sign to detect pneumonias.
Management
Antibiotic therapy according to severity:
Recommendations of antibiotic therapy are made according to severity,
thus limiting unnecessary referrals to hospital level of care.
- Non-severe pneumonia: Amoxycillin
- Severe pneumonia and infants up to 60 days: Ampicillin and
gentamicin IV
5
Pio A. Standard Case Management of pneumonia in children in developing countries: the cornerstone of
Acute Respiratory Infection Programme. Bull World Health Organ 2003;81(4):298-300.
6
Simoes EAF, Cherian T, Chow J et al. Acute Respiratory Infections Chapter in Disease Control Priorities in
Developing Countries. Editors Dean T. Jamison Oxford University Press, 2006.
55
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Amoxycillin for non-severe pneumonia, and ampicillin and gentamicin
IV for severe pneumonia including infants up to 60 days or benzyl
penicillin IV are prescribed in the STG & EDL (Hospital Level
Paediatrics).
The duration of treatment for pneumonias remains ill-defined. South
African guidelines recommend at least 2 days IV antibiotics for severe
pneumonia and 5 to 10 days for very severe pneumonia before
switching to oral antibiotics. For non-severe pneumonia, at least 3 days
of oral treatment has been confirmed to be effective. 2
Humidified oxygen therapy:
Humidified oxygen therapy by nasal prongs for severe pneumonia if
available; or if limited, to children with any of the following signs:
inability to feed and drink, cyanosis, respiratory rate greater than or
equal to 70 breaths per minute or severe chest wall retractions (WHO
1993). It should be continued until the respiratory rate is less than 60
breaths per minute and the child is improving clinically.
Adjunctive therapy:
Antipyretics, cough suppressants and mucolytics have either low or
insufficient evidence for recommendation. Chest physiotherapy is not
indicated in community acquired pneumonia, while a meta-analysis of
zinc supplementation suggests it is useful in high mortality settings
where zinc deficiency is prevalent. 5
Treatment controversies
Because antibiotic resistance is increasing, modifications to the
antibiotic policy are becoming necessary. Up to 40% of pneumococcal
isolates in developed countries have been shown to be resistant to
penicillin. 2 The use of a third generation cephalosporin or vancomycin
should be considered if clinical resistance is suspected, especially in
prolonged hospitalised and HIV infected children.
A new WHO document published in 2010 makes the following
antibiotic recommendations concerning pneumonia:
Ampicillin (or penicillin when ampicillin is not available) plus
gentamicin or ceftriaxone are recommended as a first-line
antibiotic regimen for HIV-infected and -exposed infants and
56
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
children under five years of age with severe or very severe
pneumonia.
Empirical co-trimoxazole treatment for suspected PCP is
recommended as an addition for HIV-infected and -exposed
infants from 2 months to 1 year old with severe or very severe
pneumonia.
Hospital acquired infections are predominately due to serious gram
negative organisms such as Serratia, Pseudomonas, E. Coli and
Enterobacter species, but also Staph. aureus. Children admitted to
ICUs in particular are at risk of infection with multi-drug resistant
organisms. This problem is becoming endemic and a serious challenge
to health institutions. In choosing an appropriate antibiotic, selection
must be based on surveillance of sensitivity patterns in the particular
high-care/ICU setting.
The most severe disease from pneumonia is due to the usual bacterial
organisms or due to mixed bacterial (Strep. Pneumoniae) and viral
infections. However, there is some evidence for the role of
Mycoplasma Pneumoniae and Chlamydia Pneumoniae in younger
children (< 5 years) than previously thought.7 Other than PCP, these
organisms ought to be thought of in non-responding pneumonias.
While the incidence of positive blood cultures is low in general (10% to
30%) in children with pneumonia, they are useful for ensuring that the
correct treatment is prescribed. The British Thoracic Society guidelines
recommend that a blood culture be taken in suspected bacterial
pneumonia on admission of a child to hospital. Recommendations in
South Africa are: “if facilities are available”, a blood culture should be
taken (as well as nasopharyngeal and gastric aspirates for TB or
induced sputum).
Prevention strategies
Interventions to lessen the impact of ARIs can be grouped as follows:
- Improving environmental conditions
7
Michelow IC, Olsen K, Lozano J, et al. Epidemiology and clinical characteristics of community acquired
pneumonia in hospitalized children. Pediatrics 2004;113(4)701-7.
57
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
- Nutritional interventions
- Specific immunisations
- Early detection and treatment
Nationwide improvements of socio-economic conditions (including
nutritional status) throughout developing countries lead to a significant
reduction in pneumonia in children even prior to the availability of
antibiotics. This confirms the crucial role that impoverishment plays,
though living conditions and access to healthcare, in child deaths due
to ARIs.
Vaccines
Vaccines that were previously unavailable to populations with a high
mortality rate due to ARIs are now becoming more accessible and have
been included in the national Expanded Programme on Immunisation
(EPI) schedule. A good example of the effectiveness of vaccine
prevention of ARI related disease is measles vaccine. Before the
availability of measles vaccine, respiratory disease as a consequence of
measles was the major cause of viral respiratory related deaths in
developing worlds. 6 More recently, vaccines against H. Influenzae and
S. Pneumoniae in particular, have been shown to be effective. 2
Pneumococcal conjugate vaccine reduces clinical pneumonia, severe
pneumonia, radiologically diagnosed pneumonia as well as bacteraemia
due to the serotype specific organisms. As there are more than 90
serotypes of pneumococcus, serotype-shift may reduce effectiveness
over time.2 However, newer vaccines are being introduced that
increase serotype coverage. Importantly, vaccination assists with
overcoming resistance to antimicrobials by decreasing particular
serotypes known to have a high resistance pattern.8
8
Cohen R, Levey C, de La RF, et al. Impact of pneumococcal conjugate vaccine and reduction of antibiotic
use on nasopharyngeal carriage of non-susceptible pneumococci in children with acute otitis media. Pediatr
Infect Dis J 2006;25(11):1001-7.
58
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Challenges and future research required to implement
national standards for caring for children
Challenges
Much more effort must be put into community awareness of
respiratory danger signs and the need to seek help early. A large scale
national programme is needed to inform mothers to seek help when
the baby is breathing fast, has noisy breathing, has chest wall recession,
or fails to feed, similar to the message on home oral rehydration
solution in gastroenteritis management.
While the South African vaccination programme is admirable, in order
to be effective a great number of logistical challenges, such as cold
chain management and consistent distribution of vaccine, need to be
constantly surmounted.
Once admitted, the respiratory rate and status of children with ARIs
must be monitored regularly. Children presenting to A&E need to be
more thoroughly assessed, and antibiotics and oxygen not only
prescribed but administered there and then.
The lack of adequate high-care/ICU facilities has been identified by
Child PIP users as an important in-hospital factor that must be
addressed.
Future research
In poorly resourced areas, studies are needed to define the place and
most optimal use of saturation monitors and oxygen therapy, and for
non-severe pneumonia to improve the specificity of clinical diagnostic
criteria.
Summary and conclusion
Risk factors for pneumonia include stunting and being underweight,
suboptimal breast-feeding, lack of immunisation and indoor air
pollution from household use of solid fuels.9 All these issues demand
an effective primary care approach, a well-functioning IMCI at primary
9
Niessen L, ten Hove A, Hilderink H et al. Comparative impact assessment of child pneumonia
interventions. Bull World Health Organ 2009;87:472-480.
59
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
care health facilities and trained and supportive community level health
workers, to encourage improved healthcare-seeking behaviour. Strong
breast-feeding support must urgently be provided to address the poor
nutritional status in many young children.
Simoes et al state: “The evidence clearly shows that the WHO casemanagement approach and wider use of available vaccines will reduce
ARI mortality among young children by half to two-thirds.” 6
We are fortunate in South Africa to have vaccines available in the EPI
that combat ARIs.
Our vaccination programmes must be
strengthened and effective.
To achieve the goal, we have to improve the case management of
individual cases of pneumonia. We must be able to provide good acute
care in casualty departments and adequate high care facilities with
appropriate nurse-to-patient ratios so that even the sickest of children
can be successfully treated.
60
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Chapter
6
Diarrhoeal Disease
Anthony Westwood
FCP(SA), MMed (Paed), MD, FRCPCH(UK)
Department of Health, Western Cape Province & University of Cape Town
Head of General Paediatrics, Metro West, Western Cape
Abstract
Diarrhoeal disease is the second most important direct cause of
death among children only following deaths from acute
respiratory infections as reflected in the Child PIP data from 2005
to 2009. The overwhelming majority of the 3 988 deaths due to
diarrhoea must be considered to have been unnecessary deaths,
even though the audit classified only 34.9% of deaths as
avoidable.
Almost universal protection against diarrhoea is achievable
through adequate provision and use of clean water and safe
sanitation, hand and household hygiene, breast-feeding and safe,
appropriate weaning, interruption of mother-to-child HIV
transmission, micronutrient supplementation and immunisation.
If we want to prevent deaths from diarrhoea, we have to prevent
malnutrition among children, which can be achieved through
breast-feeding, nutrient-dense weaning foods, micronutrient
supplementation, poverty reduction and relief from critical aspects
of diarrhoea prevention.
61
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
An average of 3.6 modifiable factors (MFs) were identified per
diarrhoeal death, which is significantly higher than the overall
average of 2.8 per death. Inadequate assessment of cases of
diarrhoea is the most common deficit in care reported during the
five year period.
One MF for every three deaths relates to general utilisation of
health services, suggesting a vulnerable population of children
whose caregivers cannot or will not access the full array of
prevention services. This speaks eloquently of the essential role
that community healthcare workers should take in promoting
Integrated Management of Childhood Illness (IMCI) messages, and
ensuring access to promotive and preventive care for the most
vulnerable children in South Africa.
Half of all acute diarrhoeal deaths occurred within 24 hours of
arrival at the hospital, suggesting inadequate immediate and
follow-on care in the first hours. Under-ascertainment of lifethreatening situations such as circulatory shock or dehydration in
a malnourished child can result in lethal errors in prescription of
fluids. Standardisation of regimens for fluid therapy across the
country (e.g., via the Essential Drug List (EDL)) and standardisation
of health worker training would enable a truly objective
assessment of this set of MFs, and may, in itself, reduce the
number of poorly managed dehydrated children.
62
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Key Recommendations
Ensure that household and community IMCI key household practices
reach every home through community health workers backed by
community-based health services and media messages, emphasising
oral rehydration using sugar-salt solution, and recognition of danger
signs.
Ensure that district health services help all IMCI-trained nurses at
primary health care level maintain skills in diarrhoea-related aspects
of the strategy, especially the management of children with severe
dehydration.
Expand basic and post-basic nurse training to include IMCI and
Emergency Triage Assessment and Treatment (ETAT) guidelines on
recognition and management of shock and severe dehydration.
Ensure that basic medical student training includes ETAT and IMCIbased modules on shock, dehydration and diarrhoea.
Ensure that all interns, paediatric ward staff and emergency care
health professionals receive regular training and updates on the
provisions of the EDL with respect to assessment and management of
children with acute and chronic diarrhoea.
Review the national EDLs to ensure coverage of the sequence of
diarrhoea management from primary health care through emergency
units to ward-level care for uncomplicated and complicated cases,
including those in whom diarrhoea persists.
Introduction
South Africa entered the era of the rotavirus vaccine in 2009 and
consequently childhood diarrhoeal diseases have moved from being an
inevitable incident to being preventable in most cases.1 2 Almost
universal protection against diarrhoea is achievable through adequate
provision and use of clean water and safe sanitation, hand and
1
Shabir A. Madhi, M.D., Nigel A. Cunliffe, M.B., Ch.B., Ph.D., Duncan Steele, Ph.D., Desirée Witte, M.D.,
Mari Kirsten, M.D., Cheryl Louw, M.D., et al. Effect of Human Rotavirus Vaccine on Severe Diarrhea in
African Infants. New Eng J Med 2010;362:289-298
2
Patel, Manish M.; Steele, Duncan; Gentsch, Jon R.; Wecker, John; Glass, Roger I.; Parashar, Umesh D.
Real-world Impact of Rotavirus Vaccination Pediatric Infectious Disease Journal. 30(1):S1-S5, January 2011.
63
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
household hygiene, breast-feeding with safe and appropriate weaning,
interruption of mother-to-child HIV transmission, micronutrient
supplementation and immunisation.
If disease slips through this protective net, dehydration, the main cause
of life-threatening illness from diarrhoea is to a significant extent
preventable with early home use of oral rehydration.3 If the child
becomes dehydrated, effective fluid therapy at clinic, emergency room
or hospital level prevents death and disability in almost every case.
Diarrhoeal disease is the second most important direct cause of
childhood hospital death after acute respiratory infections as reflected
in the Child PIP data from 2005 to 2009. The overwhelming majority
of the 3 988 deaths due to diarrhoeal disease must be considered to
have been unnecessary deaths, even in the absence of rotavirus vaccine
that was only introduced to South Africa in 2009. Early interventions
can interrupt a slide into life-threatening organ failure, electrolyte
disturbance or acute malnutrition, and the usual mechanisms of death.
Deaths due to chronic diarrhoea were a minority despite the prevalence
of HIV in hospital admissions; this may reflect the dominance of
district hospitals reporting to Child PIP. Many such patients may be
referred to larger hospitals for investigation and care.
Child PIP’s emphasis on knowing the nutritional and HIV status of
every child that dies is important: this information provides insight into
a child’s risk status for susceptibility to and poor outcome from acute
and chronic diarrhoea. Together with its MF approach, knowledge of
nutritional and HIV status casts a useful light on failures of primary
and secondary prevention as well as on treatment interventions. An
average of 3.6 MFs was identified per diarrhoeal death, which is higher
than the overall average of 2.8 per death. This probably reflects the
relative clarity with which MFs are identifiable with a condition as
preventable and treatable as diarrhoea. The largest proportion of MFs
was linked to pre-hospital care and emergency settings. Only a minority
of deaths (16%) were thought to be unavoidable. The remaining deaths
were either considered avoidable (34.6%) or the review meeting was
3
Cesar G. Victora, Jennifer Bryce, Olivier Fontaine, Roeland Monasch. Reducing deaths from diarrhoea
through oral rehydration therapy. Bull World Health Organ 2000;78:1246-1255
64
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
unable to form an opinion. In the under-five age group, more acute
diarrhoeal deaths were considered to have been avoidable than deaths
from all causes (35.1% of diarrhoeal deaths compared with 26.8% of all
deaths, with an odds ratio of 1.97). Even then, it is likely that a narrow
view of causation of death in children arriving at the hospital in
extremis from dehydration may have been taken when ascribing
avoidability.
This chapter analyses five years of Child PIP data to guide policy
makers (in public and health policy), programme managers and health
workers at all levels on what needs to be done to reduce gaps in the
following areas:
Primary prevention of diarrhoea in childhood.
Prevention of dehydration.
Early interventions within the health system.
Management of the sicker child with diarrhoea.
Prevention of chronic diarrhoea.
Health promotion and diarrhoeal disease prevention
The primary MFs associated with risk from diarrhoeal disease are
shown in following figures: age (figure 1), nutrition (figure 2), and
immunity (figure 3). Environment, including water supply and
sanitation method, and access to and usage of primary health care
services is depicted in the table.
Many MFs related to primary prevention can be found through death
auditing. In the case of diarrhoea, these cluster around the factors
highlighted in the following figures and table. The lack of detail that is
inevitable when the review takes place at hospital level is a limitation of
these graphic representations, and importantly, the social and domestic
circumstances of the child also influence the child’s condition. Water
and sanitation is the subject of only one MF, significantly limiting
insight into environmental factors in the causation of diarrhoeal
disease. Poverty is covered (sometimes indirectly) by more MFs and
65
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
usage of health services is covered by at least nine MFs, allowing some
insight into social factors in diarrhoeal disease deaths.
Age (Figure 1)
The overwhelming number of deaths (90%) was among children
under-five years of age, with two thirds of these occurring in infancy.
Figure 1. Age groups of children dying from diarrhoeal diseases 2005-2009 (totals
and percentages)
Nutritional status (Figure 2 & Table 1)
Inadequate nutrition appears as a common association (66.9%) among
the deaths, even when the chronic cases (who were more likely to have
malnutrition since it is a frequent consequence of ongoing diarrhoea)
are excluded (63.5%). It is no coincidence that nutrition or feeding
appeared as an MF 800 times among the 3 988 deaths (20%) in this
group. If we want to prevent deaths from diarrhoea, we have to
prevent malnutrition among children. Breast-feeding, nutrient-dense
weaning foods, micronutrient supplementation, poverty reduction and
relief are critical aspects of diarrhoea prevention (see also Chapter 4).
Some of the nutritional deficits found in this group will have been
mediated through HIV-related diseases. This is discussed in Chapter 3.
The very low number of poverty related MFs (when counted separately
from those directly related to food security) in the table is based on
only two MFs so is an insensitive marker of the true extent of income
poverty in this series of deaths.
Breast-feeding is recorded in Child PIP in categories related to HIVrisks. Unfortunately this data is missing in more than half of the
66
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
diarrhoeal deaths in this period. Of the 1 523 cases in which breastfeeding status was recorded, 1 127 (74%) had probably never received
exclusive breast-feeding.
Figure 2. Nutritional status of children dying from diarrhoeal diseases 2005-2009
(totals and percentages)
HIV (Figure 3 & Table 1)
It is not surprising that HIV was implicated in more than half the
deceased children who had an HIV test during this period.
Susceptibility to diarrhoea is greatly increased by the immunological
consequences of HIV infection and there is a significant tendency to
progress to chronic diarrhoea with its deleterious effects on nutritional
status. MFs in Child PIP relate to missed opportunities to prevent,
identify and treat HIV infection. Preventing HIV infection among
children will help to prevent serious consequences from diarrhoea
especially among babies exposed to HIV. The table shows that 117
MFs related to missed opportunities for optimal HIV assessment and
care were identified among children who died of diarrhoea.
67
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Figure 3. HIV status of children dying from diarrhoeal diseases 2005-2009 (totals
and percentages)
Social and physical environment
Child PIP is not geared to answer important questions regarding access
to water and sanitation and hygiene practices in the home in its current
format. The MFs that relate to general utilisation of health services by
the family (e.g., those relating to immunisations, Road-to-Health Card
usage) reveal one MF for every three deaths, suggesting a vulnerable
population of children whose caregivers cannot or will not access the
full array of prevention services. This speaks eloquently of the essential
place that community healthcare workers should take in promoting
access to promotive and preventive care for the most vulnerable
children in South Africa by taking the Household and Community
IMCI messages into the homes. While there is an essential role for
media, the presence within the home of a health worker is the most
powerful arbiter of change in health behaviour in the child health
arena.
Table 1. Leading primary prevention modifiable factors from the 3 988 diarrhoeal
diseases deaths: 2005-2009
Modifiable Factor Group
Poor usage of health services
Food security and nutrition
HIV-related
Poverty
Water and sanitation
68
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Prevention of dehydration
This aspect of diarrhoeal care requires a parent/caregiver who is able
to:
- Recognise the signs of diarrhoea,
- Give oral rehydration solutions safely and effectively,
- Recognise ineffective treatment by identifying danger signs,
and
- Recognise when and where to seek help.
Effective prevention of dehydration requires primary healthcare
workers who can assess, treat and counsel the parent/caregiver on
managing the child’s hydration. The IMCI patient care algorithms
should guide this process. MFs in Child PIP are constructed to explore
this set of interventions.
Delays in recognition of danger signs or care-seeking are prominent
among the recorded MFs, occurring 1 899 times or in nearly half the
deaths (47.6% of cases). In addition, in 257 cases (6.5% of deaths),
treatments other than oral rehydration (e.g., enemas, were administered
to the child at home with the risk of harm being done. The IMCI
guidelines were not used in 230 cases (5.8%). This is likely to have been
an underestimate, as some hospital staff members at larger hospitals are
ignorant of the provisions of IMCI.
An important perspective on MFs collected during the five years of
Child PIP is that a minority of the 12 418 diarrhoea-related MFs
recorded related to the home. The majority of MFs as a whole related
to activities within the health system.
Quality of care within the health system
Of the 12 418 modifiable factors that occurred in the 3 270 deaths
from acute diarrhoea, 68.7% happened within the health system and
were ascribed to clinical personnel in 58.4% of cases.
When the top 10 MFs are analysed at clinic, accident and emergency
(A&E) and ward levels and the referral and transport systems that exist
to support these unites, inadequate assessment of cases of diarrhoea is
69
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
the most common deficit in care reported during the five-year period.
The top five problems at these levels are:
-
Inadequate assessment
Inadequate monitoring
Inappropriate fluid management
Poor health records
Inadequate service provision
Inadequate assessment is a particular issue in A&E centres with three
quarters (1 492/1 952) of the instances occurring in these settings. In
addition, half of all acute diarrhoeal deaths occurred within 24 hours of
arrival at the hospital, suggesting inadequate immediate and follow-on
care in the first hours. Under-ascertainment of life-threatening
situations such as circulatory shock or dehydration in a malnourished
child can result in lethal errors in prescription of fluids. Child PIP has
one specific MF related to staff experience and training in A&E centres
(EA103 - No A&E staff trained in ETAT/BLS/APLS). This was not
used by Child PIP personnel in many cases, perhaps through lack of
knowledge of staff in the emergency units. Training and experience are
critical factors in the early recognition and resuscitation of the sickest
children with diarrhoea and dehydration. Protocols such as the EDL
exist for use in A&E settings. It is important to establish more firmly
(by audit) the relative roles of poor assessment and poor management
in these situations.
Inadequate monitoring followed inadequate assessment in frequency.
This suggests that (a) there are inadequate protocols for regular review
of dehydrated or shocked children, (b) they are not being applied, or (c)
there is insufficient staff to monitor. The frequency of death within 24
hours also speaks to deficits in this aspect of care. MFs related to staff
numbers did not appear prominently, suggesting that protocols are the
main missing factor in this area.
It is perhaps surprising that fluid management of dehydrated children
appears only third on the list of deficiencies in care. However if
assessment is wrong, fluid prescription cannot be right. Problems with
fluid therapy occurred at all levels. Child PIP does not provide a gold
standard for fluid therapy, leaving units to determine whether therapy
70
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
was adequate using standard protocols. Standardisation of regimens for
fluid therapy across the country (e.g., via the EDL) would enable a truly
objective assessment of this set of MFs to be made, and may, in itself,
reduce the number of poorly managed dehydrated children.
Prevention of chronic diarrhoea
Chronic diarrhoea most commonly follows an episode of acute
diarrhoea. Young and immune-compromised children (e.g., HIVinfected and/or malnourished children) and those who have a severe
bout of acute diarrhoea (especially with bloody diarrhoea) are most
susceptible to the intestinal damage that leads to a vicious cycle of
diarrhoea and nutritional deficit. Breast-feeding young infants,
prevention of malnutrition and HIV infection are the means to
reducing deaths from chronic diarrhoea. Recognition of the child who
has ongoing diarrhoea as being in a special risk category is important.
Early intervention such as alteration of feeds and vitamin
supplementation may prevent the severe nutritional consequences of
this condition that make it so dangerous.
A wide array of MFs make up the 2 043 factors recorded in this group.
The only MFs recorded more than 100 times related to delayed care
either by the parent or by health workers was not recognising
nutritional vulnerability secondary to the ongoing diarrhoea. Wardbased MFs were conspicuous by their absence. This probably reflects
the difficulty in reversing the damaging effects of chronic diarrhoea
once admission has become necessary. Management of these children
is complex and challenging; Mortality Review meetings are likely to
have been forgiving when analysing ward management after a child
died of chronic diarrhoea.
National guidelines for diarrhoea management
The two national guidelines that have been developed for the
management of acute gastroenteritis are the IMCI Case Management
guidelines primarily for nurses in the primary health care system, and
the guideline in the National Hospital Level EDL for children,
published by the Department of Health in 2006.
71
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Integrated Management of Child Illnesses (IMCI)
The IMCI guideline starts with an assessment of general danger signs
(i.e., child is vomiting everything, lethargic, convulsing), which should
identify the sickest children with gastroenteritis. . Treatment involves
rapid assessment of blood glucose level, treating a low reading, and
rapid referral to a senior health worker with more experience and skills.
Fluid therapy should be initiated if there is a history of diarrhoea.
An assessment of hydration status is made for the child with no danger
signs but a history of loose stools. Classification in one of three
categories is required: ‘no visible’, ‘some’ and ‘severe’ dehydration.
Each category has a set of therapies that should be initiated, involving
fluid therapy, zinc supplementation, advice to the caregiver, and
referral or follow up plans.
Oral rehydration therapy (ORT) corners are mandated at all
primary healthcare sites for rapid and controlled use of oral rehydration
fluid for children with diarrhoea, and for education of caregivers of all
children on what to do if a child develops diarrhoea.
The fluid therapy regimens for each category of dehydration are set out
in detail:
No visible dehydration: A table of volumes per hour to be
given orally.
Some dehydration: A table of volumes to be given orally
based on weight or age.
Severe dehydration: Intravenous fluids given rapidly according
to weight or age.
Regular re-assessment is required by IMCI in the four hours following
initiation of therapy. The children with severe dehydration should be
referred to the next level of care rapidly.
Essential Drug List (EDL)
This covers acute diarrhoea and chronic or persistent diarrhoea.
72
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Acute diarrhoea
Importantly, the EDL guideline is largely compatible with the IMCI
guidelines although it uses different terminology.
Detailed diagnostic criteria are given for identification of shock and
degrees of dehydration: severe dehydration, dehydration, no visible
dehydration. In addition, signs of co-morbidity and possible
complications of gastroenteritis and dehydration are given.
A set of simple investigations is required for children with more severe
illness or co-morbidity.
Non-drug therapies are set out covering re-assessment, monitoring and
feeding. The importance of continuing to feed the child during the
episode is emphasised.
Drug therapies play a very limited role in the management of
gastroenteritis. Antibiotics are advised only for dysentery; zinc and
potassium chloride are advised for children with deficiencies in these
micronutrients; and Vitamin A for those with recurrent diarrhoea.
(Note: zinc and vitamin A recommendations differ from the IMCI and
Vitamin A supplementation programmes which recommend routine
use in all cases [zinc] and those with severe diarrhoea [vitamin A]). All
other drugs should be avoided.
Fluid therapy for the first few hours for the various categories of
dehydration are set out in tabular and bullet-point form. Details are
given for volumes, fluids and methods of rehydration and maintenance
of hydration. The protocols emphasise the need for continual reassessment.
Specific circumstances of the malnourished child are highlighted: use
fluid less rapidly, and avoid the intravenous route as required by the
WHO 10 Steps protocol for severe malnutrition.
Likewise, detailed advice with dosages for children with serum
electrolyte derangements is available in the EDL.
The role of home-based sugar salt solution (i.e., 1 litre of boiled then
cooled water, half a level teaspoon of salt, 8 level teaspoons of sugar) is
highlighted for use with the formula.
73
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Referral criteria are given for children with continuing severe fluid
deficits.
Persistent or chronic diarrhoea
Persistent diarrhoea is defined as diarrhoea that continues beyond 7
days; according to the EDL, chronic diarrhoea begins at 14 days. The
EDL notes the potential dangerous nutritional consequences of
persistent or chronic diarrhoea.
The main causes of persistent/chronic diarrhoea are set out with a
simple set of criteria that suggest whether this is toddler’s diarrhoea, a
complication of acute diarrhoea, or a malabsorption syndrome. Tests
are suggested for each of these situations. A basic therapeutic approach
to this complex situation is set out, including dietary modification
(lactose-free feeds) and the ‘bowel cocktail’: non-absorbable antibiotic,
cholestyramine, and metronidazole if giardiasis is suspected.
Summary and conclusion
This five-year review of Child PIP data has provided insights into the
malign influences of poverty, nutritional deficits and HIV in the
causation of diarrhoea in young children in South Africa. Diarrhoea
remains a major preventable cause of morbidity and mortality
throughout the country. It is not possible to estimate trends in
diarrhoea mortality owing to the increasing number of sites reporting
to this database in the period under review. Nonetheless it is clear from
national data and Statistics of South Africa sources4 that considerably
more effort is required to combat these scourges. The Child PIP data
and international experience5 suggest how this may be tackled.
Preventive and promotive interventions for diarrhoea must concentrate
on infants. Breast-feeding must be ceaselessly and vigorously promoted
and supported.
4
StatsSA (Statistics South Africa). 2005, 2006, 2007, 2008. Mortality and Causes of Death in South Africa.
Findings from Death Notification. Pretoria: Statistics South Africa.
5
King CK, Glass R, Bresee JS, Duggan C. Managing acute gastroenteritis among children. MMWR
2003;52(RR16):1-1
74
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Integrated programmes are required to tackle the prevention of
malnutrition and diarrhoea, which include:
Adequate provision of clean water and sanitation services by
municipalities.
Community empowerment to enable community members to
interact with local service providers, thus ensuring that the
environment safe for children.
Community health worker programmes through which the
IMCI Household and Community Component Key Family
Practices are delivered to homes in all districts in South Africa,
particularly in vulnerable communities and households.
Community nutritional support programmes in all districts.
Primary healthcare services that reinforce the preventive
messages, recognise vulnerable children and channel them
into appropriate programmes such as the EPI, nutritional
support and social welfare interventions, and provide effective
management of sick children.
PMTCT implementation must be strengthened through integration
with basic antenatal care, safe delivery of the baby, safe infant feeding,
and post-natal care and immunisation programmes, with clear referral
pathways to ART children found to be HIV-infected.
Once the child becomes dehydrated, the safe and effective
management of the child’s fluid requirements remains a challenge at all
levels in the health system. Protocols, training and regular audits are
required to standardise care across the country and the health system.
IMCI remains a cornerstone of this effort, but improving emergency
care and compliance with national guidelines for safe and effective
diarrhoeal treatment is an essential task for the health policy-makers,
planners, managers and clinical staff throughout the health system.
A national definition of two weeks of diarrhoea being the signal for
specific interventions to prevent the consequences of chronic diarrhoea
should be introduced together with protocols covering this situation in
IMCI (i.e. referral) and the EDL.
75
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Chapter
7
Sepsis
Mark Patrick
MBChB, DCH, FCPaed
Paediatrician, Grey’s Hospital and Pietermaritzburg Metropolitan Hospitals Complex
Chairperson of Child PIP
Abstract
In high and low income countries, sepsis has long been recognised
as an important cause of morbidity and mortality. In South African
hospitals many children die of sepsis.
Child PIP data can describe the health profile of and quality of care
received by children dying of sepsis.
In the 5-year period from 2005 to 2009, 3 134 children died of
sepsis, making it the third most important cause (16%) of the
19 295 audited deaths. Forty-nine percent of these children also
had severe malnutrition, and 55 % were either HIV-exposed or infected. Very few (13%) died within a high or intensive care unit
(ICU).
Modifiable factors occurred at a rate of 2.6 per death, of which
62% occurred within the health system.
Twenty-seven percent of the deaths were regarded as NOT
avoidable, meaning that of the 3 134 children who died of sepsis,
approximately 2 200 may have survived had the process of caring
for them been different.
There is an urgent need for a countrywide initiative to improve the
clinical care of seriously sick children by applying the standards of
care, and by updating and applying guidelines for the diagnosis
and management of children developing sepsis in South Africa.
76
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Key Recommendations
Policy
A national action plan for improving the quality of care for children with
sepsis in South Africa needs to be implemented at all levels of care within
the district health system and should include:
Developing a uniform definition of sepsis and its treatment guideline
that should be disseminated to all levels of care;
Updating and communicating the Integrated Management of
Childhood Illness (IMCI)
guidelines for febrile illness to all
practitioners; and
Developing clear referral guidelines for sepsis.
Administration
Clinical managers must be held responsible for providing:
Appropriate infrastructure (e.g., adequate care facilities for children,
communication, transport);
Sufficient staffing with adequate training and skill; and
Appropriate equipment and consumables, in particular all drugs and
antibiotics according to the Essential Drugs List (EDL).
Clinical Practice
Experts in the provision of emergency and intensive care for children,
together with regional hospital paediatricians and district hospital family
medicine practitioners should:
Update the IMCI guidelines for recognition of sepsis in primary care;
Update and define a clinical care guideline for sepsis for the South
African context in the light of the major co-morbidities of
malnutrition, HIV infection and tuberculosis;
Propose such a guideline as a separate chapter in the EDL; and
Ensure countrywide dissemination of the guidelines.
Trainers and supervisors of junior doctors working in frontline clinical care
situations must ensure:
Competence in recognising the likelihood of sepsis;
77
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Compliance with the standard of care in initial clinical management of
sick children; and
Familiarity with and rigorous application of the guidelines for sepsis.
Education
Medical schools, nursing colleges and all hospitals training interns should:
Emphasise the importance of sepsis as a major cause of childhood
mortality; and
Ensure that students and interns are competent in all aspects of
sepsis management, including diagnosis, treatment, interventions,
and standards of care.
Introduction
For every hour’s delay in the administration of antibiotics to a child
with sepsis, survival decreases by 12%.1
Internationally, sepsis contributes substantially to disease burden and
mortality. In the USA, a high income country, 750 000 new cases of
sepsis occur annually. In Germany, 60 000 deaths occur from sepsis
per year.2 In high income countries, severe sepsis is more likely to occur
in patients already hospitalised, who acquire infection and develop
severe sepsis in hospital. In low income countries infections are more
likely to be community acquired, and their impact is made more severe
due to the high prevalence of co-morbidities such as malnutrition, HIV
infection and tuberculosis. The majority of South African children fall
into this category.
1
Kumar A et al. CCM 2006; 34: 1589-96: Delay in administration of effective antimicrobials from first
documented hypotension increases mortality
2
Bulletin of the World Health Organization 2010;88:839-846. doi: 10.2471/BLT.10.077073
78
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
In this survey, 16% of children who died had sepsis3 assigned as the
main cause of death. This chapter describes the health profile of and
quality of care received by children dying of sepsis.
Health profile of the children who died of sepsis
During the 5-year survey period, for the 343 408 admissions there were
19 295 audited deaths, to which 53 326 modifiable factors were
assigned, giving a rate of 2.8 modifiable factors per death. After acute
respiratory infections and diarrhoeal disease, sepsis was the third most
important cause of death, being assigned to 3 134 (16%) of the
children who died.
Age
For the 1 065 neonates dying in children’s wards, the leading cause of
death was sepsis (28% of all deaths). In infants it was the third most
important cause of death (15%), following respiratory infections and
diarrhoea and in the 1 to 5 year group, sepsis was the second most
important cause (19%).
Nutritional status
Almost half (49%) of the children dying of sepsis had severe
malnutrition (28% marasmus, 12% kwashiorkor, 9% marasmickwashiorkor), and 26% were underweight for age. This demonstrates
the substantial contribution malnutrition makes as a co-morbidity, and
quantifies the complex clinical challenge South African health workers
must confront when dealing with children with sepsis.
HIV infection
Fifty-five percent of children dying of sepsis were either HIV-exposed
(24%) or HIV-infected (31%). Thirteen percent were regarded as being
HIV clinical stage 3 and 33% were stage 4. HIV is quantified in this
data as a second serious co-morbidity for children dying of sepsis in
South Africa.
In 5% an HIV test was regarded as ‘not indicated’. In the country at
the epicentre of the global HIV pandemic, it is hard to imagine how an
3
In Child PIP, the cause of death category is ‘Septicaemia, possible serious bacterial infection’. To have died
from this condition, the child is likely to have died from ‘severe sepsis’, or ‘septic shock’, the categories used
by the international Surviving Sepsis Campaign. In this chapter, ‘sepsis’ is used for brevity, and encompasses
these three meanings.
79
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
HIV test can be ‘not indicated’ in a child dying of sepsis. In addition,
for 24% of children in this group (1 in 4), their HIV status was
unascertained.
Length of stay
Fifty-five percent of children dying of sepsis died within the first four
days of admission, with more than half of these children (29% of the
total) dying in the first 24 hours. The early demise of children dying
with sepsis points toward an urgent need for improving the care of
children whose main clinical problem at presentation is sepsis.
Place of death
Only 13% of children dying of sepsis died in intensive care (6%) or
high care (7%) units. Thirty-seven percent died in mixed medical and
surgical wards (corresponding to district hospitals) and 42% in medical
wards (regional, provincial and central hospitals).
Quality of care received by the children who died of
sepsis
For the 3 134 children dying of sepsis, 8 165 modifiable factors were
recorded, giving a rate of 2.6 modifiable factors per death. Forty- four
percent of the modifiable factors occurred within the hospital itself,
two thirds of which occurred in the ward and one-third in
casualty/out-patient department (OPD) and accident and emergencies
(A&E). Thirty-eight percent occurred at home (Figure 1). Fifty percent
of the modifiable factors were attributable to clinical personnel, 14% to
administrators and 37% to caregivers (Figure 2).
With more than one modifiable factor per death being attributed to
clinical personnel, and many of these being related to not following the
most basic principles of medical care, serious deficits in the quality of
hospital-based care are evident.
80
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
4
Figure 1: Modifiable factors (%): Where? Figure 2: Modifiable factors (%): Who?
60%
60%
50%
40%
30%
20%
10%
0%
50%
40%
30%
20%
10%
Clinical
Personnel
Caregiver
Administrator
0%
Clinical
Personnel
Caregiver
Administrator
Table 1 lists the leading modifiable factors at each point in the health
system. At hospital level the data highlight the lack of high and
intensive care facilities for children with sepsis, and alarmingly, the
failure to prescribe antibiotics.
In clinics IMCI, which aims to identify, treat and refer children with
possible serious bacterial infection (severe pneumonia or very severe
disease) was frequently not used. At home, delay in seeking care and
not recognising danger signs recur as serious problems.
Table 1. Modifiable factors 2005-2009, for children dying of sepsis (n=8 165)
Care in the Ward
Lack of high care and/or ICU facilities for children in own and higher level facility
Inadequate monitoring of blood glucose in ward
Lack of professional nurse in children's ward 24 hours a day
Lack of experienced doctors (post-community service), for children's ward
Inadequate antibiotics prescribed in ward
Care in Admission and Emergency
Appropriate antibiotics not prescribed at A&E
Inadequate investigations (blood, x-ray, other) at A&E
Inadequate history taken at A&E
Inadequate physical examination at A&E
Blood glucose not monitored in child with danger signs at A&E
Transit Care
No or delayed referral to higher level
Inappropriate care or late referral from private sector/GP
Inadequate ambulance service from health facility to receiving hospital
Severity of child’s condition incorrectly assessed at referring facility
4
Child PIP user feedback indicated a need for categorising and recording modifiable factors occurring in the
referring facility and during transport. This was done and incorporated into the programme in 2009.
Quantification of transit-related factors only began in 2009.
81
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Care in Clinics
Delayed referral for severe malnutrition, weight loss, or growth faltering from clinic/OPD
Child's growth problem (severe malnutrition, not growing well) inadequately identified or
classified
IMCI not used for patient assessment at clinic/OPD
IMCI not used for case management at clinic/OPD
Inadequate notes on clinical care (assess, classify, treat) at clinic
Care at Home
Caregiver delayed seeking care
Child not provided with adequate (quality and/or quantity) food at home
Caregiver did not recognise danger signs/severity of illness
Inappropriate treatment given at home with negative effect on the child (e.g., enema)
Caregiver took child to clinic infrequently
Was the death avoidable?
After reviewing each death in the Child PIP process, an assessment is
made taking all factors into consideration, of whether or not the death
was avoidable. For children dying of sepsis, only 27% were regarded as
NOT avoidable. Twenty-four percent were regarded as avoidable and
in 35% there was uncertainty (Figure 3).
It is possible that of the 3 134 children who died of sepsis,
approximately 2 200 may have survived had the process of caring for
them been different.
Figure 3. Was the death avoidable?
14%
27%
Yes
Not sure
24%
No
35%
82
Unknown
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
South African national standards for caring for children
who died of sepsis
The international Surviving Sepsis Campaign provides three severity
grades for sepsis: uncomplicated sepsis, severe sepsis and septic shock.
Children who died with their main cause of death assignment being
‘Septicaemia, possible serious bacterial infection’ are likely to have
fallen into the latter two grades. Progression through the three grades
of severity begins at the onset of illness, and rapidity of progression
may be difficult to predict. Proper recognition and management of the
child with sepsis is therefore required at all levels within the health
system, with survival being dependent on recognition of severe sepsis
and septic shock, and on an urgent and appropriate response.
At clinic level, IMCI is the national standard for identifying, classifying
and treating children with severe sepsis. In the emergency care setting,
ad hoc training of health workers in life support using Basic Life
Support (BLS), Advanced Paediatric Life Support (APLS), Paediatric
Advanced Life Support (PALS) and Emergency Triage, Assessment
and Treatment (ETAT), and others (see also Chapter 11) occurs. All of
these systems advocate early recognition and treatment of the child
with sepsis.
At ward level, the South African Standard Treatment Guidelines and
Essential Drugs List, Hospital Level, Paediatrics,5 the country standard
for paediatric care, management of sepsis is covered in Chapter 1,
under the heading “Shock’’ and in Chapter 8 under “Sepsis (outside the
neonatal period)”. Both sections emphasise the need for early and
‘aggressive’ management.
International guidelines for surviving sepsis are available through the
Surviving Sepsis Campaign, the most recent revision being in 2008.6
While debate and controversy necessarily continue in the finer detail of
the guidelines, there is clear, evidence-based international consensus on
5
6
Standard Treatment Guidelines and Essential Drugs List for South Africa: Hospital Level Paediatrics,
National Department of Health, South Africa, 2006
Surviving Sepsis http://www.survivingsepsis.org
83
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
the most fundamental aspects of responding to the needs of the child
with severe sepsis and septic shock:
Early recognition.
Early initiation of antibiotics.
Early establishment of an effective circulating blood volume.
Challenges
A national plan for improving the quality of care for children with
severe sepsis in South Africa should be devised. The international
Surviving Sepsis Campaign suggests a six-point action plan for
countries wishing to improve survival from sepsis, which entails:
1. Awareness: Increase awareness of healthcare professionals,
governments, health and funding agencies, and the public of
the high frequency of and mortality associated with sepsis.
2. Diagnosis: Improve the early and accurate diagnosis of
sepsis by developing a clear and clinically relevant definition
of sepsis and disseminating it to all health workers working
with children.
3. Treatment: Increase the use of appropriate treatments and
interventions by disseminating the range of care options and
urging their timely use.
4. Education: Encourage the education of all healthcare
professionals who manage sepsis patients by providing
leadership, support and information to them about all
aspects of sepsis management, including diagnosis,
treatments and interventions, and standards of care.
5. Counselling: Provide a framework for improving and
accelerating access to post-ICU care and counselling for
sepsis patients.
84
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
6. Referral: Recognise the need for clear referral guidelines that
are accepted and adopted at a local level in all countries by
initiating the development of global guidelines.7
Local (South African) expertise and guidelines for children with sepsis
exist in abundance, but perhaps the call now should be for a more
unified, standardised and national response, similar to that of the
Surviving Sepsis Campaign.
Summary and conclusion
Sepsis is the third most important cause of death for children dying in
hospitals conducting the Child PIP mortality audit.
In children dying of sepsis, malnutrition and HIV infection are
frequent co-morbidities. Children with sepsis die soon after admission
to hospital and have limited access to high and intensive care facilities.
They present late, without the severity of their condition being
recognised in the home setting, and problems in the process of caring
for them in the health system have to do with basic standards of care
not being followed.
The problem of sepsis for South African children is of such magnitude
that responding to it must become a priority in particular at the clinical
coalface, but also for all politicians, administrators, educators, and
community leaders responsible for delivering health services to
children.
The data provided by Child PIP demonstrate a countrywide need for
an action plan. Experts in the provision of emergency and intensive
care for children based in medical schools throughout the country need
to gather, with paediatricians working in regional hospitals, and family
medicine practitioners in district hospitals, to work out a country
strategy for improving the quality of care for children with severe
sepsis and septic shock in South Africa.
.
7
Surviving Sepsis http://www.survivingsepsis.org
85
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Chapter
8
Tuberculosis
Barnesh L Dhada
MBChB (Natal), DA (SA), FC Paeds (SA)
Head of Department of Paediatrics, Grey’s Hospital
Pietermaritzburg Metropolitan Hospitals Complex
Abstract
Childhood tuberculosis (CH TB) is among the leading causes of
death (7.1 % of all audited deaths) among children in South African
hospitals according to Child PIP. More children with HIV-TB coinfection and malnutrition died than those with TB without these
conditions. Re-admissions (29.4%) and modifiable factors (MFs)
among clinical personnel (accounting for 51.6% of MF among TB
deaths) reflect chances for interventions that were missed. The
quality of TB care can be improved if healthcare workers follow
best practice guidelines and eliminate “missed opportunities”.
This includes proper HIV, TB and malnutrition prevention and
treatment. We must ensure proper implementation of what we
know works, until we can find a better way to do it.
86
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Key Recommendations
Policy level: Ensure that either a single, up-to-date clinical care
guideline is available, or if there are many guidelines, that all are
compatible.
Administrator level: Ensure that funding and resources are adequate
and appropriate for the standard of TB care for children (and adults)
expected to meet set targets.
Clinical Practice level: Ensure that all clinical personnel are aware of
and are assisted with implementing the standard of care case
detection and management guidelines as best as possible in their
daily practice.
Clinical Practice level: Ensure that all clinical personnel take
cognisance of the three inextricably linked conditions of TB, HIV and
severe malnutrition in children under five years by ensuring that
management plans are integrated.
Education level: Ensure that quality continuing medical and nursing
education on childhood TB at undergraduate, post-graduate and
vocational level is adequately delivered.
Introduction
Childhood tuberculosis (CH TB) continues to act as a sentinel of
performance reflecting inadequate TB control.1 South Africa’s
performance in the areas of prevention of spread from adults to
children through intensified case finding and treatment, using the 1994
World Health Organisation (WHO) DOTS strategy, and in contact
tracing of highly vulnerable and susceptible groups (impoverished,
close contacts, children under five, HIV-infected and malnourished),
has been below the targets in the STOP TB strategy and the Global
Plan to Stop TB 2006-2015. These were launched in 2006 and
“…explicitly aim to redress the chronic neglect of childhood TB.”2
1
van Rie A, Beyers N, Gie R, Kunneke M, Zietsman L, Donald PR. Childhood tuberculosis in an urban
population in South Africa: burden and risk factor. Archives of diseases in childhood 1990; 80: 433-437.
2
World Health Organization. Guidance for national tuberculosis programmes on the management of
tuberculosis in children. WHO/HTM/TB/2006.371. Geneva: WHO.
87
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
These targets are linked to the Millennium Development Goal 6, target
8 and include:
By 2005: to detect at least 70% of new sputum smear-positive
TB cases and cure at least 85% of these cases;
By 2015: to reduce prevalence of and deaths due to TB by
50% relative to 1990; and
By 2050: eliminate TB as a public health problem (< 1 case
per million population).
The performance against these targets is reflected in the WHO Global
TB Report 20103 indicating that the trend in South Africa (and SubSaharan Africa) shows a deteriorating picture despite an international
trend that is showing promising signs of improving TB incidence,
prevalence and mortality rates. South Africa remains among the top
five (India, China, SA, Nigeria, Indonesia) of the 22 high burden
countries that account for more than 80% of the world’s TB burden.
CH TB in all its forms is one of the “Big 5” leading causes of inhospital deaths of children in South Africa as recorded in the Child PIP
audit process. TB accounts for 7.1% (1 368 / 19 295) of all deaths
audited between 2005 and 2009, using the main cause of death
diagnosis recorded on the Child PIP death data capture sheet
completed for all deaths. There is no discernible downward trend in
TB mortality over the five years of Child PIP data, using either main or
all (main and other)* cause of death diagnoses as reflected in Table 1.
Table 1 further reflects data for year on year trends including the main
cause of death, and all (main and other) diagnoses recorded for
individual patients who died.
Table 1. TB deaths (%) using cause of death diagnosis in Child PIP 2005-2009
2005
2006
2007
2008
2009 Total
Main cause of death diagnosis
All diagnoses (main and other)*
6.9
8.2
7.7
9.6
7.7
8.7
6.1
7.7
7.4
8.3
7.1
8.4
Note that TB mortality occurs against a background of severe
malnutrition in one third of children who died (6 623 / 19 295) and of
3
World Health Organisation. WHO Report 2010 Global Tuberculosis Control. Geneva: WHO.
88
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
association with HIV exposure and infection in half (9 847 / 19 295) of
the children who died. These three conditions seem to be inextricably
linked but will be interrogated separately in this report.
Epidemiologic*ally, HIV infection has been driving the increasing TB
incidence with parallel trends in the last decade.
Tuberculosis in South Africa
South Africa is classified by the WHO as a country with a high burden
of HIV, high TB and high multi-drug resistant TB (MDR-TB). Table
2 compares the main indicators assessing performance of the global
(2008) and South African (2009) TB Control Programme indicating
that our national performance is poor on all indicators and against all
targets. The case detection rate has just climbed over the target and
may herald the beginning of the change in the tide against TB.
Table 2. From WHO TB Report 2010 & Country Profile (Annexure 1)
TB Indicator
Global (2008)
RSA (2009)
TB Incidence*
TB Prevalence*
TB Mortality*
Case Detection Rate** (Target 70%)
Treatment success rate** (Target 85%)
Treatment success rate**
139
164
20
61%
87%
971
808
52
74% (all forms)
76% (Smear +ve)
68%(***Smear-ve and
extrapulmonary TB)
64% (Retreatment)
Treatment success rate**
*
per 100 000 population
**
reported for the preceding year (2008)
***
more likely to indicate CH TB Treatment success rate as most cases in children
would fall in this category
According to the WHO Country Profile: Tuberculosis – South Africa
(See Annexure 1), there were 49 825 new notifications of TB in
children under 15 years of age in 2009, compared with a total of 340
066 in all ages (14.7%). This proportion is similar to our previous
estimate in Saving Children 2005-20074 indicating that children account
for 15% to 20% of the TB burden, when estimating from smear
positive cases.
4
Stephen CR, Mulaudzi MC, Kauchali S, Patrick ME eds. Saving Children 2005-2007: A fourth survey of
child healthcare in South Africa. Pretoria: University of Pretoria, MRC,CDC; 2009.
89
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
The WHO TB indicators are not reported by age and thus do not
provide a picture of HIV co-infection or MDR-TB burden in children.
The following data from the WHO Report illustrate the enormity of
the challenge. In 2008, 1.8% of new TB cases, and 6.7% of retreatment
cases were infected with MDR-TB. Further, 58% of TB patients tested
for HIV were found to be HIV-infected, but only 49% of TB patients
had a known HIV status, and only 42% of HIV-positive TB patients
had been started on antiretroviral therapy in 2009. For South Africa,
drug resistance and HIV co-infection remain major obstacles to
successfully overcoming these epidemics.
The gaps of knowledge concerning childhood TB will hopefully
change with data recording and reporting now being requested by
WHO in two age bands for children with TB: viz. under 5 years and 5
to 15 years.
Health profile of children who died of TB
The data presented in the tables and the conclusions made do not
differ substantially from the review of three years of data in the Saving
Children 2005-2007 report. 4 However, an increased number of audited
cases have made the data more representative and this provides a
clearer picture of the quality of care received by children in hospitals in
South Africa.
Children under five accounted for 70% of all audited childhood TB
deaths. TB patients had a HIV co-infection rate of 54.7% and a 4 times
higher number of deaths with co-infection in comparison to TB deaths
in HIV uninfected children. It is unacceptable that in 241 of 1368 TB
deaths (17.6%) the HIV status of children who had been diagnosed
with TB were unknown. The staging data suggest that some patients
were staged without confirmation of their HIV status; however the
majority displayed features of advanced HIV disease needing highly
active antiretroviral therapy (HAART) when co-infected with TB (stage
3 or 4 defining diagnosis).
Only 14.7% of children who died of TB had a normal weight, while
37.3% were marasmic and 29.9% were underweight for age (Figure 1).
Not growing well or failure to thrive is therefore an important clinical
90
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
feature during screening that may indicate TB, both as a risk factor in
the development of TB disease and as a consequence of tuberculosis.
Figure 1. Weight categories for deaths from TB (all) reflected as a percentage: Child
PIP 2005-2009 (n=1368)
Quality of care received by children who died of TB
TB is a curable infectious disease, yet only 21.9% of cases were thought
to be avoidable by Child PIP users (Table 3). This is somewhat less
than the overall figure, possibly indicating how severely ill TB-infected
patients were, but this needs further evaluation to interpret what it
means for the quality of TB care offered to children. Are we doing well
and therefore are these deaths really unavoidable?
Table 3. Deaths from TB: Audit opinion on “Was the death avoidable?”
Yes
No.
%
TB (all)
299
21.9
Grand total 5 013 26.0
Not sure
No.
%
502
36.7
6 708 34.8
No
No.
%
394
28.8
4 702 24.4
Unknown
No.
%
173
12.6
2 872 14.9
Total
No.
%
1 368 100
19 295 100
TB deaths and those due to chronic diarrhoea tended to occur after a
longer hospital stay than acute diarrhoea, pneumonia and septicaemia,
which were more likely to cause death within the first three days after
admission (Table 4). Later deaths suggest a slower, less acute final
pathway to death than septic shock in septicaemia or hypovolaemic
shock in gastroenteritis. It is not surprising that the attending clinical
personnel (ward level) considered such deaths less likely to be
preventable and had fewer MFs.
91
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Table 4. Length of stay: Comparison of TB deaths with the overall picture
%
0.8
< 24
hrs
%
15.5
1-3
days
%
23.2
4-7
days
%
20.7
8-14
days
%
17.0
> 14
days
%
22.7
2.6
31.5
25.7
16.7
11.3
12.0
DOA
TB (all)
Grand
total
Unkn
Total
0.1
No.
1 368
%
100
0.2
19 295
100
Table 5 shows that more patients with HIV infection, chronic
diarrhoea and TB disease had been re-admitted before they died, in
comparison with acute diarrhoea or the overall picture. This may
reflect “missed opportunities” in their care on previous admissions that
could have prevented the re-admission or the death.
Table 5. Readmission status for deaths from TB in comparison with all deaths from
leading causes
Not readmitted
No.
%
TB (all forms)
Diarrhoea - chronic
HIV-infected
Diarrhoea - acute
Total (“Big Five”)
771
363
2 863
2 248
9 668
56.4
50.6
52.7
68.7
64.9
Unknown
No.
%
Readmitted
No.
%
195
141
696
472
2 250
402
214
1 873
550
2 983
14.3
19.6
12.8
14.4
15.1
29.4
29.8
34.5
16.8
20.0
Total
No.
%
1 368
718
5 432
3 270
14 901
100
100
100
100
100
Table 6 shows that slightly more than half (51.6%) of the MFs
recorded among TB deaths are attributable to clinical personnel, with
TB deaths showing fewer MFs than the acute diseases, pneumonia or
diarrhoea. This may reflect the audit team’s opinion that with the
longer hospital stay there was more chance of avoiding MFs.
Nevertheless, clinical care can be significantly improved if these factors
are addressed. A closer look at the 3 416 recorded TB related MFs
shows that just over 68% of MFs were found within the health system
(administrators and clinical personnel). This proportion is similar for all
causes of death, and higher than for HIV-infected or malnourished
deaths.
Table 6. Proportion of modifiable factors ascribed to responsible persons
Administrator
TB (all forms)
HIV-infected
Severe malnutrition
Overall
16.5%
13.5%
11.9%
13.6%
Clinical
Personnel
51.6%
50.1%
50.2%
54.4%
92
Caregiver
31.9%
36.4%
37.8%
32.0%
Total
3 416
12 869
19 800
53 328
100
100
100
100
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
If clinical personnel (1 762 MFs = 51.6%) were to make change
happen at their workplace, this would account for improving nearly
half (46.2%) of the 2397 MFs reflected in Tables 7 and 8. Furthermore,
the tables reflect that these MFs could be reversed if clinical personnel
took more care and performed better in doing those activities that are
routine and standard of care: a detailed history, thorough clinical
examination, basic and relevant investigations, improved note-keeping,
comprehensive assessment of illness and appropriate management.
Table 7 also shows that more than half of MFs occurred in the health
system (56%) rather than at home. This reinforces the statements
above concerning the quality of care received by the patients.
Nevertheless, caregiver factors at home or the community were
considered to contribute nearly 40% of the MFs in TB deaths. Table 8
highlights the most common occurrences of MFs according to the
person responsible. It can be seen that clinical management is often
considered to be inadequate.
Table 7. Top five Modifiable Factors in TB at each site of occurrence (n = 2397)
Home
Health System (n=1 350)
(n=1047)
Clinic/OPD
Transit
A&E
Ward
Delayed seeking
care
Did not
recognise danger
signs or severity
of illness
Child not
provided with
adequate (quality
and/or quantity)
food at home
Caregiver took
child to clinic
infrequently
Road-to-Health
Card (RTHC)
not used or lost
by caregiver
(n=447)
IMCI not used
for case
management at
clinic/outpatient
department
(OPD)
(n=26)
Inadequate
ambulance
service from
health facility to
receiving
hospital
IMCI not used
for patient
assessment at
clinic/OPD
No or delayed
referral to higher
level
Delayed referral
for severe
malnutrition,
weight loss, or
growth faltering
from
clinic/OPD
(n=373)
Inadequate
investigations
(blood, x-ray,
other) at
accidents and
emergencies
(A&E)
Appropriate
antibiotics not
prescribed at
A&E
Inadequate
history taken at
A&E
Inadequate
physical
examination at
A&E
Did not arrive at
clinic/OPD on
day of
referral/did not
Not classified as
critically ill
93
(n=504)
Lack of
experienced
doctors (post
community
service), for
children's ward
Lack of
professional
nurse in
children's ward
24 hours a day
Inadequate
antibiotics
prescribed in
ward
Inadequate case
assessment and
management at
previous
admission to
ward
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
keep
appointment
D A T A
despite presence
of danger signs
at A&E
Child's growth
problem (severe
malnutrition, not
growing well)
inadequately
identified or
classified
Lack of hospital
beds and/or
ward
overcrowded
About 14% of MFs are related to lack of personnel, equipment and
beds. Administrators and managers need to be held responsible and
accountable for system failures.
Table 8. Most common Modifiable Factors by Responsible Person in TB (n = 2397)
Caregiver (31.9%)
Clinical Personnel (51.6%)
Administrator (16.5%)
Caregiver delayed seeking
care
IMCI not used for case
management at clinic/OPD
Caregiver did not recognise
danger signs/severity of
illness
Delayed referral for severe
malnutrition, weight loss, or
growth faltering from
clinic/OPD
Child not provided with
adequate (quality and/or
quantity) food at home
Inadequate investigations
(blood, x-ray, other) at A&E
Caregiver took child to clinic
infrequently
Appropriate antibiotics not
prescribed at A&E
Other caregiver MFs at
home/in community
Inadequate history taken at
A&E
Insufficient notes on home
circumstances or child's
health history
Inadequate case assessment
and management at previous
admission to ward
Lack of experienced doctors
(post community service), for
children's ward
Lack of professional nurse in
children's ward 24 hours a
day
Lack of hospital beds and/or
ward overcrowded
Inadequate resuscitation area
and/or trolley in ward
Lack of high care and/or
intensive care unit for
children in own and higher
level facility
No TB contact tracing or
treatment at home
South African national standards for caring for children
with TB
Standards of care must reflect the bare minimum that one has to do to
achieve the goal of quality health (TB) care. South Africa has
committed to follow the WHO guidelines for TB control, adopting the
94
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
DOTS, Stop TB and Global Plan to Stop TB strategies and therefore
must adhere to the “Standards” as well.
The International Standards for Tuberculosis Care5 developed in 2006
with prominent South African contributors serves as the source
document for any health care worker managing patients with TB.
“The Standards are intended to be complementary to local and national
tuberculosis control policies that are consistent with the World Health
Organisation (WHO) recommendations.” They clearly outline what is
expected of all role-players, including the patient for successful
outcomes.
In terms of clinical guidelines, the WHO published CH TB Guidelines
in 2006. 2 As a result, several national CH TB experts led a meeting in
Cape Town in 2007 concluding with a mandate to forge the way
forward towards developing a single, consensus guideline for South
Africa.6 International guidelines and tools are also available from the
International Union against Tuberculosis and Lung Disease, WHO and
several other organisations. The South African Society for Paediatric
Infectious Disease published their guidelines for CH TB in 2009,7 and
this is an excellent resource aiming at “providing consensus
management guidelines …” and “…current best practice guidelines in
resource-limited settings…”
The South African National Department of Health has updated and
published the “National Tuberculosis Management Guidelines 2009”.8
Chapter 11 traditionally deals with TB in children, and has
incorporated newer recommendations but there are still concerns.
These include: the use of the Bacille Calmette Guerin (BCG) vaccine in
HIV-infected newborns; the prophylaxis regimen and doses in children
with latent TB infection or those in contact with drug-resistant source
5
Tuberculosis Coalition for Technical Assistance. International Standards for Tuberculosis Care (ISTC). The
Hague: Tuberculosis Coalition for Technical Assistance, 2006.
6
Proceedings Paediatric TB Symposium & SATS CME – Combined Congress of SA Thoracic Society,
Allergy Society of SA and Cystic Fibrosis Association, Cape Town International Convention Centre, 02
March 2007. (Published in SAMJ, October 2007, Vol 97,No 10.
7
Moore DP, Schaaf HS, Nuttall J, Marais B. Childhood tuberculosis guidelines of the Southern African
Society for Paediatric Infectious Diseases. South Afr J Epidemiol Infect 2009; 24(3): 57-68.
8
RSA National Department of Health. National Tuberculosis Management Guidelines 2009.Pretoria, SA.
95
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
cases; the routine use of Isoniazid Prophylactic Therapy (IPT) in all
HIV-infected children after active TB disease has been excluded; the
extent to which one goes in terms of diagnostic tests to exclude active
disease, given the limited diagnostic yield of confirmed TB in children,
especially in HIV infected children; and drug dosages for children using
current preparations being inadequate given the weight bands in use.
Finally, the “Standard Treatment Guidelines and Essential Drug List”9
for hospital level care for children 2006 also needs to be updated.
Having several guidelines that do not always agree can be difficult and
the call in the “South African Child Gauge 2009/2010” to “clearly
define and formalise the process of updating and disseminating
treatment guidelines” is important.10 If there is more than one
guideline, all should be compatible with each other.
The National Tuberculosis Control Program manages the systems for
TB care to occur successfully, setting standards for prevention and
treatment, diagnostics in laboratories, pharmacy and drug supply,
reporting and recording, infection control, human resource and budget
needs.
Internationally and nationally, the resources needed for TB control are
not adequate, jeopardising progress being made. The emergence of
drug-resistant TB attests to this fact and also significantly increases the
costs of care. According to a publication in the Bulletin of the WHO in
200811 “African countries require the largest increases in funding” and
“achieving the 2015 global targets set for TB control requires a major
increase in funding.” Furthermore, some believe that the TB budget in
South Africa is inequitably skewed against children, further
compromising their TB care.
9
RSA National Department of Health. Standard Treatment Guidelines and Essential Drugs List Hospital
level Paediatrics 2006. Pretoria, SA.
10
Eley B. HIV, TB and child health. Kibel M, Lake L, Pendelbury S & Smith C eds. In South African Child
Gauge 2009/2010. Cape Town: Children’s Institute, UCT.
11
Floyd K & A Pantoja A. Financial resources required for tuberculosis control to achieve global targets set
for 2015. Bulletin of the WHO 2008; 86: 568-576.
96
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Controversies within the standard for caring for children
with TB
All clinical care guidelines must follow the evidence available at the
time of writing. In CH TB many grey areas and unanswered questions
persist. As a result, controversies will inevitably arise, including:
The place of BCG vaccine in the context of a high prevalence
of HIV and its complications in HIV infected infants.
Uncertainty about optimal prophylactic therapy such as IPT
in HIV-infected children, and the best regimen for childhood
contacts of drug-resistant TB cases.
Continuing developments in
combinations with higher
formulations for children,
probable TB disease when in
cases.
treatment such as fixed drug
drug doses for children,
regimens for children with
contact with drug-resistant TB
The increasing use of ethambutol in children.
The optimal management for those with HIV-TB coinfection, including aspects such as timing of HAART, the
Immune Reconstitution Inflammatory Syndrome and its
management.
Controversies in the retreatment regimens, such as the use of
injectable streptomycin as a single drug addition to a failing
regimen.
Congenital or perinatally acquired TB in neonates and infants,
and the optimum management of osteo-articular TB.
Challenges and further research needs
Education: Families, caregivers, clinical personnel should be
educated on TB, and proper undergraduate training of nurses
and doctors for TB should be provided.
Prevention measures: There are new vaccines and drugs for
effective, shorter treatment.
97
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Epidemiology and statistics: Clinicians may not fully
appreciate the total burden of disease that TB represents for
children and reporting and recording is not as pristine as it
should be. Additionally, the diagnosis is difficult in children
especially with HIV co-infection7 therefore the “true extent” is
not fully established. This makes planning and budgeting
services difficult.
Adults: TB control in adults to date has been poor and this is
reflected in WHO TB report trends in the last decade, so
spread to children is inevitable. For example, contact tracing is
poorly done as is seen from “missed opportunities” of
children who acquired TB from an adult on treatment (Child
PIP MFs12). This results in poor outcomes for individual
patients and results in increased costs for the health system.
Diagnostics: Newer laboratory and bedside diagnostic testing
is needed. This is even more difficult in high HIV burden
areas. This is mainly a research challenge with high activity.13
Laboratory services: Human resource capacity and equipment
should be improved. In addition, better microbiological
confirmation and Direct Sensitivity Testing is necessary as we
are also regarded by WHO and local data14 as a high MDR-TB
burden area, with children also presenting with drug-resistant
TB at the first episode of disease.
Summary and conclusion
The Child PIP data have confirmed the important role of TB in child
deaths. The close relationship between TB, HIV and malnutrition has
been highlighted. Death auditing has helped to uncover numerous
12
De Maayer T & Saloojee H. Clinical outcomes of severe malnutrition in a high tuberculosis and HIV
setting. Arch Dis Child (2011). doi:10.1136/2 of 6 adc.2010.205039
13
Marais B, Pai M. Recent advances in the diagnosis of childhood tuberculosis. Archives of diseases in
childhood 2007; 92; 446-452 doi:10.1136/adc.2006.104976
14
Fairlie l, Beylis NC, Reubenson G, Moore DP, Madhi SA. High prevalence of childhood multi-drug
resistant tuberculosis in Johannesburg, South Africa: a cross sectional study. BMC Infectious Diseases
2011.Vol 11:28 doi:10.1186/1471-2334-11-28
98
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
modifiable factors in the care of children with TB who died, allowing
extrapolation to routine management of such cases.
Improved clinical personnel awareness and adherence to clinical
guidelines and TB control program activities is a challenge for clinicians
and implementation managers of the TB control program. Clinicians
should be assisted through education and knowledge and skills updates
(especially with so many changes and the pace of changes in CH TB)
to enable thorough history and clinical assessments that will help
recognise TB infection for IPT and TB disease for treatment, proper
diagnostic efforts, proper recording and reporting on CH TB indices.
MFs related to clinical care practices and “missed opportunities” to
deliver the standard of care can be corrected by clinical personnel
through following the principles of thorough history-taking, meticulous
examination and holistic management taught during training.
In addition, clinical personnel should be trained to investigate patients
better and more intensively for diagnosis with the available tools (i.e.,
microscopy, culture and histology). A challenge for clinicians in our
high prevalence TB setting is the ongoing exposure of patients to the
bacillus, requiring repeated screening and ongoing surveillance as
recommended in the new Road to Health Card. Clinicians should
always maintain a high index of suspicion, especially in high-risk
patients (i.e., an HIV-infected child should be assessed for TB at every
visit).
Reporting and recording must be improved through proper use of the
TB register to get accurate data for monitoring performance and
planning of TB care for children (and adults). In addition, access to
HIV care should be improved by investigating the HIV status of all
patients diagnosed with TB.
Major efforts are needed at home and community level to help prevent
TB. Health promotion with clear healthcare knowledge and messages
about TB symptoms together with prompts as to when patients (and
their families) need to seek care will empower caregivers to seek care
early and not miss danger signs.
Further efforts at preventing TB in children include contact tracing
with early detection of susceptible at-risk (i.e., HIV-infected,
99
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
malnourished) children under five using growth monitoring in the
Road-to-Health Card and Integrated Management of Childhood
Illnesses protocols during clinic and ambulatory care visits.
Contact tracing and source case identification must be prioritised as a
major activity with appropriate capacity to be successful.15 Children add
to the new infections accounting for 14.7% of the total TB burden in
2008 and furthermore remain the future reservoir of TB for the
country. This must then be sustained with ongoing support done
through dedicated TB teams or as part of the infection control team at
institutions.
System failures are challenges that need to be overcome to improve
outcomes. These include staffing, equipment and drug supply and
quality.
Budgets are inadequate internationally and nationally. This needs to be
addressed, including increasing the public-private mix in TB care.
Finally, TB, especially TB meningitis, results in long-standing,
devastating effects on those who survive. Providing ongoing care for
these patients is a challenge for clinical personnel and families alike.
Acknowledgements:
Dr NH McKerrow (Chief Specialist Paediatrics), Pietermaritzburg Metropolitan
Hospitals Complex – Department of Paediatrics.
Dr S Kauchali (Senior Specialist), Pietermaritzburg Metropolitan Hospitals Complex –
Department of Paediatrics.
Prof PM Jeena – University of Kwazulu Natal, Department of Paediatrics and Child
Health.
15
Morrison J, Pai M, Hopewell PC. Tuberculosis and latent tuberculosis infection in close contacts of people
with pulmonary tuberculosis in low-income and middle-income countries: a systematic review and metaanalysis. Lancet Infectious Disease 2008; 8: 359-68
100
South Africa
Tuberculosis profile
| High TB burden | High HIV burden | High MDR-TB burden |
Population 2009 (millions)
Estimates of burden * 2009
Case notifications 2009
New cases
Smear-positive
139 468
Smear-negative
55 083
Smear unknown
92 104
Extrapulmonary
53 411
Other
0
Total new
340 066
Total < 15 years
49 825
Rate
(per 100 000 pop)
52 (29–85)
808 (362–1 288)
971 (791–1 169)
563 (461–675)
Number (thousands)
Mortality (excluding HIV)
Prevalence (incl HIV)
Incidence (incl HIV)
Incidence (HIV-positive)
Case detection, all forms (%)
Incidence (HIV+TB in orange), notifications (black)
(rates per 100 000 population)
50
26
400
490
280
74
(14–42)
(180–650)
(400–590)
(230–340)
(61–91)
(%)
(41)
(16)
(27)
(16)
(0)
Retreatment cases
Relapse
Treatment after failure
Treatment after default
Other
1200
800
400
0
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008
Prevalence (rate per 100 000 population)
Total retreatment
Total new and relapse
Total cases notified
360 183
405 982
20 117
2 895
5 671
37 233
(%)
(31)
(4)
(9)
(56)
1500
1000
500
0
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008
65 916
Mortality excluding HIV
(rate per 100 000 population)
(89% of total)
120
80
Drug regimens
Rifampicin used throughout treatment
% of patients treated with fixed-dose combinations (FDCs)
Paediatric formulations procured
Treatment success rate 2008 (%)
New smear-positive
New smear-negative/extrapulmonary
Retreatment
Treatment success rate (%)
80
New smear-/extrap
19
9
19 5
9
19 6
9
19 7
9
19 8
9
20 9
0
20 0
0
20 1
0
20 2
0
20 3
0
20 4
0
20 5
0
20 6
2007
08
New smear +
0
Retreatment
197 448
49
114 523
58
71
42
433 662
23 583
CPT (orange) and ART (green) for HIV-positive TB patients (blue)
120000
MDR-TB, Estimates among notified cases * 2008
% of new TB cases with MDR-TB
1.8 (1.5–2.3)
% of retreatment TB cases with MDR-TB
6.7 (5.5–8.1)
Estimated MDR-TB cases among notified
9 600 (7 900–12 000)
pulmonary TB cases 2009
MDR-TB reported cases 2009
0
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008
TB/HIV 2009
TB patients with known HIV status
% of TB patients with known HIV status
TB patients that are HIV-positive
% of tested TB patients that are HIV-positive
% HIV-positive TB patients started on CPT
% HIV-positive TB patients started on ART
HIV-positive people screened for TB
HIV-positive people provided with IPT
76
68
64
40
40
Yes
100
Yes
New
Retreatment
Cases tested for MDR-TB
% of notified tested for MDR-TB
Confirmed cases of MDR-TB
MDR-TB patients started treatment
Total
9 070
4 143
80000
40000
0
2003
2004
2005
2006
Financing
Total budget (US$ millions)
Available funding (US$ millions)
% of budget funded
% available funding from domestic sources
% available funding from Global Fund
2007
2008
2009
2010
386
2011
436
386
100
77
23
436
100
75
25
NTP Budget (blue) and available funding (green) (US$ millions)
Laboratories
Smear (per 100 000 population)
Culture (per 5 million population)
DST (per 10 million population)
Second-line DST available
National Reference Laboratory
* Ranges represent uncertainty intervals
2008
0.5
1.5
2.4
2009
0.5
1.6
3.2
In and outside country
Yes
2010
0.5
2
4
400
300
200
100
0
2006
Generated: March 14, 2011
2007
2008
2009
2010
2011
Source: www.who.int/tb/data
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
PART THREE
CHILDREN AT SPECIAL RISK
O F
D A T A
Chapter
9
The Very Young Child:
Neonates (0-28 days)
Cindy Stephen
MB ChB (UCT) DCH (SA)
Medical Officer, Department of Paediatrics, Pietermaritzburg Metropolitan Hospitals Complex
National Child PIP Coordinator
Abstract
A reduction in the number of deaths in the neonatal period (0-28
days) is a key component in efforts to reduce under-five mortality
in children.
Child PIP data describes more than 1 000 neonatal deaths in
detail. The in-hospital mortality rate (IHMR) for neonates is
increasing (to a high of 6.8% in 2009), which reflects deteriorating
care for newborns in the South African health system.
The majority of deaths are caused by infections. At least 36% of
newborns dying in South African hospitals are HIV-infected or exposed.
Many newborns are admitted to children’s wards and more than
half die within 24 hours of admission. Only one quarter of neonatal
deaths are considered unavoidable, with the majority of modifiable
factors relating to clinical personnel and infrastructural
inadequacies.
In an effort to fill the gaps identified in quality of care, an effective
system for post-natal care (with community-based strategies)
102
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
must be defined and implemented. In addition, sick newborns seen
in clinics and admitted to hospital must receive appropriate care
designed to meet their specific needs.
Improvement in overall neonatal data collection is essential for
appropriate health planning. Child PIP serves an important
function in providing information on the quality of neonatal care in
the South African health system.
Key Recommendations
Strengthen antenatal and primary obstetric care, with particular
emphasis on maternal HIV and tuberculosis (TB), and integrate these
services with newborn care.
Ensure skilled birth attendants are present at delivery and are also
competent in neonatal resuscitation.
Perinatal care must include:
- Full examination and assessment of the newborn;
- Early detection and referral of complications;
- Discharge 24 hours or more post-delivery; and
- Pre-discharge education of mothers feeding and care, recognition of
illness in their babies, and where to access help.
Develop a national uniform guideline for post-natal care and early
follow-up of discharged newborn babies during the first week of life.
Define and implement a formal policy on the principles of care for
hospitalised newborns.
Improve training and supervision of health workers, with an emphasis
on the Integrated Management of Childhood Illness (IMCI) guidelines
and neonatal care, and including the appointment of regional
neonatologists or equivalent medical practitioners.
Introduction
Newborns are a particularly vulnerable group of children in any
population, and especially in the developing world. The neonatal
period is only 28 days long, yet deaths during this period account for
more than one third of deaths in children under five years of age.
103
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Furthermore, because worldwide neonatal mortality rates have fallen
more slowly than overall under-five mortality rates, the proportion of
deaths which occur during the newborn period has increased over
time.1 Globally, neonatal deaths account for 40% of all under-5 deaths,
whereas in South Africa the proportion is lower at about 30%.2
Historically attention has focussed on perinatal care, which includes
early newborn deaths, but the Millennium Development Goal (MDG)
process, and MDG 4 in particular which emphasises reducing underfive deaths, has re-focussed attention on all newborn deaths, both early
(0-7 days) and late (8-28 days).
Reductions in the number of deaths during the neonatal period are a
key component of overall efforts to improve survival of young
children. Nevertheless, demographic and health information about
newborn deaths, especially those occurring in the late neonatal period,
is often incomplete. Current sources of neonatal data include vital
registration (StatsSA), the Perinatal Problem Identification Programme
(PPIP) and more recently, the Child Healthcare Problem Identification
Programme (Child PIP). It must be noted that both PPIP and Child
PIP include only a select group of newborns admitted to maternity
units or hospitals and therefore cannot contribute to population-based
mortality rate data. However, these data do form an important source
of quality of care information that can make a significant contribution
to reducing mortality in this group of vulnerable newborns.
In this chapter, Child PIP data are used to describe the experience of
newborns admitted to the children’s wards in South African hospitals
using the Child PIP mortality review process.
Health profile of newborns
In-hospital mortality rates
Child PIP data on all child admissions and deaths from 2005 to 2009
are shown in Table 1. A total of 23 127 newborns were admitted to
1
Lawn JE, Cousens S, Zupan, J. 4 million neonatal deaths: When? Where? Why? Lancet 2005; 365: 9-18.
2
Black E, Cousens S, Johnson L, et al. Global, regional, and national causes of child mortality in 2008: a
systematic analysis. Lancet 2010; 375:1969-1987.
104
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
participating hospitals during this period accounting for almost 7% of
all admissions to children’s wards.
The IHMR for children dropped over the years, whereas the IHMR
for neonates increased to a high of 6.8% in 2009. Although the
proportion of neonatal admissions remained almost constant over the
five years, the proportion of neonatal deaths increased, as reflected in
the increasing neonatal IHMR, particularly during the last three years.
Table 1. Admission and death Child PIP data with IHMRs (2005-2009)
Year
2005
2006
2007
2008
2009
Total
NN admissions
NN deaths
Neonatal IHMR
All child admissions
All child deaths
Paediatric IHMR
NN proportion of all
admissions
1 355
76
5.6
23653
1543
6.5
2 769
150
5.4
40665
2393
5.9
4 354
192
4.2
63378
3190
5.0
7 867
440
5.6
106860
5379
5.0
6 602
426
6.8
108852
5398
5.0
23 127
1 284
5.6
343408
17903
5.2
5.7
6.8
7.2
7.4
6.1
6.7
Newborn deaths accounted for approximately 6% of all audited deaths
in Child PIP, with early newborn deaths contributing approximately
40% of the neonatal deaths, as shown in Table 2. A striking feature is
the increasing proportion over the years of early deaths (0-7 days)
versus late neonatal deaths (8-28 days).
On average, 2.5 modifiable factors were identified for each audited
neonatal death, but the rate has increased from 1.7 per death in 2005 to
3.2 per death in 2009. This may reflect either deterioration in neonatal
care, and/or greater attention to the audit process of the newborn
deaths.
Table 2. Audited newborn deaths (2005-2009)
Year
2005
2006
2007
2008
2009
Total
Early neonatal deaths
Late neonatal deaths
% Early neonatal deaths
All neonatal deaths
NN proportion of all
audited deaths
Neonatal MFs
MF rate (per death)
8
23
26%
31
35
87
29%
122
84
123
41%
207
122
206
37%
328
184
193
49%
377
433
632
41%
1 065
2.0
4.2
5.4
5.9
6.8
5.5
56
1.7
206
1.7
394
1.9
810
2.5
1 200
3.2
2 664
2.5
105
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Nutritional status and feeding practice
About one half of neonates as shown in Figure 1 were below normal
weight for age (UWFA and severe malnutrition), which most probably
reflects low birth weight in most cases. However, the proportion of
neonates with severe malnutrition increases from 7% in early neonates
to 11% in the late neonatal group, suggesting that poor feeding
practices may also play a role.
Of concern is that the weight of 9% of neonates dying in hospital was
not known.
With regard to feeding practice, whilst the feeding choice for 37% of
newborns was not known, almost 60% of sick babies (where the
choice was known) did not receive exclusive breastfeeding.
Figure 1. Nutritional status: 2005-2009
Figure 2. Feeding practice: 2005-2009
HIV context
The HIV status for half of all neonates dying in children’s wards was
unknown as shown in Figure 3. About 36% of babies dying in hospital
were either HIV-exposed or already diagnosed as infected with HIV,
and 13% tested HIV-negative.
A similar picture emerged when looking at babies receiving nevirapine
and/or AZT as part of prevention of mother-to-child transmission of
HIV (PMTCT) shown in Figure 4, where, for almost half of the babies
who died, it was unknown whether they received perinatal
antiretrovirals (ARVs). Furthermore, nearly half of the babies known to
have been eligible for perinatal ARVs did not receive them. At least
one third of babies dying had mothers known to be HIV-infected.
106
S A V I N G
C H I L D R E N
Figure 3. HIV status of neonates 2005-2009
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Figure 4. Perinatal ARVs 2005-2009
It is important to remember that these HIV data highlight significant
gaps in quality of care rather than absolute HIV surveillance data.
Cause of death
The majority of deaths in neonates were caused by infections. Of the
infections, septicaemia was the most common cause, followed by
pneumonia, diarrhoeal disease, meningitis, PCP, and surprisingly
perhaps in this age group, TB was the sixth most common cause of
death. HIV also contributes significantly with more than one third of
newborn deaths being either HIV-exposed or HIV-infected, as
described above.
Figure 5. Cause of death in neonates 2005-2009
36% of
deaths HIV
exposed or
infected
Quality of care for newborns
Child PIP data describing the quality of care received by neonates in
children’s wards in South Africa are vital for identifying problem areas
in care and beginning to formulate solutions. HIV quality of care has
already been described and data relating to other indicators of care will
107
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
be highlighted in this section, particularly length of stay, place of death
and modifiable factor data.
Length of stay
It is significant that 50% of all neonates dying in children’s wards died
within 24 hours of admission as shown in Figure 6, which is a much
higher proportion than that recorded for all child deaths (32%).
Furthermore, a higher proportion of these deaths occurred in the first
week of life (58%) as compared to deaths from 8-28 days of life (45%).
Figure 6. Length of stay for neonates 2005-2009
Modifiable factors
For newborn deaths, 59% of all modifiable factors were identified in
hospitals, almost equally distributed between the emergency
department and the wards, as compared to 50% for all children. Nine
percent were identified in clinics, and 29% at home. This means that
nearly 1.5 modifiable factors per death arose in the hospital care of
these babies.
About half of all modifiable factors were attributed to clinical
personnel, about one fifth to administrators, and the remainder to
caregivers. The modifiable factor rate attributed to clinical personnel
was 1.3 reflecting that clinical personnel were responsible for more
than one modifiable factor per death.
108
S A V I N G
C H I L D R E N
Figure 7. MFs (%): Where? 2005-2009
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Figure 8. MFs (%): Who? 2005-2009
The leading modifiable factors in newborn deaths are listed in Table 3
(note that there may be more than one modifiable factor identified per
death).
Of significance in hospital is the frequent inadequacy of basic clinical
care practices, particularly assessment in the emergency setting (11% of
deaths), and routine monitoring in the wards (17% of deaths). A lack of
infrastructure and staff were noted in at least 11% of deaths.
In clinics, IMCI is not being used effectively contributing to 7% of
deaths, and in the community and at home, delays in seeking care as
well as the inability to recognise the severity of a newborn’s condition
contribute to very significant proportion (47%) of avoidable deaths.
Having reviewed all the modifiable factors, healthcare workers
considered that three quarters of neonatal deaths occurring in
children’s wards were potentially avoidable. This amounts to a very
significant number of deaths, many due to infections, where a
difference in overall care could have changed the outcome for the
baby.
Table 3. Modifiable factors according to where they occurred 2005-2009 (n=1 065)
% of
Care in the Ward
No.
deaths
Monitoring (e.g. RR, SATS, blood glucose)
185
17.4
Buildings/Beds (e.g. lack high care/ICU facilities)
Clinical management (e.g. IV fluids incorrect)
Staff (e.g. lack of professional nurse and experienced doctors)
119
76
74
11.2
7.1
6.9
Care in Admission and Emergency
No.
Monitoring (e.g. O2 saturation, blood glucose)
Assessment (e.g. RR not taken, not classified as critically ill)
Clinical management (e.g. appropriate antibiotics not prescribed )
Investigations (e.g. inadequate bloods, x-ray etc.)
118
114
81
74
% of
deaths
11.1
109
10.7
7.6
6.9
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Care in Clinics
No.
Assessment (e.g. IMCI not used, malnutrition not identified)
Clinical management (e.g. inadequate fluids in DD, delay in referring
acute
Notesproblems)
(e.g. on clinical care)
Treatment (e.g. IMCI not used)
74
45
26
19
Care at Home
No.
Care seeking and Compliance (e.g. delay, recognition of danger signs)
Home treatment (e.g. inappropriate remedies)
Growth & Development (e.g. inadequate nutrition)
Notes (e.g. home circumstances, health history)
495
84
58
47
% of
deaths
6.9
4.2
2.4
1.8
% of
deaths
46.5
7.9
5.4
4.4
South African national standards for newborn care
Although National Maternity and Newborn Guidelines are available
from the National Department of Health (www.doh.gov.za), the
newborn section is brief and not widely known nor used, and many
provinces have developed their own neonatal guidelines for clinical use.
However, both IMCI and the Standard Treatment Guidelines and
Essential Drugs List (EDL)3 contain widely accepted guidelines for
assessing and treating sick neonates.
The Newborn Care Charts, developed by the Limpopo Initiative for
Newborn Care4 (LINC) and based on IMCI principles, are an excellent
resource for the management of sick and small newborns in hospital.
Apart from a comprehensive approach to routine newborn care, they
include specific guidelines for the management of respiratory distress,
sepsis, jaundice, diarrhoeal disease, low birth weight infants, syphilis,
tuberculosis and HIV. The following summary provides an overview
of basic neonatal care standards as described in the charts.
3
Standard Treatment Guidelines and Essential Drugs List for South Africa: Hospital Level Paediatrics
National Department of Health, South Africa, 2006.
4
Newborn Care Charts, Management of Sick and Small Newborns in Hospital, Limpopo Initiative for
Newborn Care, University of Limpopo and Department of Health, Limpopo Province, First edition, 2009.
110
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Routine care at birth
Dry baby with warm towel.
Resuscitate if needed (i.e., assess baby, and ‘No’ to any of the
following questions):
- Is the baby breathing?
- Is the heart rate > 100/min?
- Is the baby centrally pink?
Check and record Apgar score.
Observe and assess for risk factors requiring admission to a
neonatal ward by examining baby fully as follows:
- Birth weight > 4 kg or < 2 kg
- Any respiratory distress (grunting, fast breathing, chest
indrawing)
- Prolonged rupture of membranes (> 18 hours)
- Maternal infections (especially HIV, syphilis and TB)
- Floppy baby
- Congenital abnormality
Give Vitamin K (1 mg intramuscularly in anterolateral aspect
of thigh).
Provide eye prophylaxis (chloramphenicol eye ointment to
both eyes).
Initiate bonding and feeding with skin-to-skin contact with
mother.
Routine postnatal care prior to discharge
Conduct a full examination of baby from head-to-toe, using a
checklist (e.g., the Maternity Case Record, NDOH).
Give vaccinations as follows:
- BCG
- Polio drops
111
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Check treatment of maternal infections and manage
appropriately:
- HIV: Assess mother’s HIV status and treatment received,
and manage baby according to national guidelines.
- Syphilis: If RPR reactive and treatment incomplete in
clinically normal baby give benzathine penicillin 50 000u/kg
imi stat; or if symptomatic, give procaine penicillin
50 000u/kg imi daily for 10 days.
- Tuberculosis: If mother has been on treatment for > 2
months, give baby INH for 6 months. If treatment < 2
months or active maternal disease or HIV infected, give baby
three drug treatment for 6 months.
Check for jaundice (daily TSB) if mother has O blood group
or is Rhesus negative.
Ensure feeding established, preferably breast-feeding
Complete the Road to Health Chart (RTHC) for the baby.
Counsel mother clearly about where and when to return for
routine follow-up, and to return immediately if danger signs
develop in baby (see box below).
Routine postnatal follow-up (Discharge after 24 hours, if possible)
All babies: Day 3 at facility or at home (healthcare worker visit), at 6
weeks, 10 weeks, 14 weeks and 6 months.
Low birth weight babies: Day 3, then weekly until 2.5 kg, at 6 weeks,
10 weeks, 14 weeks and 6 months.
High risk babies: Day 3, weekly until 2.5 kg, at 6 weeks (with a PCR
in HIV-exposed babies), 10 weeks, 14 weeks, at 4 months, 6 months
and 9 months.
Check weight gain and growth, assess feeding, immunise according to
schedule, and observe for and presence of priority signs at each visit.
112
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Routine care for sick neonates at clinics
IMCI clearly recognises neonates as a vulnerable group as it contains
specific guidelines for managing the sick young infant (birth up to 2
months). These guidelines describe the assessment, classification and
first-line treatment for neonates with sepsis (possible serious bacterial
infection and jaundice), diarrhoea and HIV infection. They also
provide healthcare workers with guidance regarding feeding and
growth monitoring, as well as immunisations.
Danger signs for which mothers are advised to return immediately to
the clinic include:
× Breast-feeding poorly or drinking poorly
× Irritable or lethargic
× Vomiting everything or diarrhoea
× Convulsions
× Fast breathing (> 60/min) or difficult breathing (grunting
or chest indrawing)
× Fever (> 38 ̊ C)
× Blood in stool
(Based on IMCI, the current NDOH RTHC of 2011, LINC and Opportunities
5
for Africa’s Newborns )
Routine care for sick neonates requiring hospital admission
Warmth: maintain body temperature > 36 ̊ C.
Oxygen for babies (via head box with Venturi, nasal prongs,
CPAP or ventilation):
- During resuscitation
- With severe hypothermia
- With respiratory distress (RR > 80/min, severe chest
indrawing or grunting, SATS < 88% or central cyanosis)
5
Opportunities for Africa’s Newborns: Practical data, policy and programmatic support for newborn care in
Africa. Joy Lawn and Kate Kerber, eds. PMNCH, Cape Town, 2006.
113
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Maintain normal blood glucose, especially in small, sick babies
and infants of diabetic mothers, and monitor for at least the
first 24 hours (must be > 2.5 mmol/l).
Feeds to be given at three hourly intervals if birth weight >
1.5 kg via breast (or cup), or nasogastric tube if unable to
suckle OR if baby to be kept nil per mouth, commence IV
fluids.
Infection control (hand washing, isolation for babies with
gastroenteritis, proper cleaning of equipment, no
overcrowding, and adequate staff who are patient-allocated
rather than task- allocated).
Condition-specific management according to guidelines,
either EDL, Newborn Care Charts or locally adapted and
accepted standards.
Note: Returning neonates must not be admitted to a paediatric
ward but should be sent to a newborn ward or cubicle specifically
equipped and staffed for their needs.
Controversies
For increased survival of newborns it is vital that their mothers receive
adequate antenatal care and skilled perinatal care during labour and
delivery, all newborns are offered effective postnatal care (PNC), and
sick newborns have access to appropriate facilities and proper care.
There is reasonable consensus on the contents of effective PNC but
controversies remain, about the timing and frequency, and about who
should provide it. 5
The World Health Organisation (WHO) 1998 model of PNC suggests
visits within 6 hours of birth, 3 to 6 days, 6 weeks and 6 months (‘6-66-6’).6 However the most critical time for newborns is the first 24 hours
and three-quarters of neonatal deaths occur during the first week. 1 It
has therefore been suggested that discharge be delayed until 24 hours
after birth and that the first postnatal visit should happen on the
6
WHO. Postpartum care of the mother and newborn: a practical guide. Report of a technical working group.
1998. Geneva, Switzerland: World Health Organization. (WHO/RHT/MSM/98.3)
114
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
second or third day of life and, if resources allow, a second visit
towards the end of the first week (Day six).
The current South African guideline contained in the Maternity Case
Record published by the Department of Health regarding postpartum
visits states that these should be conducted at a postnatal clinic, if
possible, at three days of life and again at six weeks. It is implied that
the baby must also be seen but there is no explicit checklist of aspects
to attend to at these visits.
PNC can take place at a facility, or at home, or a combination of both.
Currently, home visits by skilled healthcare workers are not available in
South Africa due to resource constraints. The ideal may be for a skilled
community health worker (CHW) to visit the home during the first
week of life (preferably day two or three), with subsequent postnatal
visits to the facility.
Regarding appropriate care for sick newborns admitted to hospital,
there is still considerable controversy around admitting babies from the
community back into the nursery due to the presumed risk of
infection. It is essential that this prejudice be reversed and that
newborns be admitted to facilities specially equipped and staffed for
them, including proper infection control measures.
Lastly, data describing the newborn period, whilst they have improved
over time, remain incomplete, especially for the late neonatal period.
Controversies exist particularly around neonatal mortality rates for
South Africa which vary from the UN estimate of 21 per 100 births7 to
approximately 10 per 1000 reported by StatsSA8 and the most recent
Saving Babies 2008-2009 report.9 These differences are probably due to
basic under reporting and inadequate health information systems.
7
Countdown to 2015: Maternal, Newborn and Child Survival. UNICEF and WHO, 2010.
8
Based on StatsSA figures adjusted by Bradshaw D based on adjustment factor described in: Darikwa TB
(2009) Estimating the level and trends of child mortality in South Africa, 1996–2006 MPhil, University of
Cape Town.
9
Saving babies 2008-2009: Seventh report on perinatal care in South Africa. Ed RC Pattinson. Tshepesa
Press, Pretoria, 2011.
115
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Challenges and further research
There are a number of challenges and opportunities for research
concerning the optimal care provision for newborn children.
Antenatal and perinatal care needs to be strengthened, with special
emphasis on skilled birth attendants, neonatal resuscitation and routine
care for well and sick newborns.
A feasible and adequate package of postnatal care urgently should be
defined (what, when, who and how) and implemented in the South
African context.
Integration of postnatal care between the community and facilities,
with an emphasis on empowering communities, and with other health
programmes (particularly HIV and TB), remains an urgent challenge.
This includes ensuring that health records, particularly the Maternity
Case Record and the RTHC, contain appropriate newborn care
strategies and health messages.
Resources should be specifically allocated to provide sufficient
neonatal-friendly facilities for hospitalised newborns and to ensure that
equipment, guidelines and staff are in place.
Specific pre- and in-service training in newborn care for healthcare
workers (including CHW), as well as supervision enabling healthcare
workers to ‘do what they know’, would significantly impact newborn
survival. The appointment of regional neonatologists, or an equivalent
medical practitioner, supported by medical schools is also essential in
meeting the challenges of improving overall neonatal care.
Operational research to assess the feasibility and impact of the various
models of postnatal care in the South African context, as well as
improved data collection and the monitoring of care and outcomes for
newborns, would assist in appropriate health planning and provision
for this group of very young children.
Summary and conclusion
The IHMR for neonates has increased in the five-year period under
review, reflecting an increasing proportion of neonatal deaths,
especially in the early neonatal period. Large numbers of neonates (7%
116
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
of all admissions) are still admitted to children’s wards where their
specific needs such as extra warmth (e.g., Kangaroo Mother Care),
special feeding requirements (e.g., breast milk) and protection from
infectious diseases (e.g., diarrhoeal disease and acute respiratory
infections), are difficult to meet.
The reasons for the worsening survival of newborns are not clear but
may include overwhelmed obstetric services, an ill-defined and poorly
implemented system of post-natal primary health care and poor
community understanding of healthcare needs of newborns,
specifically danger sign recognition. Particularly in the case of early
neonatal deaths, the question must be asked whether these babies were
already ill at discharge, whether the babies were not seen for their
scheduled postnatal visit for a number of possible reasons, or whether
postnatal care had been inadequate.
Almost half of the newborns dying in hospital are underweight which
may largely be accounted for by low birth weight infants, but the
majority is not exclusively breastfed. Inadequate feeding could
contribute not only to poor nutrition but also to the increased risk of
sepsis. Breast-feeding is known to have a very strong protective effect
against infections and must be reinforced at all times, particularly as the
implementation of the new PMTCT policy will reduce the risk of HIV
transmission through breast milk to very low levels.
Infections are the leading cause of death in the newborn period, with
septicaemia being the main contributor, followed by pneumonia,
diarrhoeal disease, meningitis, PCP and TB.
More than one third of neonates are HIV-infected or -exposed. The
HIV status for more than half of the neonates is unknown, and of
those babies eligible for perinatal antiretroviral therapy (ART), only half
receive it. These omissions represent enormous gaps in perinatal HIV
care, particularly as all pregnant women should be tested for HIV
during pregnancy, both at the initial booking visit and again at 32
weeks. HIV remains rampant and there is a great need for
improvement in the quality of perinatal HIV care, particularly
integration of antenatal ART, labour ward, nursery and neonatal
PMTCT.
117
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
An alarming 50% of newborns admitted to hospital die within the first
24 hours, with the majority being under one week of age. This may
reflect a combination of factors including inadequate routine post-natal
care with too early discharge post-delivery even after complications of
pregnancy and labour, late presentation for a variety of reasons (such as
the rapid progression and often non-specific features of illness in
newborns, inadequate caregiver knowledge and/or lack of transport),
and a failure to provide adequate emergency care for sick newborns at
clinics and hospitals.
Many gaps in the quality of care provided to newborns in the health
system can be identified. The most significant are inadequate basic
clinical care practices, especially the monitoring and assessment of sick
newborns in clinics and hospitals, as well as the inadequate provision of
emergency care for neonates. Both of these highlight the need for
improved and specialised neonatal education for nurses and doctors, at
nursing colleges and medical schools, as well as continuous in-service
training.
In the home and community, there is the ongoing serious problem of
delay in seeking care which has been discussed above and has also been
shown by PPIP in the Saving Babies reports10 and by the Confidential
Enquiry into Maternal Deaths in the Saving Mothers reports11 and by
Child PIP for older children in the Saving Children reports.12
Resource allocation problems may prevent hospitals from providing
facilities designed for small babies, with the appropriate equipment and
skilled clinical personnel. This means that ill newborns are most often
admitted to paediatric wards, which are not designed or equipped to
care for them. It is essential that this practice be stopped and that
newborns requiring re-admission are admitted to nurseries or
equivalent facilities specifically designed for their needs.
10
11
12
Saving Babies, see www.ppip.co.za
Saving Mothers, see www.doh.gov.za/docs/reports/2004/savings.pdf
Saving Children, see www.childpip.org.za
118
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Chapter
10
The Critically Ill Child:
Deaths within 24 hours
Mark Patrick
MBChB, DCH, FCPaed
Paediatrician, Grey’s Hospital and Pietermaritzburg Metropolitan Hospitals Complex
Chairperson of Child PIP
Abstract
The pathway that sick children follow through the health system
is, at its simplest: home, clinic, admissions and emergency (A&E),
and children’s ward. Emergency care is provided to critically ill
children in the casualty and outpatients settings, or in children’s
wards soon after arrival, having been transferred from another
hospital. Children who die within 24 hours of admission, will have
arrived at hospital critically ill.
Child PIP data can describe the health profile of and quality of care
received by children who died within 24 hours of admission.
In the 5-year period from 2005 to 2009, 6 074 children died within
24 hours, and 507 were dead on arrival, in the Child PIP sites,
amounting to 34% of all child deaths. The three main causes of
death were diarrhoeal disease (27%), acute respiratory infections
(26%) and sepsis (21%), on a background of the rampant HIV
pandemic (21% exposed; 19% infected). In the vulnerable neonatal
age group, 51% died in the first 24 hours.
Modifiable factors occurred at a rate of 3 per death for children
dying within 24 hours. These modifiable factors were spread
throughout the health system, but 53% occurred within the
119
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
hospital setting, with more than half occurring in A&E.
Only one fifth of the deaths were regarded as unavoidable,
meaning that approximately 4 800 children in this survey could
have survived had their process of caring been different.
Addressing problems identified with the provision of emergency
care to critically ill children may lead to an improvement in the
quality of emergency care provided in the South African health
system, and to a significant reduction in child deaths.
Key Recommendations
Policy
A national plan for improving the quality of emergency care for critically ill
children in South Africa should address:
Providing improved training in and support for community Integrated
Management of Childhood Illness (IMCI);
Ensuring accessible health care and efficient referral systems
including communication and transport;
Training in emergency triage and assessment for health workers;
Training in and application of critical care guidelines within health
facilities; and
Allocating emergency care resources to children.
Administration
Clinical managers must be held responsible for providing:
Appropriate infrastructure (i.e., adequate care facilities for children,
communication, transport);
Sufficient staffing with adequate training and skills; and
Appropriate equipment and consumables, in particular all drugs and
antibiotics according to the Essential Drugs List.
120
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Clinical Practice
Experts in the provision of emergency care for children, regional hospital
paediatricians and district hospital family medicine practitioners should:
Define the training requirements for paediatric critical care in South
Africa; and
Agree on a uniform protocol of triage and emergency care for varying
levels of health care in South Africa.
Trainers and supervisors of junior doctors working in frontline clinical care
situations must ensure:
Competence in triage and emergency care of critically ill children;
Compliance with the standard of care in initial clinical management of
sick children; and
Familiarity with and rigorous application of the guidelines for all
critical conditions.
Education
Certified paediatric and neonatal emergency care training and regular recertification must become a prerequisite for registration and practice for
all doctors, professional nurses and paramedics (all grades).
Introduction
Previous Saving Children, Saving Babies and Saving Mothers reports
have all documented the problem of patients and caregivers not
recognising severity of illness and danger signs, and of presenting to
health services late.1 By implication, this group of mothers and
children is arriving at hospital in a critically ill state, and probably dying
soon after arrival, with advanced disease.
For children dying in children’s wards, it is reasonable to assume that
those dying within 24 hours of arrival in hospital would have presented
critically ill. For the purpose of this chapter death within 24 hours of
arrival in hospital is taken as the proxy for having been critically ill on
arrival.
1
In Child PIP, ‘Admissions and Emergency’ is taken to mean the point and event of entry into the hospital in
which the child died. The point may be an outpatient or casualty department, or the children’s ward if the
child was transferred from another hospital. The event is the admission process (see also ‘Glossary’).
121
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
The pathway that children follow through the health system is, at its
simplest: home, clinic, A&E,2 and ward. It is notable that in South
Africa excellent ‘nationalised’ paediatric guidelines are available for all
the points on this pathway, with the notable exception of guidelines for
A&E.
This chapter describes the health profile of, and quality of care received
by, children dying within 24 hours of arrival in hospital, and provides
some information on those children who were dead on arrival.
Health profile of the children who died within 24 hours
of arrival in hospital
During the 5-year survey period, for the 343 408 admissions there were
19 295 audited deaths, of which 507 (3%) were already dead on arrival
in hospital and 6 074 (31%) died within 24 hours of arrival in hospital.
Over the 5-year period the proportion of children dead on arrival and
dying in the first 24 hours varied little from 31 to 34%.3
Table 1. Number (and %) of children dead on arrival or dying within 24 hours per
year
Dying in
< 24 hrs
Dead on
arrival
Total
2005
No.
%
2006
No.
%
2007
No.
%
2008
No.
%
2009
No.
%
Total
No.
%
450
29.3
864
30.1
1189
31.1
1788
32.3
1783
32.3
6074
31.5
33
2.1
102
3.6
117
3.1
141
2.5
114
2.1
507
2.6
1537
31.4
2871
33.6
3828
34.1
5539
34.8
5520
34.4
19295
34.1
Age
Of the 1 065 neonates (0-28 days) who died, for 538 (51%) this
occurred within 24 hours of admission (48 were dead on arrival). This
is a far greater proportion than the 30% of children aged 28 days to 5
years who died within 24 hours. This emphasises the vulnerability of
the neonatal population and its susceptibility to more rapid disease
2
In Child PIP, ‘Admissions and Emergency’ is taken to mean the point and event of entry into the hospital in
which the child died. The point may be an outpatient or casualty department, or the children’s ward if the
child was transferred from another hospital. The event is the admission process (see also ‘Glossary’).
3
The health profile of children dying before arrival is similar to that of those dying within 24 hours, and data
from those dying within 24 hours only, are reported in this chapter.
122
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
progression. It may also indicate difficulty in assessing severity of illness
in this age group, and suggests that children’s wards are not the place
to care for critically ill neonates (see also Chapter 9).
Nutritional status
The weight category distribution shows a slight difference among all
children, with the proportion of children of normal weight for age at
early death being 34% as compared with 28% for all children. There is
therefore a higher proportion of normal weight children dying within
24 hours than after longer stays in hospital, indicating that
overwhelming disease is often of acute onset in otherwise normal
children.
HIV infection
For 26% of children who die in the first 24 hours, their HIV status is
unknown, this being a bigger proportion than for all children. This may
indicate a breakdown in the testing and documentation of HIV
infection in children during previous encounters with the health
system. It is surprising that in 9% of children, an HIV test was thought
by the health workers responsible for the child’s care, to be ‘not
indicated’, even though the child died, and even though South Africa is
at the global epicentre of the pandemic.
Nineteen percent, as compared with 28% of those children dying later,
were known to be HIV-infected. This may not reflect at-admission
HIV status as results indicating HIV infection may only have become
available post mortem. The same could be said of the 21% recorded as
being HIV-exposed. Knowledge of the at-admission HIV status is
therefore likely to be even lower than these levels, further emphasising
the poor quality of HIV testing and of documentation for HIV
exposure and infection in South African children. However, the HIV
status may not always be of critical importance in the immediate
survival from acute overwhelming disease.
HIV clinical category
Nineteen percent of children dying within 24 hours were known to be
HIV-infected and 33% were staged as I, II, III or IV. Thus 14% of
children dying within 24 hours were assigned an HIV clinical stage
123
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
without being confirmed as having HIV infection. The impact of this
finding on clinical care planning may need to be elucidated.
Cause of death
The three main causes of death in the first 24 hours (Figure 1) were
diarrhoeal disease (27%), acute respiratory infections, including
pneumocystis pneumonia (26%) and sepsis (21%). Forty percent of
these children were known to be HIV-exposed (21%) or -infected
(19%). Diarrhoeal disease is overall the third most important cause of
death. In the A&E setting, its greater prominence as a cause of early
death may be due to problems with the management of shock and
severe dehydration. Early death in respiratory conditions is commonly
associated with an inability to maintain oxygenation. Septicaemia
deaths relate to septic shock and multi-organ failure: in all these
conditions it is critical to provide adequate fluid resuscitation,
oxygenation, early administration of antibiotics and on-going
monitoring in order to save the life.
Figure 1. Main cause of death
Quality of care received by the children who died within
24 hours of arrival in hospital
It is logical that some children who die soon after arrival in hospital or
who are dead on arrival may have experienced problems with caring at
home or in their community setting, and/or at their referring health
facility, and/or (for those arriving alive) with the emergency care they
received on arrival at hospital. By analysing this group of children in
124
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Child PIP, it is possible to begin describing the quality of care they
received, with a view to making recommendations for improvement.
Referral status
Forty-three percent of children dying within 24 hours and 29% of
children who were dead on arrival were referred from elsewhere in the
health system. This might indicate serious and deep problems with pretransfer assessment and care, and a possible need for improving quality
of emergency assessment and care at referring sites (i.e., all sites) within
the health system. It may also reflect lack of recognition of danger signs
and delayed presentation to the first point of entry into the health
system.
Readmissions
In Child PIP, a ‘readmission’ is defined as a child presenting again to
the health service within four weeks of discharge for the same or
similar condition. Eighteen percent of children dead on arrival, and
17% of those dying within 24 hours were regarded as re-admissions.
This suggests problems with premature discharge, or with the child’s
caregiver not being given adequate advice about danger signs indicating
relapse or deterioration. Re-admissions may also result from premature
discharge enforced by inadequate resource allocation to sick children
(beds and staffing) or to industrial action.
Modifiable factors
For the 6 074 children dying within 24 hours, 19 123 modifiable factors
were recorded, giving a rate of 3.1 modifiable factors per death. Fiftythree percent of these modifiable factors occurred within the receiving
hospital, with more than half (55%) occurring in the emergency
department. Thirty-one percent occurred at home (Figure 2). Fifty-six
percent of the modifiable factors were attributed to clinical personnel,
30% to caregivers, and 13% to administrators (Figure 3).
125
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
4
Figure 2. Modifiable factors (%): Where? Figure 3. Modifiable factors (%): Who?
The leading modifiable factors in children dying within 24 hours are
listed in Table 2. The data highlight inadequate high and intensive care
facilities in hospitals, as well as staffing deficiencies. With specific
regard to emergency care, deficiencies identified have largely to do with
the assessment and monitoring of critically ill children. At clinic level,
failure to use the IMCI system was a frequent occurrence.
At home, the major problems are delay in seeking care and nonrecognition of severity of illness.
Table 2. Modifiable factors according to where they occurred 2005-2009 (n=19 123)
Care in the Ward
Lack of high care and/or intensive care units (ICU) for children in own and higher level
facility
Inadequate investigations in ward
Inadequate monitoring of respiratory rate and/or oxygen saturation in ward
Lack of professional nurse in children`s ward 24 hours a day
Inadequate monitoring of blood glucose in ward
Care in Admission and Emergency
Inadequate history taken
Inadequate investigations (blood, x-ray, other)
Inadequate physical examination
Inadequate assessment of shock
Appropriate antibiotics not prescribed
4
Child PIP user feedback indicated a need for categorising and recording modifiable factors occurring in the
referring facility and during transport. This was done and incorporated into the programme in 2009.
Quantification of transit-related factors only began in 2009.
126
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
Transit Care
Inappropriate care or late referral from private sector/general practitioner
Inadequate ambulance service from health facility to receiving hospital
No or delayed referral to higher level
Severity of child`s condition incorrectly assessed at referring facility
No ambulance available for transfer from referring to receiving hospital
Care in Clinics
IMCI not used for patient assessment
Inadequate notes on clinical care (i.e., assess, classify, treat)
IMCI not used for case management
Growth problem (severe malnutrition, not growing well) inadequately identified
Inadequate fluid management for diarrhoeal disease with dehydration
Care at Home
Caregiver delayed seeking care
Caregiver did not recognise danger signs/severity of illness
Child not provided with adequate (quality and/or quantity) food at home
Inappropriate treatment given at home with negative effect on the child (e.g., enema)
Insufficient clinical notes on home circumstances or child`s health history
Was the death avoidable?
After reviewing each death in the Child PIP process, an assessment is
made taking all modifiable factors into consideration, of whether the
death was avoidable. For children dying within 24 hours, only 21%
were regarded as NOT avoidable. Twenty-eight percent were regarded
as avoidable and in 37% there was uncertainty.
It is possible that of the 6 074 children who died, approximately 4 800
could still be alive had their process of caring been different.
Figure 4. Was the death avoidable?
127
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
South African national standards for providing
emergency care for critically ill children
There are several systems for providing emergency care to critically ill
children in South Africa, but there is no national consensus on the
optimal system.
The South African Standard Treatment Guidelines and Essential
Drugs List (EDL),5 Hospital Level, Paediatrics contains a chapter on
Emergency Care and Trauma, which has an algorithm for provision of
emergency care. This is the closest South Africa comes to having a
national standard for provision of emergency care.
Paediatric Life Support courses are also provided countrywide, and
attendance of these courses is ad hoc, and based primarily on individual
health worker enthusiasm and dedication. No medical schools require
certified training in emergency care as a prerequisite for graduation as a
medical intern.
The WHO Emergency Triage Assessment and Treatment (ETAT)
course has been piloted in three centres in South Africa during 2010.
ETAT includes triage of sick children and is based on recognised life
support teaching and protocols. ETAT suggests and promotes the
improvement of systems for emergency care service provision in the
sites from which participants come. At clinic level, the IMCI approach,
which is specifically designed to manage children based on a severity
assessment rather than diagnosis, was adopted as THE standard of
care for South Africa in the 1990s. However implementation remains
problematic.
Community IMCI is designed to assist caregivers with the recognition
of danger signs and although this programme is aligned with national
policy, it has thus far not been adequately implemented countrywide,
and has made little impact on severity of children’s illness being
adequately identified in the home setting. Actual delays in obtaining
5
Standard Treatment Guidelines and Essential Drugs List for South Africa: Hospital Level
Paediatrics National Department of Health, South Africa, 2006.
128
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
care may be associated with a lot more than danger sign recognition,
such as transport problems, and barriers to access at health facilities
themselves.
Emergency care principles and protocols, regardless of where they are
‘housed’, taught, and then applied, need to be in line with the evidencebased framework and processes suggested by the International Liaison
Committee on Resuscitation.6 Local (South African) expertise and
guidelines for children requiring emergency care exist in abundance,
but perhaps the call now should be for a more unified, standardised
and national response.
Challenges
The country standards for provision of adequate healthcare to children
are clear for home (Community IMCI7), clinics (IMCI), and children’s
wards (EDL, Hospital Paediatrics). However, there are gaps at the
emergency care point in the health care pathway in:
The current skill levels of South African health workers.
National consensus on the most desirable system for training
in and provision of emergency care to children within the
South African context at all levels in the District Health
System.
Resource allocation to children requiring emergency care.
Experts in the provision of emergency care for children based in
medical schools throughout the country need to gather with
paediatricians working in regional hospitals and family medicine
practitioners in district hospitals to begin working out a country
strategy for improving the quality of emergency care children receive in
South Africa.
6
http://www.ilcor.org/en/home/
7
‘Community IMCI’ is accepted as national policy but has not had a country implementation strategy.
Current efforts to standardise the national Community Health Worker Programme with its clear emphasis on
maternal, newborn and child health, present an important opportunity for improving danger sign recognition
and response within communities at country scale, and for better roll-out of ‘Community IMCI’ as the tool.
129
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
For the provision of emergency care to children, the strategy needs to
address the following key components for delivery:
-
Infrastructural needs at all facility levels
Equipment
Staffing
Record keeping
Clinical protocols
Monitoring and evaluation
Good clinical protocols already exist in the different training
programmes on offer, but a country standard should be decided upon.
Child PIP is a tool that can be used for monitoring and evaluation but
a lot of work needs to be done with regard to the other key
components.
There appears to be a crying need for further research into the
problems broadly categorised as failure to recognise severity of illness
and late presentation. Understanding and addressing the actual factors
within these problem categories may then well lead to additional
reductions in child mortality.
Summary and conclusion
Critically ill children are presenting to hospitals with predominantly
three conditions: acute respiratory infections, diarrhoeal disease and
sepsis, on a backdrop of a rampant HIV pandemic.
Problems in the process of caring for critically ill children are spread
throughout the health care pathway. Within the health system itself
lack of application of basic standards of care by health workers and
health resource allocation to critically ill children by administrators are
primary reasons for these problems.
Notwithstanding the prevalence of HIV, managing children properly
with the three predominant conditions is dependent on the basic
elements (A, B, C, D8 etc.) of appropriate resuscitation care.
8
For example Airway, Breathing, Circulation, Dehydration
130
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
In the home and community, there is the ongoing serious problem of
delay in seeking care, which has also been shown by the Perinatal
Problem Identification Programme in the Saving Babies reports9 and
by the Confidential Enquiry into Maternal Deaths in the Saving
Mothers reports,10 and by Child PIP for older children in the previous
Saving Children reports.11
A national plan for improving the quality of emergency care for
critically ill children in South Africa should be devised for all levels of
care within the district health system, and should address capacity
building for health workers, improving systems for provision of
emergency care within health facilities, and emergency care resource
allocation to children.
9
Saving Babies, see www.ppip.co.za
10
11
Saving Mothers, see www.doh.gov.za/docs/reports/2004/savings.pdf
Saving Children, see www.childpip.org.za
131
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
132
O F
D A T A
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
List of Abbreviations and Definitions
Definitions
A&E
Admission and Emergency
ARI
Acute Respiratory Infection
ART
Antiretroviral Treatment
CFR
Case Fatality Rate: Number of deaths in a
specific age group during a specific period
divided by number of admissions in the same
age group and period. Can also be calculated
for specific disease categories.
Case management
Drug treatment and non-drug treatment,
intravenous fluids, feeding, communication
with the caregiver and follow-up.
CHC
Community Health Centre
Child PIP or ChIP
Child Healthcare
Programme
Clinic nurse
Nurse employed by the district, working in a
peripheral clinic
Clinical personnel
Nurses and doctors
DOA
Dead On Arrival
EDL
Essential Drug List
EPI-SA
Expanded Programme
South Africa
133
Problem
on
Identification
Immunisation
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
ETAT
Emergency Triage Assessment and Treatment
FIO
Facility Information Officer
HIV & AIDS
Human Immunodeficiency Virus & Acquired
Immunodeficiency Syndrome
HSRC
Human Sciences Research Council
IHMR
In-Hospital Mortality Rate
IMCI
Integrated Management of Childhood Illness
(a WHO training and implementation
programme for paediatric primary care)
IV fluids
Intravenous fluids
Health worker
Doctors, nurses, paramedical health workers
(P)ICU
(Paediatric) Intensive Care Unit
IMR
Infant Mortality Rate
INP
Integrated Nutrition Programme
KMC
Kangaroo Mother Care
MCWH
Maternal, Child and Women’s Health
MF
Modifiable Factor: Events, actions, omissions
contributing to the death of a child or to
substandard care, in a child who died.
M&M
Morbidity and Mortality meeting
Mortality review
meetings
Regular audit meetings with all health workers
involved, to discuss paediatric deaths that have
occurred in health institutions.
MRC
Medical Research Council
NCCEMD
National Committee for
Enquiries into Maternal Deaths
134
Confidential
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y A E R S
O F
D A T A
NDOH
National Department of Health
NGO
Non-Governmental Organisation
NIMART
Nurse initiation and management of patients
on ART
OPD
Outpatient Department
OWFA
Overweight-For-Age
PCP
Pneumocystis carinii or Pneumocystis jirovecii
pneumonia
PCR
Polymerase Chain Reaction blood test
PHC
Primary Health Care
PITC
Provider Initiated Testing and Counselling
PMTCT
Prevention of Mother-To-Child Transmission
of HIV
QI
Quality Improvement
PPIP
Perinatal Problem Identification Programme
RTHC
Road-to-Health Chart: Under-five chart,
patient-retained record of the child’s weights,
immunizations and health problems.
SAPA
South African Paediatric Association
Severe malnutrition
Marasmus, kwashiorkor
kwashiorkor
U5PIP
Under-five Problem Identification Programme
UWFA
Underweight For Age (below the 3rd centile for
weight-for-age, according to the Wellcome
classification)
WHO
World Health Organisation
135
and
marasmic
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
136
O F
D A T A
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
Appendices
Appendix A:
Data tables for Chapter 1
Appendix B:
Provincial Data
Eastern Cape
Free State
Gauteng
KwaZulu-Natal
Limpopo
Mpumalanga
North West
Northern Cape
Western Cape
Appendix C:
Child PIP Data Capture Sheets & Code Lists
Monthly Tally
Death Data Capture Sheet
Cause of Death
Modifiable Factors
Appendix D:
Additional Tools
Child PIP Mortality Review Process Guideline
137
D A T A
S A V I N G
C H I L D R E N
Appendix A:
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Data Tables for Chapter 1
Table A1. In-hospital mortality rates by age, nutritional status and illness: 2005-2009
All admissions
0-28 days
28 days-1 year
1-5 years
5-13 years
13-18 years
Unknown
Total
Admissions (no.)
Deaths (no.)
IHMR (%)
23127
124440
122940
63179
3271
6451
343408
1284
9832
4544
1830
74
339
17903
5.6
7.9
3.7
2.9
2.3
5.3
5.2
125425
48166
24450
72466
270507
4044
4352
4290
2974
15660
3.2
9.0
17.5
4.1
5.8
76013
67080
127414
270507
5332
3973
6355
15660
7.0
5.9
5.0
5.8
Under-5 admissions
Nutritional status
≥ 3rd centile
< 3rd centile
Severe malnutrition
Unknown
Total
Illness
ARI
Acute diarrhoea
Other
Total
Table A2. Age and Gender of Deaths: 2005-2009
Age
0 - 7 days
8 - 28 days
1 month - 1 year
1 - 5 years
5 - 13 years
13 - 18 years
Unknown
Total
Gender
Female
Male
Unknown
Total
Number
% of all deaths
433
632
11141
4865
1885
72
267
19295
2.2
3.3
57.7
25.2
9.8
0.4
1.4
100.0
Number
% of all deaths
8905
10070
320
19295
46.2
52.2
1.7
100.0
138
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
Table A3. Referral Patterns of Deaths: 2005-2009
Referred
Number
% of all deaths
Yes
No
Unknown
Total
8200
9569
1526
19295
42.5
49.6
7.9
100.0
Number
% of all deaths
2011
4755
1329
9569
1631
19295
10.4
24.6
6.9
49.6
8.5
100.0
Number
% of all deaths
3869
12399
3027
19295
20.1
64.3
15.7
100.0
Referred from
Another hospital
A clinic
Private practitioner
Not referred
Unknown
Total
Re-admission
Yes
No
Unknown
Total
Table A4. Length of Stay and Time of Deaths: 2005-2009
Length of stay
DOA
< 24 hours
1 - 3 days
4 - 7 days
8 - 14 days
> 14 days
Unknown
Total
Time of death
Weekday (07:00 - 19:00)
Weeknight (19:00 - 07:00)
Weekend/Public Holiday
Unknown
Total
Number
% of all deaths
507
6074
4968
3220
2182
2315
29
19295
2.6
31.5
25.7
16.7
11.3
12.0
0.2
100.0
Number
% of all deaths
6881
6595
5520
299
19295
35.7
34.2
28.6
1.5
100.0
139
D A T A
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Table A5. Caregiver Data for Deaths: 2005-2009
Primary Caregiver
Mother
Grandmother
Father
Other
Unknown
Total
Mother’s wellbeing
Alive and Well
Dead
Sick
Unknown
Total
Father’s wellbeing
Alive and Well
Dead
Sick
Unknown
Total
Number
% of all deaths
14062
2097
178
709
2249
19295
72.9
10.9
0.9
3.7
11.7
100.0
Number
% of all deaths
14005
1203
1560
2527
19295
72.6
6.2
8.1
13.1
100.0
Number
% of all deaths
6290
699
316
11990
19295
32.6
3.6
1.6
62.1
100.0
Table A6. Nutritional Status of All Deaths: 2005-2009
Nutritional category
OWFA
Normal
UWFA
Marasmus
Kwashiorkor
Marasmic-Kwash (M-K)
Unknown
Total
Number
% of all deaths
156
5321
5694
4511
1242
870
1501
19295
0.8
27.6
29.5
23.4
6.4
4.5
7.8
100
Table A7. HIV Status of Deaths: 2005-2009
Laboratory category
Negative
Exposed
Infected
No result
Not tested (but indicated)
Not tested (not indicated)
Unknown
Total
Clinical HIV staging
Stage I
Stage II
Stage III
Stage IV
Not staged (but indicated)
Not staged (not indicated)
Unknown
Total
Number
% of all deaths
2722
4414
5432
567
1534
1119
3507
19295
14.1
22.9
28.2
2.9
8.0
5.8
18.2
100.0
Number
% of all deaths
297
519
2472
5525
2519
4042
3921
19295
1.5
2.7
12.8
28.6
13.1
20.9
20.3
100.0
140
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Table A8. Cause of Death Under 5 Years - Main Diagnosis: 2005-2009
Under 5 Years - Main Diagnosis
Pneumonia, ARI
Acute diarrhoea, hypovolaemic shock
Septicaemia, possible serious bacterial infection
PCP (suspected and/or confirmed)
TB: Pulmonary/ Miliary, other extra-pulmonary
Meningitis: bacterial
Chronic diarrhoea
Other Endocr, Nutritional, Metabol. (specify)
Other Respiratory System (specify)
TB: Meningitis
Other diagnosis (specify)
Ill-defined/Unknown cause of mortality
Cirrhosis, Portal Hypertension, Liver Failure, Hepatitis
Hospital-acquired infection
Other Nervous System (specify)
Other possible serious infection (specify)
Other Poisoning (specify)
Heart failure, Pulmonary Oedema
Inhalation of foreign body or gastric content
Status epilepticus
Unknown
Hypoglycaemia
Congenital Heart Disease
Cardiomyopathy
Oncology/Leukaemias/Tumours
Other Circulatory System (specify)
Surgical (Appendix, hernia, intestines, peritoneum)
Acute renal failure
Anaemia
Burns
Meningitis: viral (meningo-encephalitis)
Dysentery
Croup
Other Digestive System (specify)
Myocarditis, Endocarditis
Other inflammatory disease of CNS (e.g. abscess)
Pneumothorax, Pyothorax, Pleural effusion
Non-accidental injury, Abuse-related, Neglect, Homicide, Suicide
Other accidents (incl. Drowning; Paraffin; specify)
Congenital Infections (not HIV)
Chronic renal disease
Cong malformations of the respiratory system
AIDS
Acute nephritis
Bites and Stings, Toxic plants
Other Genito-urinary System (specify)
Transport-related accidents
Measles
IDDM, DKA
Asthma
Malaria
RHD, Rheumatic fever
Total
141
No.
Percent
3553
3103
2933
1656
668
627
605
374
351
289
222
212
187
158
147
143
134
119
111
98
96
72
69
69
64
62
62
54
54
54
50
43
39
37
35
32
27
22
22
21
15
13
11
11
10
9
8
8
7
7
5
5
17071
20.8
18.2
17.2
9.7
3.9
3.7
3.5
2.2
2.1
1.7
1.3
1.2
1.1
0.9
0.9
0.8
0.8
0.7
0.7
0.6
0.6
0.4
0.4
0.4
0.4
0.4
0.4
0.3
0.3
0.3
0.3
0.3
0.2
0.2
0.2
0.2
0.2
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0
0
0
0
0
0
100
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Table A9. Cause of Death: Main and Other Diagnosis (All Ages): 2005-2009
Diagnosis
Main
Other
Total
Percent
Pneumonia, ARI
Acute diarrhoea, hypovolaemic shock
Septicaemia, possible serious bacterial infection
PCP (suspected and/or confirmed)
TB: Pulmonary/Miliary, other extra-pulmonary
Chronic diarrhoea
Other Endocr, Nutritional, Metabol. (specify)
Meningitis: bacterial
Anaemia
Other diagnosis (specify)
Other Respiratory System (specify)
TB: Meningitis
Oncology/Leukaemias/Tumours
Other possible serious infection (specify)
Other Nervous System (specify)
Other Poisoning (specify)
Cirrhosis, Portal Hypertension, Liver Failure, Hepatitis
Acute renal failure
Heart failure, Pulmonary Oedema
Hospital-acquired infection
Hypoglycaemia
Ill-defined/Unknown cause of mortality
Status epilepticus
Other Circulatory System (specify)
Unknown
Inhalation of foreign body or gastric content
Other Digestive System (specify)
Cardiomyopathy
Congenital Heart Disease
Surgical (Appendix, hernia, intestines, peritoneum)
Meningitis: viral (meningo-encephalitis)
Dysentery
Pneumothorax, Pyothorax, Pleural effusion
Other inflammatory disease of CNS (e.g. abscess)
Other Genito-urinary System (specify)
Burns
Non-accidental injury, Abuse-related, Neglect, Homicide etc
Croup
Chronic renal disease
Congenital Infections (not HIV)
Myocarditis
AIDS
Bites and Stings, Toxic plants
Transport-related accidents
Other accidents (incl. Drowning; Paraffin; Corrosives etc)
Acute nephritis
Cong malformations of the respiratory system
RHD, Rheumatic fever
Measles
Asthma
IDDM, DKA
Malaria
Total
3876
3270
3134
1696
727
718
408
761
70
275
403
438
369
191
203
142
214
76
174
171
76
243
126
96
127
117
46
85
78
71
78
44
34
63
14
58
26
42
34
22
35
19
14
26
21
14
14
9
9
9
10
8
19295
1819
1427
1366
605
1204
641
892
214
765
331
192
95
139
262
238
269
122
240
134
122
201
12
127
92
56
50
112
60
63
60
47
57
60
21
67
16
42
19
23
26
12
25
29
2
4
8
6
8
7
6
4
2
12462
5695
4697
4500
2301
1931
1359
1300
975
835
606
595
533
508
453
441
411
336
316
308
293
277
255
253
188
183
167
158
145
141
131
125
101
94
84
81
74
68
61
57
48
47
44
43
28
25
22
20
17
16
15
14
10
31757
29.5
24.3
23.3
11.9
10.0
7.0
6.7
5.1
4.3
3.1
3.1
2.8
2.6
2.3
2.3
2.1
1.7
1.6
1.6
1.5
1.4
1.3
1.3
1.0
0.9
0.9
0.8
0.8
0.7
0.7
0.6
0.5
0.5
0.4
0.4
0.4
0.3
0.3
0.3
0.2
0.2
0.2
0.2
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
100.0
142
CoD
Sepsis
No.
%
No.
%
0 - 28 days
28 days - 1 yr
1 - 5 years
5 - 13 years
13 - 18 years
Unknown
Total
189
4167
853
296
14
53
5572
120
2492
1096
201
4
75
3988
295
1709
929
146
6
49
3134
9
426
522
381
15
15
1368
0.8
3.8
10.7
20.2
20.8
5.6
7.1
17.7
37.4
17.5
15.7
19.4
19.9
28.9
11.3
22.4
22.5
10.7
5.6
28.1
20.7
27.7
15.3
19.1
7.7
8.3
18.4
16.2
TB
Meningitis
No.
%
56
535
375
300
4
7
1277
5.3
4.8
7.7
15.9
5.6
2.6
6.6
HIV exp & inf
No.
%
387
5855
2377
1093
38
96
9846
36.3
52.6
48.9
58.0
52.8
36.0
51.0
Severe maln
No.
%
101
3194
2582
640
27
79
6623
9.5
28.7
53.1
34.0
37.5
29.6
34.3
Total
No.
%
1065
11141
4865
1885
72
267
19295
100.0
100.0
100.0
100.0
100.0
100.0
100.0
F I V E
Table A11. Age distribution of Leading Causes of Death (CoD): 2005-2009
Sepsis
No.
%
No.
%
0 - 28 days
28 days - 1 yr
1 - 5 years
5 - 13 years
13 - 18 years
Unknown
Total
189
4167
853
296
14
53
5572
120
2492
1096
201
4
75
3988
295
1709
929
146
6
49
3134
9
426
522
381
15
15
1368
0.7
31.1
38.2
27.9
1.1
1.1
100.0
3.4
74.8
15.3
503
0.3
1.0
100.0
3.0
62.5
27.5
5.0
0.1
1.9
100.0
9.4
54.5
29.6
4.7
0.2
1.6
100.0
TB
Meningitis
No.
%
56
535
375
300
4
7
1277
4.4
41.9
29.4
23.5
0.3
0.5
100.0
HIV exp & inf
No.
%
387
5855
2377
1093
38
96
9846
3.9
59.5
24.1
11.1
0.4
1.0
100.0
Severe maln
No.
%
101
3194
2582
640
27
79
6623
1.5
48.2
39.0
9.7
0.4
1.2
100.0
Total
No.
%
1065
11141
4865
1885
72
267
19295
5.5
57.7
25.2
9.8
0.4
1.4
100.0
D A T A
Diarrhoea
No.
%
O F
Age
ARIs
No.
%
Y E A R S
143
CoD
2 0 0 9 :
Diarrhoea
No.
%
C H I L D R E N
Age
ARIs
No.
%
S A V I N G
Table A10. Leading Causes of Death (CoD) and Age category: 2005-2009
Table A12. Leading Causes of Death (CoD) and Referral patterns & Re-admissions: 2005-2009
CoD
Referred
ARIs
No.
%
Diarrhoea
No.
%
Sepsis
No.
%
No.
%
Not referred
Referred
Unknown
2763
434
2375
49.6
7.8
42.6
2238
360
1390
56.1
9.0
34.9
1443
200
1491
46.0
6.4
47.6
577
90
701
42.2
6.6
51.2
534
83
660
41.8
6.5
51.7
5160
601
4085
Total
5572
100.0
3988
100.0
3134
100.0
1368
100.0
1277
100.0
9846
Diarrhoea
No.
%
Sepsis
No.
%
No.
%
Yes
No
Unknown
Total
1027
3688
857
5572
764
2611
613
3988
659
2008
467
3134
402
771
195
1368
29.4
56.4
14.3
100.0
18.4
66.2
15.4
100.0
19.2
65.5
15.4
100.0
21.0
64.1
14.9
100.0
TB
Meningitis
No.
%
232
853
192
1277
18.2
66.8
15.0
100.0
HIV exp & inf
No.
%
Severe maln
No.
%
Total
No.
%
52.4
6.1
41.5
3364
415
2844
50.8
6.3
42.9
9569
1526
8200
49.6
7.9
42.5
100.0
6623
100.0
19295
100.0
HIV exp & inf
No.
%
2551
5906
1389
9846
25.9
60.0
14.1
100.0
Severe maln
No.
%
1861
3768
796
6623
28.1
56.9
12.0
100.0
Total
No.
%
3869
12399
3027
19295
20.1
64.3
15.7
100.0
CoD
TB
Meningitis
No.
%
HIV exp & inf
No.
%
%
DOA
< 24 hours
1 - 3 days
4 - 7 days
8 - 14 days
> 14 days
Unknown
106
1596
1528
1047
673
620
2
1.9
28.6
27.4
18.8
12.1
11.1
0.0
152
1612
1024
571
329
295
5
3.8
40.4
25.7
14.3
8.2
7.4
0.1
53
904
814
553
383
426
1
1.7
28.8
26.0
17.6
12.2
13.6
0.0
11
212
318
283
232
311
1
0.8
15.5
23.2
20.7
17.0
22.7
0.1
20
408
338
221
133
156
1
1.6
31.9
26.5
17.3
10.4
12.2
0.1
153
2442
2475
1820
1342
1611
3
Total
5572
100.0
3988
100.0
3134
100.0
1368
100.0
1277
100.0
9846
Total
No.
%
1.6
24.8
25.1
18.5
13.6
16.4
0.0
76
1544
1689
1329
966
1019
0
1.1
23.3
25.5
20.1
14.6
15.4
0.0
507
6074
4968
3220
2182
2315
29
2.6
31.5
25.7
16.7
11.3
12.0
0.2
100.0
6623
100.0
19295
100.0
D A T A
No.
Severe maln
No.
%
O F
Sepsis
No.
%
Y E A R S
Diarrhoea
No.
%
F I V E
LOS
ARIs
No.
%
2 0 0 9 :
Table A13. Leading Causes of Death (CoD) and Length of Stay (LOS): 2005-2009
C H I L D R E N
144
Re-adm
ARIs
No.
%
Meningitis
No.
%
S A V I N G
CoD
TB
CoD
ARIs
TB
No.
%
Meningitis
No.
%
HIV exp & inf
No.
%
Total
No.
%
No.
%
OWFA
Normal
UWFA
Marasmus
Kwashiorkor
M-K
Unknown
44
1894
1857
1110
178
130
359
0.8
34.0
33.3
19.9
3.2
2.3
6.4
21
939
1104
1111
254
197
362
0.5
23.5
27.7
27.9
6.4
4.9
9.1
19
646
816
878
384
286
105
0.6
20.6
26.0
28.0
12.3
9.1
3.4
9
201
409
510
83
88
68
0.7
14.7
29.9
37.3
6.1
6.4
5.0
18
454
400
246
25
19
115
1.4
35.6
31.3
19.3
2.0
1.5
9.0
68
2130
3130
3150
535
469
364
0.7
21.6
31.8
32.0
5.4
4.8
3.7
156
5321
5694
4511
1242
870
1501
0.8
27.6
29.5
23.4
6.4
4.5
7.8
Total
5572
100.0
3988
100.0
3134
100.0
1368
100.0
1277
100.0
9846
100.0
19295
100.0
CoD
Sepsis
No.
%
No.
%
Stage I
Stage II
Stage III
Stage IV
Not staged Not staged Unknown
Total
91
177
873
1766
833
814
1018
5572
64
113
530
852
594
809
1026
3988
50
93
421
1030
394
697
449
3134
5
16
160
757
113
175
142
1368
0.4
1.2
11.7
55.3
8.3
12.8
10.4
100.0
1.6
3.2
15.7
31.7
14.9
14.6
18.3
100.0
- but indicated - not indicated
1.6
2.8
13.3
21.4
14.9
20.3
25.7
100.0
1.6
3.0
13.4
32.9
12.6
22.2
14.3
100.0
TB
Meningitis
No.
%
23
29
121
370
161
315
258
1277
1.8
2.3
9.5
29.0
12.6
24.7
20.2
100.0
HIV exp & inf
No.
%
209
396
2044
4885
1394
376
542
9846
2.1
4.0
20.8
49.6
14.2
3.8
5.5
100.0
Severe maln
No.
%
13
88
833
3300
740
952
697
6623
0.2
1.3
12.6
49.8
11.2
14.4
10.5
100.0
Total
No.
%
297
519
2472
5525
2519
4042
3921
19295
1.5
2.7
12.8
28.6
13.1
20.9
20.3
100.0
D A T A
Diarrhoea
No.
%
O F
HIV Stage
ARIs
No.
%
Y E A R S
145
Table A15. Leading Causes of Death (CoD) and Clinical HIV Staging: 2005-2009
2 0 0 9 :
Nutrition
F I V E
Sepsis
No.
%
C H I L D R E N
Diarrhoea
No.
%
S A V I N G
Table A14. Leading Causes of Death (CoD) and Nutritional Status: 2005-2009
Table A16. Leading Causes of Death (CoD) and HIV Laboratory Status: 2005-2009
CoD
ARIs
Diarrhoea
No.
%
Sepsis
No.
%
TB
No.
%
Meningitis
No.
%
Severe maln
No.
%
Total
No.
%
%
Negative
Exposed
Infected
No result
Not tested Not tested Unknown
480
1729
1563
171
498
241
890
8.6
31.0
28.1
3.1
8.9
4.3
16.0
442
955
934
114
326
233
984
11.1
23.9
23.4
2.9
8.2
5.8
24.7
516
750
963
110
251
163
381
16.5
23.9
30.7
3.5
8.0
5.2
12.2
185
161
748
33
88
22
131
13.5
11.8
54.7
2.4
6.4
1.6
9.6
237
189
385
40
110
80
236
18.6
14.8
30.1
3.1
8.6
6.3
18.5
813
1271
2883
215
603
145
693
12.3
19.2
43.5
3.2
9.1
2.2
10.5
2722
4414
5432
567
1534
1119
3507
14.1
22.9
28.2
2.9
8.0
5.8
18.2
Total
5572
100.0
3988
100.0
3134
100.0
1368
100.0
1277
100.0
6623
100.0
19295
100.0
146
- but indicated - not indicated
Year
2005
2006
2007
2008
2009
%
No.
%
No.
%
No.
%
Negative
Exposed
Infected
No result
Not tested (but indicated)
Not tested (not indicated)
Unknown
133
346
429
34
237
29
329
8.7
22.5
27.9
2.2
15.4
1.9
21.4
298
670
886
79
253
115
570
10.4
23.3
30.9
2.8
8.8
4
19.9
483
922
1239
112
291
232
549
12.6
24.1
32.4
2.9
7.6
6.1
14.3
835
1230
1431
162
408
389
1084
15.1
22.2
25.8
2.9
7.4
7
19.6
973
1246
1447
180
345
354
975
17.6
22.6
26.2
3.3
6.3
6.4
17.7
Total
1537
100
2871
100
3828
100
5539
100
5520
100
D A T A
No.
O F
%
Y E A R S
No.
F I V E
HIV Lab
2 0 0 9 :
Table A17. HIV Laboratory Status by Year of Deaths: 2005-2009
C H I L D R E N
No.
S A V I N G
HIV Lab
1-5 years
No.
%
5-13 years
No.
%
Negative
Exposed
Infected
No result
Not tested Not tested Unknown
136
338
49
27
66
153
296
12.8
31.7
4.6
2.5
6.2
14.4
27.8
1419
3441
2414
367
936
606
1958
12.7
30.9
21.7
3.3
8.4
5.4
17.6
851
537
1840
137
416
263
821
17.5
11.0
37.8
2.8
8.6
5.4
16.9
263
59
1034
29
103
84
313
14.0
3.1
54.9
1.5
5.5
4.5
16.6
22
0
38
0
3
3
6
Total
1065
100.0
11141
100.0
4865
100.0
1885
100.0
72
Unknown
No.
%
No.
%
30.6
0.0
52.8
0.0
4.2
4.2
8.3
31
39
57
7
10
10
113
11.6
14.6
21.3
2.6
3.7
3.7
42.3
2722
4414
5432
567
1534
1119
3507
14.1
22.9
28.2
2.9
8.0
5.8
18.2
100.0
267
100.0
19295
100.0
F I V E
28 days-1 year
No.
%
13-18 years
No.
%
Total
Y E A R S
147
- but indicated - not indicated
Table A19. Perinatal ARV Uptake by Year of Deaths: 2005-2009
2005
2006
2007
2008
2009
O F
Year
2 0 0 9 :
HIV Lab
0-28 days
No.
%
C H I L D R E N
Age
S A V I N G
Table A18. HIV Laboratory Status in Different Age Groups: 2005-2009
Total
No.
%
No.
%
No.
%
No.
%
No.
%
No.
%
Given
Not given
Mother negative
Unknown
Total
121
335
139
942
1537
7.9
21.8
9.0
61.3
100.0
432
466
270
1703
2871
15.0
16.2
9.4
59.3
100.0
668
596
545
2019
3828
17.5
15.6
14.2
52.7
100.0
980
651
915
2993
5539
17.7
11.8
16.5
54.0
100.0
1130
680
1040
2670
5520
20.5
12.3
18.8
48.4
100.0
3331
2728
2909
10327
19295
17.3
14.1
15.1
53.5
100.0
D A T A
Perinatal ARVs
Table A20. Clinical HIV Staging and ART in Deaths: 2005-2009
HIV Stage
Stage II
No.
%
Stage III
No.
%
Stage IV
No.
%
Current
Ever
Never Never Unknown
Total
6
7
82
156
46
297
12
3
225
171
108
519
153
45
1656
301
317
2472
884
131
3692
373
445
5525
2.0
2.4
27.6
52.5
15.5
100.0
2.3
0.6
43.4
32.9
20.8
100.0
6.2
1.8
67.0
12.2
12.8
100.0
16.0
2.4
66.8
6.8
8.1
100.0
Not staged No.
%
52
26
969
636
836
2519
2.1
1.0
38.5
25.2
33.2
100.0
Not staged No.
%
Unknown
No.
%
11
15
92
3424
500
4042
29
14
224
651
3003
3921
0.3
0.4
2.3
84.7
12.4
100.0
0.7
0.4
5.7
16.6
76.6
100.0
Total
No.
%
1147
241
6940
5712
5255
19295
5.9
1.2
36.0
29.6
27.2
100.0
- but indicated - not indicated
CoD
TB
Meningitis
No.
%
HIV exp & inf
No.
%
No.
%
Yes
Not sure
No
Unknown
1436
1956
1319
861
25.8
35.1
23.7
15.5
1350
1470
665
503
33.9
36.9
16.7
12.6
859
1084
754
437
27.4
34.6
24.1
13.9
299
502
394
173
21.9
36.7
28.8
12.6
266
463
371
177
20.8
36.3
29.1
13.9
2587
3296
2681
1282
Total
5572
100.0
3988
100.0
3134
100.0
1368
100.0
1277
100.0
9846
Severe maln
No.
%
Total
No.
%
26.3
33.5
27.2
13.0
1791
2299
1639
894
27.0
34.7
24.7
13.5
5013
6708
4702
2872
26.0
34.8
24.4
14.9
100.0
6623
100.0
19295
100.0
Y E A R S
Sepsis
No.
%
F I V E
Diarrhoea
No.
%
2 0 0 9 :
Avoidable?
ARIs
No.
%
C H I L D R E N
148
Table A21. Avoidable Deaths for Leading Causes of Death (CoD): 2005-2009
S A V I N G
ART
Stage I
No.
%
O F
D A T A
CoD
‘Who’
9191
2145
4777
16113
8288
1637
4535
14460
2.9
57.3
11.3
31.4
100.0
Sepsis
No.
%
No.
%
4051
1118
2996
8165
1762
564
1090
3416
51.6
16.5
31.9
100.0
3.6
49.6
13.7
36.7
100.0
TB
2.6
Meningitis
No.
%
HIV exp & inf
No.
%
1610
518
916
3044
13262
3489
8641
25392
2.5
52.9
17.0
30.1
100.0
2.4
52.2
13.7
34.0
100.0
Severe maln
No.
%
9944
2364
7492
19800
50.2
11.9
37.8
100.0
Total
No.
%
29013
7246
17069
53328
54.4
13.6
32.0
100.0
2.6
3.0
2.8
HIV exp & inf
No.
%
Severe maln
No.
%
Total
No.
%
2 0 0 9 :
MF per death
57.0
13.3
29.6
100.0
Diarrhoea
No.
%
TB
Meningitis
No.
%
No.
%
Ward
A&E
Ref & Transit
Clinic/OPD
Home
4239
4242
193
2506
4933
26.3
26.3
1.2
15.6
30.6
4059
3873
140
1674
4714
28.1
26.8
1.0
11.6
32.6
2138
1442
133
1345
3107
26.2
17.7
1.6
16.5
38.1
914
609
51
693
1149
26.8
17.8
1.5
20.3
33.6
783
688
76
526
971
25.7
22.6
2.5
17.3
31.9
6781
5184
305
4326
8796
26.7
20.4
1.2
17.0
34.6
5048
3603
193
3329
7627
25.5
18.2
1.0
16.8
38.5
14510
12608
791
7691
17728
27.2
23.7
1.5
14.4
33.2
Total
16113
100.0
14460
100.0
8165
100.0
3416
100.0
3044
100.0
25392
100.0
19800
100.0
55328
100.0
D A T A
Sepsis
No.
%
O F
Diarrhoea
No.
%
Y E A R S
149
‘Where’
ARIs
No.
%
F I V E
Table A23. ‘Where’ Modifiable Factors – Place in Leading Causes of Death (CoD): 2005-2009
CoD
C H I L D R E N
Clinical pers.
Administrator
Caregiver
Total
ARIs
No.
%
S A V I N G
Table A22. ‘Who’ Modifiable Factors – Responsible person in Leading Causes of Death (CoD): 2005-2009
Table A24. Category of Modifiable Factors by Leading Causes of Death (CoD): 2005-2009
CoD
Total
Care Seeking & Compliance
Manage
Assess
Monitor
Notes
Growth & Development
Investigation
Other
Buildings/Beds
Staff
History
Examination
Home Treatment
Communication
Disease Prevention
Equipment
Treat
Problem List
Transport
Laboratory
Policy
Consumables
Plan
Classify
Total
3304
1616
1791
1756
1186
715
817
637
640
447
558
452
266
368
314
294
250
192
105
150
97
94
49
15
16113
3017
1666
1479
1723
967
792
798
419
358
360
450
429
411
370
236
212
132
152
109
120
152
50
51
7
14460
1900
1016
832
626
440
637
328
401
297
310
138
134
216
158
131
99
125
116
74
62
57
51
13
4
8165
777
450
341
193
198
178
135
150
119
212
64
60
41
61
107
60
95
70
30
17
26
18
10
4
3416
701
375
307
186
209
82
140
132
129
172
80
79
52
52
70
49
65
56
43
17
24
12
10
2
3044
11433
5946
5343
5281
3599
2870
2606
2293
1934
1748
1512
1328
1282
1221
957
848
708
662
446
431
419
274
155
32
53328
21.4
11.1
10.0
9.9
6.7
5.4
4.9
4.3
3.6
3.3
2.8
2.5
2.4
2.3
1.8
1.6
1.3
1.2
0.8
0.8
0.8
0.5
0.3
0.1
100.0
D A T A
%
O F
No.
Y E A R S
Meningitis
No.
F I V E
TB
No.
2 0 0 9 :
Sepsis
No.
C H I L D R E N
Diarrhoea
No.
S A V I N G
150
MF category
ARIs
No.
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Table A25. Quality of Records: 2005-2009
Records
Number
954
7239
10632
470
19295
Folder not available
Folder available: incomplete and/or inadequate
Folder available: OK
Unknown
Total
% of all deaths
4.9
42.5
55.1
2.4
100
Table A26. ‘Where’ and ‘Who’ of Modifiable factors (MFs): 2005-2009
Where they occur
No.
%*
14510
12608
791
7691
17728
53328
27.2
23.7
1.5
14.4
33.2
100
Who is responsible
No.
Rate †
Clinical personnel
Administrator
Caregiver and family
Total
29013
7246
17069
53328
1.5
0.4
0.9
2.8
Ward
Admission & Emergency care
Referring facility and Transit
Clinic/Outpatients
Home
Total
* Proportion of total MFs
†
Rate, i.e. number of MFs per death
Table A27. Avoidable Deaths: 2005-2009
‘Was this death avoidable?’
Yes
Not sure
No
Unknown
Total
151
Number
% of all deaths
5013
6708
4702
2872
19295
26.0
34.8
24.4
14.9
100.0
S A V I N G
C H I L D R E N
Appendix B:
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Provincial Data
Eastern Cape
Baseline data
Eastern Cape
2005
2006
2007
2008
2009
0
-
1
87
1
0
3
3836
135
3.5
170
254
1.5
3
5640
332
5.9
310
972
3.1
6
8235
384
4.7
339
1440
4.2
2005
2006
2007
2008
2009
-
11.4
56.3
29.9
2.3
100
4.1
60.6
27.6
5.9
1.8
100
3.2
61.6
25.8
8.4
0.3
0.6
100
5.6
56.0
18.0
10.3
0.6
9.4
100
NUTRITION Nutritional category
OWFA
Normal
UWFA
Marasmus
Kwashiorkor
Marasmic-Kwashiorkor
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
-
1.1
11.5
33.3
16.1
4.6
3.4
29.9
100
1.2
23.5
24.7
24.1
9.4
5.3
11.8
100
0.3
24.2
37.4
17.7
6.8
3.2
10.3
100
0.3
25.7
32.7
15.0
5.9
3.5
16.8
100
HIV&AIDS Laboratory category
Negative
Exposed
Infected
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
-
3.4
28.7
17.2
50.7
100
14.1
25.9
30.6
29.4
100
23.2
24.5
19.7
32.6
100
20.6
19.5
21.2
38.7
100
Clinical HIV staging
2005
2006
2007
2008
2009
-
2.3
3.4
5.7
4.6
28.7
55.2
100
2.9
4.7
9.4
17.6
37.6
27.6
100
0.3
2.3
7.1
14.5
51.3
24.5
100
0.6
1.8
5.3
10.9
53.4
28.0
100
2005
2006
2007
2008
2009
-
5.7
1.1
4.6
88.5
100
17.6
8.8
12.9
60.6
100
15.2
7.1
20
57.7
100
12.4
8.6
25.4
53.7
100
No. of sites
Total admissions
Total deaths
In-hospital mortality rate (%)
Audited deaths
Total modifiable factors
Modifiable factor rate (per death)
Information about children who died
Demographics Age
0-28 days
28 days-1 year
1-5 years
5-13 years
13-18 years
Unknown
Total (% of all deaths)
Health context
Stage I
Stage II
Stage III
Stage IV
Not staged (± indicated)
Unknown
Total (% of all deaths)
PMTCT Nevirapine (NVP) prophylaxis
NVP given
NVP not given
Mother negative
Unknown
Total (% of all deaths)
152
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
FEEDING PRACTICE Infant feeding
Exclusive breast
No breast, ever
Mixed
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
-
8.0
16.1
4.6
71.3
100
4.1
27.6
5.9
62.4
100
8.7
28.1
11.0
52.3
100
7.7
18.0
15.6
58.7
100
ART (CHILD ART – child deaths
DEATHS) Current
Ever
Never (but indicated)
Never (not indicated)
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
-
2.3
16.1
31.0
50.6
100
3.5
1.8
30
34.7
30.0
100
3.9
1.9
21.6
43.2
29.4
100
5.0
1.2
18.9
44.0
31.0
100
MOTHER’S Mother’s wellbeing
WELLBEING Alive and well
Sick
Dead
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
-
58.6
2.3
4.6
34.5
100
61.2
5.3
11.8
21.8
100
64.5
3.9
13.5
18.1
100
79.4
2.9
6.2
11.5
100
Causes of child Main cause of death: top 5
deaths Diarrhoeal disease (acute and chronic)
2005-2009: % of all deaths
25.1
21.0
9.9
7.1
5.7
Pneumonia, ARI
Septicaemia
TB (pulmonary, meningitis and miliary)
PCP (suspected and confirmed)
Information about quality of child healthcare
Records
2005
2006
2007
2008
2009
-
67.8
32.2
100
5.9
57.7
35.3
1.2
100
5.2
43.8
49.4
1.6
100
2.4
42.4
54.0
1.2
100
Modifiable factors: where?
2005
2006
2007
2008
2009
Ward
Admission & emergency care
Referring facility and transit
Clinic/OPD
Home
Total (% of MFs)
-
-
13.4
18.5
0.8
12.6
54.7
100
29.7
13
0.7
14.7
41.9
100
29.9
26.5
3.0
13.4
27.2
100
Modifiable factors: who?
2005
2006
2007
2008
2009
Clinical personnel
Administrator
Caregiver and family
Total (MF rate per death)
-
-
0.5
0.1
0.9
1.5
1.0
0.8
1.3
3.1
2.6
0.7
0.9
4.2
Folder not available
Folder incomplete / inadequate
Folder available: OK
Unknown
Total (% of all deaths)
153
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Free State
Baseline data
Free State
2005
2006
2007
2008
2009
No. of sites
Total admissions
Total deaths
In-hospital mortality rate (%)
Audited deaths
Total modifiable factors
Modifiable factor rate (per death)
2
1811
86
4.7
86
223
2.6
2
862
20
2.3
84
385
4.6
3
3356
207
6.2
231
600
2.6
4
3364
233
6.9
268
702
2.6
2
2792
149
5.3
155
502
3.2
2005
2006
2007
2008
2009
1.2
64.0
24.4
9.3
1.2
100
8.3
65.5
16.7
7.1
2.4
100
2.6
56.7
25.1
15.2
0.4
100
4.5
53.0
31.3
10.1
0.7
0.4
100
4.5
60.0
26.5
8.4
0.6
100
2005
2006
2007
2008
2009
1.2
20.9
32.6
33.7
1.2
4.7
5.8
100
31.0
27.4
32.1
2.4
4.8
2.4
100
24.2
36.4
26.0
4.3
3.9
5.2
100
1.5
19.4
30.6
32.5
6.3
7.1
2.6
100
8.4
20.0
37.4
11.0
16.8
6.5
100
Information about children who died
Demographics Age
0-28 days
28 days-1 year
1-5 years
5-13 years
13-18 years
Unknown
Total (% of all deaths)
Health context
NUTRITION Nutritional category
OWFA
Normal
UWFA
Marasmus
Kwashiorkor
Marasmic-Kwashiorkor
Unknown
Total (% of all deaths)
HIV&AIDS Laboratory category
Negative
Exposed
Infected
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
1.2
7.0
36.0
55.8
100
3.6
20.2
31.0
45.2
100
11.3
11.7
45.5
31.5
100
12.3
11.9
43.7
32.1
100
20.0
16.1
40.0
23.9
100
Clinical HIV staging
2005
2006
2007
2008
2009
Stage I
Stage II
Stage III
Stage IV
Not staged (± indicated)
Unknown
Total (% of all deaths)
2.3
4.7
27.9
20.9
7.0
37.2
100
2.4
9.5
25.0
46.4
16.7
100
3.0
2.2
10.0
42.0
30.4
12.6
100
3.4
7.5
40.7
37.3
11.2
100
8.4
40.0
36.2
15.5
100
2005
2006
2007
2008
2009
3.5
25.6
70.9
100
7.1
19.0
3.6
70.2
100
8.7
19.5
13.9
58.0
100
11.2
27.2
20.9
40.7
100
13.5
15.5
25.2
45.8
100
PMTCT Nevirapine (NVP) prophylaxis
NVP given
NVP not given
Mother negative
Unknown
Total (% of all deaths)
154
S A V I N G
FEEDING PRACTICE Infant feeding
Exclusive breast
No breast, ever
Mixed
Unknown
Total (% of all deaths)
ART (CHILD ART – child deaths
DEATHS) Current
Ever
Never (but indicated)
Never (not indicated)
Unknown
Total (% of all deaths)
MOTHER’S Mother’s wellbeing
WELLBEING Alive and well
Sick
Dead
Unknown
Total (% of all deaths)
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
2005
2006
2007
2008
2009
9.3
11.6
29.1
50.0
100
15.5
27.4
21.4
35.7
100
9.1
17.7
31.2
42.0
100
18.3
22.0
29.5
30.2
100
14.8
18.7
32.9
33.5
100
2005
2006
2007
2008
2009
7.0
43.0
50.0
100
13.1
1.2
33.3
4.8
47.6
100
9.5
0.4
43.3
30.7
16.0
100
6.3
0.7
44.0
35.1
13.8
100
13.5
36.1
27.7
22.6
100
2005
2006
2007
2008
2009
31.4
14.0
11.6
43.0
100
58.3
14.3
14.3
13.1
100
68.0
11.7
11.3
9.1
100
70.9
8.2
13.8
7.1
100
50.3
12.3
23.2
14.2
100
Causes of child Main cause of death: top 5
deaths Diarrhoeal disease (acute and chronic)
2005-2009: % of all deaths
22.7
22.5
17.1
13.2
5.5
Septicaemia
Pneumonia, ARI
PCP (suspected and confirmed)
TB (pulmonary, meningitis and miliary)
Information about quality of child healthcare
Records
2005
2006
2007
2008
2009
Folder not available
Folder incomplete / inadequate
Folder available: OK
Unknown
Total (% of all deaths)
34.9
65.1
100
29.8
70.2
100
3.0
35.5
60.6
0.9
100
3.4
22.4
74.3
100
7.7
18.7
72.3
1.3
100
Modifiable factors: where?
2005
2006
2007
2008
2009
Ward
Admission & emergency care
Referring facility and transit
Clinic/OPD
Home
Total (% of MFs)
29.1
43.0
0.4
3.1
24.2
100
20.3
57.9
0.5
2.9
18.4
100
18.3
23.3
0.5
15.2
42.7
100
16.2
9.3
2.3
17.5
54.7
100
29.3
5.6
4.6
17.5
43.0
100
Modifiable factors: who?
2005
2006
2007
2008
2009
Clinical personnel
Administrator
Caregiver and family
Total (MF rate per death)
1.8
0.2
0.7
2.6
3.7
0.1
0.7
4.6
1.4
0.2
1.1
2.6
1.1
0.2
1.3
2.6
1.5
0.5
1.2
3.2
155
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Gauteng
Baseline data
Gauteng
2005
2006
2007
2008
2009
No. of sites
Total admissions
Total deaths
In-hospital mortality rate (%)
Audited deaths
Total modifiable factors
Modifiable factor rate (per death)
2
3169
201
6.3
227
376
1.7
2
3798
193
5.1
235
305
1.3
2
6734
217
3.2
242
343
1.4
2
6424
240
3.7
229
460
2.0
3
7899
325
4.1
343
517
1.5
2005
2006
2007
2008
2009
1.8
64.3
23.3
7.9
2.2
0.4
100
6.0
66.4
19.1
7.7
0.4
0.4
100
6.2
57.4
20.2
14.0
1.7
0.4
100
4.8
65.5
26.6
2.6
0.4
100
5.8
58.0
26.8
7.6
1.5
0.3
100
Information about children who died
Demographics Age
0-28 days
28 days-1 year
1-5 years
5-13 years
13-18 years
Unknown
Total (% of all deaths)
Health context
NUTRITION Nutritional category
OWFA
Normal
UWFA
Marasmus
Kwashiorkor
Marasmic-Kwashiorkor
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
1.3
23.3
35.7
25.6
3.5
1.8
8.8
100
1.7
30.2
31.1
24.7
4.3
5.1
3.0
100
0.8
22.7
35.1
19.8
4.5
4.1
12.8
100
0.4
24.5
31.4
20.1
10.0
9.6
3.9
100
0.6
35.6
28.0
19.0
9.3
4.1
3.5
100
HIV&AIDS Laboratory category
Negative
Exposed
Infected
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
15.9
12.3
48.9
22.9
100
23.0
13.6
46.4
17.0
100
27.7
9.9
50.0
12.4
100
32.8
13.1
44.1
10.0
100
35.0
13.1
38.2
13.7
100
Clinical HIV staging
2005
2006
2007
2008
2009
Stage I
Stage II
Stage III
Stage IV
Not staged (± indicated)
Unknown
Total (% of all deaths)
0.4
4.4
10.1
33.9
44.1
7.0
100
1.3
5.1
10.6
40.4
36.6
6.0
100
1.2
5.4
42.1
43.8
7.4
100
0.9
7.0
38.9
50.2
3.1
100
0.9
2.3
7.0
28.9
50.7
10.2
100
2005
2006
2007
2008
2009
8.8
26.4
20.7
44.1
100
14.5
24.7
24.3
36.6
100
21.1
22.7
30.2
26.0
100
25.3
15.7
38.9
20.1
100
14.3
19.0
35.9
30.9
100
PMTCT Nevirapine (NVP) prophylaxis
NVP given
NVP not given
Mother negative
Unknown
Total (% of all deaths)
156
S A V I N G
FEEDING PRACTICE Infant feeding
Exclusive breast
No breast, ever
Mixed
Unknown
Total (% of all deaths)
ART (CHILD ART – child deaths
DEATHS) Current
Ever
Never (but indicated)
Never (not indicated)
Unknown
Total (% of all deaths)
MOTHER’S Mother’s wellbeing
WELLBEING Alive and well
Sick
Dead
Unknown
Total (% of all deaths)
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
2005
2006
2007
2008
2009
12.8
17.6
22.9
46.7
100
14.0
22.6
26.0
37.4
100
9.1
25.2
31.4
34.3
100
20.5
27.9
33.2
18.3
100
11.4
20.4
29.2
39.1
100
2005
2006
2007
2008
2009
8.4
1.3
41.0
17.2
32.2
100
9.4
3.4
38.7
33.6
14.9
100
9.5
0.8
40.1
41.7
7.9
100
10.5
0.4
33.2
50.2
5.7
100
10.5
1.2
28.6
42.9
16.9
100
2005
2006
2007
2008
2009
72.2
2.6
5.7
19.4
100
84.3
3.0
4.7
8.1
100
74.4
2.9
5.8
16.9
100
84.3
1.3
10.0
4.4
100
74.3
2.3
5.8
17.5
100
Causes of child Main cause of death: top 5
deaths Diarrhoeal disease (acute and chronic)
2005-2009: % of all deaths
18.5
15.5
14.8
12.0
7.1
Septicaemia
Pneumonia, ARI
PCP (suspected and confirmed)
TB (pulmonary, meningitis and miliary)
Information about quality of child healthcare
Records
2005
2006
2007
2008
2009
Folder not available
Folder incomplete / inadequate
Folder available: OK
Unknown
Total (% of all deaths)
20.7
13.3
65.2
0.9
100
5.1
22.6
71.9
0.4
100
15.3
13.2
69.4
2.1
100
3.1
23.1
73.4
0.4
100
12.5
15.8
71.4
0.3
100
Modifiable factors: where?
2005
2006
2007
2008
2009
Ward
Admission & emergency care
Referring facility and transit
Clinic/OPD
Home
Total (% of MFs)
27.1
11.2
0.5
24.7
36.4
100
22.3
4.3
0.7
7.9
64.9
100
25.1
3.8
0.9
16.6
53.6
100
23.7
5.2
4.1
17.0
50.0
100
22.6
8.1
3.9
18.2
47.2
100
Modifiable factors: who?
2005
2006
2007
2008
2009
Clinical personnel
Administrator
Caregiver and family
Total (MF rate per death)
0.8
0.3
0.6
1.7
0.4
0.1
0.9
1.3
0.5
0.1
0.9
1.4
0.8
0.2
1.0
2.0
0.6
0.1
0.7
1.5
157
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
KwaZulu-Natal
Baseline data
KwaZulu-Natal
2005
2006
2007
2008
2009
No. of sites
Total admissions
Total deaths
In-hospital mortality rate (%)
Audited deaths
Total modifiable factors
Modifiable factor rate (per death)
4
6346
565
8.9
483
973
2.0
11
15786
1254
7.9
1470
2527
1.7
17
23189
1556
6.7
1993
3386
1.7
21
27057
1839
6.8
1910
5329
2.8
30
35377
2149
6.1
2103
6681
3.2
2005
2006
2007
2008
2009
3.3
51.6
26.7
15.3
0.2
2.9
100
4.8
55.4
22.4
10.7
0.1
6.6
100
6.0
57.2
23.8
11.6
0.3
1.1
100
5.5
57.8
25.9
9.5
0.2
1.0
100
4.7
56.5
26.9
9.3
0.5
2.0
100
2005
2006
2007
2008
2009
1.7
23.0
25.5
23.8
4.6
4.8
16.8
100
1.3
27.6
26.9
23.8
6.0
3.5
10.8
100
1.4
27.6
30.2
23.0
5.7
3.1
9.0
100
1.3
29.8
26.6
24.1
6.8
3.4
8.1
100
0.8
30.5
30.2
20.7
8.8
4.3
4.6
100
Information about children who died
Demographics Age
0-28 days
28 days-1 year
1-5 years
5-13 years
13-18 years
Unknown
Total (% of all deaths)
Health context
NUTRITION Nutritional category
OWFA
Normal
UWFA
Marasmus
Kwashiorkor
Marasmic-Kwashiorkor
Unknown
Total (% of all deaths)
HIV&AIDS Laboratory category
Negative
Exposed
Infected
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
7.5
25.9
24.0
42.6
100
8.8
29.5
24.6
37.1
100
9.8
30.7
27.3
32.4
100
13.4
29.5
24.7
32.4
100
17.8
27.2
27.5
27.4
100
Clinical HIV staging
2005
2006
2007
2008
2009
Stage I
Stage II
Stage III
Stage IV
Not staged (± indicated)
Unknown
Total (% of all deaths)
2.7
4.3
18.0
18.8
28.3
27.7
100
2.9
3.8
15.9
26.6
25.7
25.1
100
2.2
3.3
15.5
30.4
27.7
21.0
100
2.1
3.6
19.1
25.8
26.1
23.3
100
2.6
3.0
14.9
28.4
34.0
17.1
100
2005
2006
2007
2008
2009
7.0
8.3
6.2
78.5
100
18.2
13.7
7.5
60.7
100
18.6
12.6
11.7
57.0
100
23.5
12.8
15.1
48.6
100
29.4
12.9
19.3
38.4
100
PMTCT Nevirapine (NVP) prophylaxis
NVP given
NVP not given
Mother negative
Unknown
Total (% of all deaths)
158
S A V I N G
FEEDING PRACTICE Infant feeding
Exclusive breast
No breast, ever
Mixed
Unknown
Total (% of all deaths)
ART (CHILD ART – child deaths
DEATHS) Current
Ever
Never (but indicated)
Never (not indicated)
Unknown
Total (% of all deaths)
MOTHER’S Mother’s wellbeing
WELLBEING Alive and well
Sick
Dead
Unknown
Total (% of all deaths)
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
2005
2006
2007
2008
2009
7.5
13.5
15.1
64.0
100
14.6
16.2
14.0
55.2
100
14.6
24.5
14.5
46.4
100
17.5
25.9
13.5
43.0
100
20.5
30.0
14.8
34.7
100
2005
2006
2007
2008
2009
0.6
0.4
30.8
9.9
58.2
100
4.2
1.3
39.4
23.4
31.7
100
7.2
1.0
38.1
28.3
25.3
100
6.2
1.0
35.1
28.2
29.5
100
7.5
2.3
32.5
36.3
21.4
100
2005
2006
2007
2008
2009
60.0
6.0
5.4
28.6
100
63.3
6.4
5.5
24.8
100
75.9
7.5
5.5
11.1
100
75.7
4.6
4.9
14.9
100
80.9
6.0
4.9
8.1
100
Causes of child Main cause of death: top 5
deaths Diarrhoeal disease (acute and chronic)
2005-2009: % of all deaths
23.4
18.6
17.4
6.6
6.3
Pneumonia, ARI
Septicaemia
PCP (suspected and confirmed)
TB (pulmonary, meningitis and miliary)
Information about quality of child healthcare
Records
2005
2006
2007
2008
2009
Folder not available
Folder incomplete / inadequate
Folder available: OK
Unknown
Total (% of all deaths)
11.2
52.6
25.9
10.4
100
7.6
40.9
46.5
5.1
100
4.0
25.1
65.3
5.7
100
4.1
28.0
63.5
4.3
100
3.3
23.5
70.5
2.7
100
Modifiable factors: where?
2005
2006
2007
2008
2009
Ward
Admission & emergency care
Referring facility and transit
Clinic/OPD
Home
Total (% of MFs)
26.1
17.9
1.4
15.2
39.4
100
26.0
20.8
1.6
12.7
38.9
100
24.7
16.5
1.0
13.6
44.3
100
31.1
26.0
0.5
11.0
31.3
100
25.4
25.0
2.2
15.3
32.0
100
Modifiable factors: who?
2005
2006
2007
2008
2009
Clinical personnel
Administrator
Caregiver and family
Total (MF rate per death)
1.2
0.2
0.6
2.0
0.8
0.2
0.6
1.7
0.8
0.2
0.7
1.7
1.6
0.4
0.8
2.8
1.8
0.4
0.9
3.2
159
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Limpopo
Baseline data
Limpopo
2005
2006
2007
2008
2009
No. of sites
Total admissions
Total deaths
In-hospital mortality rate (%)
Audited deaths
Total modifiable factors
Modifiable factor rate (per death)
1
864
42
4.9
41
136
3.3
3
1971
123
6.2
110
297
2.7
1
955
48
5
48
118
2.5
3
1730
98
5.7
85
459
5.4
5
2585
137
5.3
144
698
4.8
2005
2006
2007
2008
2009
61.0
39.0
100
60.9
29.1
10.0
100
4.2
60.4
29.2
6.3
100
16.4
49.4
31.8
2.4
100
9.0
60.4
19.4
11.1
100
2005
2006
2007
2008
2009
12.2
46.3
36.6
2.4
2.4
100
0.9
16.4
25.5
42.7
3.6
6.4
4.5
100
4.2
18.8
16.7
45.8
4.2
6.3
4.2
100
21.2
25.9
40.0
10.6
2.4
100
0.7
24.3
27.1
20.8
6.9
10.4
9.7
100
2005
2006
2007
2008
2009
2.4
39.0
22.0
36.6
100
4.5
23.6
33.6
38.1
100
39.6
35.4
25.1
100
10.6
29.4
28.2
31.7
100
16.0
27.1
30.6
26.5
100
Clinical HIV staging
2005
2006
2007
2008
2009
Stage I
Stage II
Stage III
Stage IV
Not staged (± indicated)
Unknown
Total (% of all deaths)
14.6
7.3
41.5
24.4
9.8
2.4
100
0.9
3.6
4.5
49.1
27.3
14.5
100
6.3
8.3
35.4
45.8
4.2
100
1.2
4.7
32.9
56.5
4.7
100
1.4
5.6
25.0
50.7
17.4
100
2005
2006
2007
2008
2009
12.2
22.0
65.9
100
8.2
14.5
2.7
74.5
100
27.1
16.7
56.3
100
24.7
17.6
11.8
45.9
100
20.1
29.9
15.3
34.7
100
Information about children who died
Demographics Age
0-28 days
28 days-1 year
1-5 years
5-13 years
13-18 years
Unknown
Total (% of all deaths)
Health context
NUTRITION Nutritional category
OWFA
Normal
UWFA
Marasmus
Kwashiorkor
Marasmic-Kwashiorkor
Unknown
Total (% of all deaths)
HIV&AIDS Laboratory category
Negative
Exposed
Infected
Unknown
Total (% of all deaths)
PMTCT Nevirapine (NVP) prophylaxis
NVP given
NVP not given
Mother negative
Unknown
Total (% of all deaths)
160
S A V I N G
FEEDING PRACTICE Infant feeding
Exclusive breast
No breast, ever
Mixed
Unknown
Total (% of all deaths)
ART (CHILD ART – child deaths
DEATHS) Current
Ever
Never (but indicated)
Never (not indicated)
Unknown
Total (% of all deaths)
MOTHER’S Mother’s wellbeing
WELLBEING Alive and well
Sick
Dead
Unknown
Total (% of all deaths)
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
2005
2006
2007
2008
2009
17.1
36.6
19.5
26.8
100
18.2
31.8
20.0
30.0
100
39.6
27.1
2.1
31.3
100
28.2
31.8
22.4
17.6
100
25.7
16.7
26.4
31.3
100
2005
2006
2007
2008
2009
2.4
51.2
46.3
100
3.6
1.8
49.1
13.6
31.8
100
2.1
2.1
68.8
4.2
22.9
100
3.5
42.4
49.4
4.7
100
7.6
0.7
34.0
41.0
16.7
100
2005
2006
2007
2008
2009
22.0
4.9
34.1
39.0
100
50.0
6.4
30.0
13.6
100
47.9
31.3
20.8
100
80.0
4.7
11.8
3.5
100
80.6
4.9
9.0
5.6
100
Causes of child Main cause of death: top 5
deaths Pneumonia, ARI
2005-2009: % of all deaths
29.4
18.0
10.7
10.3
8.4
Diarrhoeal disease (acute and chronic)
Septicaemia
PCP (suspected and confirmed)
TB (pulmonary, meningitis and miliary)
Information about quality of child healthcare
Records
2005
2006
2007
2008
2009
Folder not available
Folder incomplete / inadequate
Folder available: OK
Unknown
Total (% of all deaths)
65.9
31.7
2.4
100
49.9
50.0
100
60.4
39.6
100
1.2
74.0
24.7
100
1.4
54.9
43.1
0.7
100
Modifiable factors: where?
2005
2006
2007
2008
2009
Ward
Admission & emergency care
Referring facility and transit
Clinic/OPD
Home
Total (% of MFs)
26.5
30.1
14.7
28.7
100
14.8
27.3
12.5
45.5
100
25.4
27.1
6.8
40.7
100
34.4
20.9
1.7
10.0
32.9
100
36.4
20.2
2.6
6.7
34.1
100
Modifiable factors: who?
2005
2006
2007
2008
2009
Clinical personnel
Administrator
Caregiver and family
Total (MF rate per death)
1.9
0.3
1.1
3.3
1.2
0.3
1.2
2.7
1.1
0.3
1.0
2.5
2.5
1.1
1.8
5.4
2.1
0.9
1.8
4.8
161
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Mpumalanga
Baseline data
Mpumalanga
2005
2006
2007
2008
2009
No. of sites
Total admissions
Total deaths
In-hospital mortality rate (%)
Audited deaths
Total modifiable factors
Modifiable factor rate (per death)
2
2349
125
5.3
126
170
1.3
2
3598
168
4.7
167
282
1.7
8
7751
463
6.0
480
1254
2.6
24
20787
1502
7.2
1508
6210
4.1
26
25572
1635
6.4
1675
7582
4.5
2005
2006
2007
2008
2009
3.2
61.9
19.8
9.5
5.6
100
6.0
55.7
25.1
10.8
2.4
100
6.3
64.2
21.0
8.3
0.2
100
7.0
58.7
25.3
9.0
0.1
100
11.7
55.8
22.0
10.5
0.1
100
2005
2006
2007
2008
2009
2.4
26.2
24.6
28.6
2.4
4.8
11.1
100
1.2
24.6
29.3
32.9
1.8
8.4
1.8
100
0.4
32.1
31.3
19
5.0
4.2
8.1
100
0.3
31.1
28.4
17.8
6.2
3.6
12.5
100
0.3
30.9
28.2
20.1
6.1
3.4
10.9
100
Information about children who died
Demographics Age
0-28 days
28 days-1 year
1-5 years
5-13 years
13-18 years
Unknown
Total (% of all deaths)
Health context
NUTRITION Nutritional category
OWFA
Normal
UWFA
Marasmus
Kwashiorkor
Marasmic-Kwashiorkor
Unknown
Total (% of all deaths)
HIV&AIDS Laboratory category
Negative
Exposed
Infected
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
6.3
39.7
24.6
29.4
100
5.4
26.3
51.5
16.8
100
9.0
23.5
34.8
32.7
100
5.8
21.0
19.6
53.6
100
8.5
22.7
21.8
47.0
100
Clinical HIV staging
2005
2006
2007
2008
2009
Stage I
Stage II
Stage III
Stage IV
Not staged (± indicated)
Unknown
Total (% of all deaths)
0.8
4.0
18.3
47.6
20.7
8.7
100
3.6
8.4
15.0
40.1
31.2
1.8
100
1.7
3.8
18.1
23.8
38.7
14.0
100
0.8
1.7
11.1
21.1
37.2
28.2
100
0.5
1.6
9.3
19.2
37.6
31.9
100
2005
2006
2007
2008
2009
2.4
57.9
6.3
33.3
100
10.8
45.5
6.0
37.7
100
14.8
27.9
12.1
45.2
100
8.4
8.2
6.8
76.6
100
12.1
9.6
9.2
69.1
100
PMTCT Nevirapine (NVP) prophylaxis
NVP given
NVP not given
Mother negative
Unknown
Total (% of all deaths)
162
S A V I N G
FEEDING PRACTICE Infant feeding
Exclusive breast
No breast, ever
Mixed
Unknown
Total (% of all deaths)
ART (CHILD ART – child deaths
DEATHS) Current
Ever
Never (but indicated)
Never (not indicated)
Unknown
Total (% of all deaths)
MOTHER’S Mother’s wellbeing
WELLBEING Alive and well
Sick
Dead
Unknown
Total (% of all deaths)
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
2005
2006
2007
2008
2009
21.4
14.3
32.5
31.7
100
8.4
18.0
47.3
26.3
100
16.3
20.2
26.3
37.3
100
11.5
16.5
16.0
56.0
100
13.4
17.2
16.8
52.6
100
2005
2006
2007
2008
2009
2.4
0.8
68.3
5.6
23.0
100
6.6
0.6
57.5
24.0
11.4
100
5.0
1.9
31.0
43.3
18.8
100
3.1
1.3
22.9
30.0
42.7
100
5.6
1.6
37.0
22.2
33.7
100
2005
2006
2007
2008
2009
51.6
11.1
15.9
21.4
100
64.7
13.2
12.6
9.6
100
75.6
7.7
10.4
6.3
100
80.6
4.7
5.0
9.7
100
85.5
5.2
5.0
4.3
100
Causes of child Main cause of death: top 5
deaths Pneumonia, ARI
2005-2009: % of all deaths
28.1
22.5
9.1
8.7
5.6
Diarrhoeal disease (acute and chronic)
PCP (suspected and confirmed)
Septicaemia
TB (pulmonary, meningitis and miliary)
Information about quality of child healthcare
Records
2005
2006
2007
2008
2009
Folder not available
Folder incomplete / inadequate
Folder available: OK
Unknown
Total (% of all deaths)
9.5
31.7
57.9
0.8
100
39.5
60.5
100
2.7
53.6
43.3
0.4
100
4.2
69.4
25.2
1.1
100
7.3
60.4
32.2
0.1
100
Modifiable factors: where?
2005
2006
2007
2008
2009
Ward
Admission & emergency care
Referring facility and transit
Clinic/OPD
Home
Total (% of MFs)
20.0
21.2
0.6
9.4
48.8
100
26.6
13.5
0.7
11.3
47.9
100
22.6
26.6
1.6
9.6
39.6
100
25.6
32.5
0.3
8.4
33.2
100
25.1
30.9
1.4
13.7
29.0
100
Modifiable factors: who?
2005
2006
2007
2008
2009
Clinical personnel
Administrator
Caregiver and family
Total (MF rate per death)
0.5
0.3
0.5
1.3
0.7
0.1
0.8
1.7
1.4
0.2
1.0
2.6
2.5
0.2
1.3
4.1
2.8
0.4
1.3
4.5
163
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
North West
Baseline data
North West
2005
2006
2007
2008
2009
No. of sites
Total admissions
Total deaths
In-hospital mortality rate (%)
Audited deaths
Total modifiable factors
Modifiable factor rate (per death)
5
5334
386
7.2
381
1526
4.0
5
5921
466
7.9
438
1121
2.6
5
5327
340
6.4
341
1298
3.8
6
8697
619
7.1
602
1663
2.8
8
7195
359
5.0
363
1120
3.1
2005
2006
2007
2008
2009
0.5
65.4
24.4
9.4
0.3
100
1.2
58.4
29.9
10
0.5
100
5.5
65.1
22.9
6.5
100
4.1
64.3
23.4
8.0
0.2
100
3.9
54.3
28.4
10.2
0.3
3.0
100
2005
2006
2007
2008
2009
22.8
31.0
31.0
5.5
6.6
3.1
100
19.4
26.5
38.6
4.6
3.9
7.1
100
1.2
21.1
32.8
33.1
5.9
3.2
2.6
100
0.2
17.8
30.4
34.2
5.6
8.8
3.0
100
0.8
12.7
23.1
39.4
10.2
11.0
2.8
100
Information about children who died
Demographics Age
0-28 days
28 days-1 year
1-5 years
5-13 years
13-18 years
Unknown
Total (% of all deaths)
Health context
NUTRITION Nutritional category
OWFA
Normal
UWFA
Marasmus
Kwashiorkor
Marasmic-Kwashiorkor
Unknown
Total (% of all deaths)
HIV&AIDS Laboratory category
Negative
Exposed
Infected
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
4.2
24.4
16.3
55.2
100
5.9
15.5
37.9
40.7
100
10.6
16.4
36.1
37.0
100
15.0
19.6
32.2
33.3
100
16.5
20.4
28.7
34.5
100
Clinical HIV staging
2005
2006
2007
2008
2009
Stage I
Stage II
Stage III
Stage IV
Not staged (± indicated)
Unknown
Total (% of all deaths)
1.8
6.3
23.4
44.9
4.5
19.2
100
1.1
2.5
10.5
61.4
4.8
19.6
100
0.6
19.9
44.6
18.8
16.1
100
0.5
1.2
13.8
40.9
31.3
12.5
100
0.3
0.3
9.9
40.2
33.9
15.4
100
2005
2006
2007
2008
2009
7.3
22.0
4.5
66.1
100
11.2
10.0
5.0
73.7
100
15.5
7.3
7.9
69.2
100
20.9
10.8
11.6
56.6
100
23.7
12.1
10.5
53.7
100
PMTCT Nevirapine (NVP) prophylaxis
NVP given
NVP not given
Mother negative
Unknown
Total (% of all deaths)
164
S A V I N G
FEEDING PRACTICE Infant feeding
Exclusive breast
No breast, ever
Mixed
Unknown
Total (% of all deaths)
ART (CHILD ART – child deaths
DEATHS) Current
Ever
Never (but indicated)
Never (not indicated)
Unknown
Total (% of all deaths)
MOTHER’S Mother’s wellbeing
WELLBEING Alive and well
Sick
Dead
Unknown
Total (% of all deaths)
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
2005
2006
2007
2008
2009
15.7
9.7
20.7
53.8
100
17.8
7.8
11.2
63.2
100
12.9
20.5
10.3
56.3
100
11.1
32.4
13.8
42.7
100
12.9
33.6
16.8
36.6
100
2005
2006
2007
2008
2009
0.3
0.5
67.7
3.1
28.3
100
3.9
57.3
6.8
32.0
100
1.5
0.9
67.4
12.0
18.2
100
6.0
1.2
49.0
22.6
21.3
100
7.7
1.9
38.8
28.7
22.9
100
2005
2006
2007
2008
2009
56.4
9.4
19.9
14.2
100
36.3
13.0
17.6
33.1
100
36.4
14.4
24.0
25.2
100
53.7
11.6
14.1
20.6
100
56.5
10.2
19.6
13.8
100
Causes of child Main cause of death: top 5
deaths Septicaemia
2005-2009: % of all deaths
18.7
16.3
14.4
13.5
11.8
TB (pulmonary, meningitis and miliary)
Pneumonia, ARI
Diarrhoeal disease (acute and chronic)
PCP (suspected and confirmed)
Information about quality of child healthcare
Records
2005
2006
2007
2008
2009
Folder not available
Folder incomplete / inadequate
Folder available: OK
Unknown
Total (% of all deaths)
1.6
57.2
41.2
100
10.3
53.2
36.5
100
1.8
45.2
53.1
100
3.3
38.0
57.6
1.0
100
4.1
34.2
60.3
1.4
100
Modifiable factors: where?
2005
2006
2007
2008
2009
Ward
Admission & emergency care
Referring facility and transit
Clinic/OPD
Home
Total (% of MFs)
30.6
37.3
0.6
20.7
10.8
100
43.8
24.4
1.1
22.8
7.9
100
49.6
21.7
0.9
20.3
7.5
100
47.3
16.0
1.8
19.9
15.0
100
30.7
19.8
4.6
26.5
18.3
100
Modifiable factors: who?
2005
2006
2007
2008
2009
Clinical personnel
Administrator
Caregiver and family
Total (MF rate per death)
2.6
1.1
0.3
4.0
1.6
0.8
0.2
2.6
1.9
1.6
0.3
3.8
1.4
0.9
0.4
2.8
2.0
0.5
0.6
3.1
165
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Northern Cape
Baseline data
Northern Capet
2005
2006
2007
2008
2009
No. of sites
Total admissions
Total deaths
In-hospital mortality rate (%)
Audited deaths
Total modifiable factors
Modifiable factor rate (per death)
1
3251
132
4.1
166
323
1.9
1
2751
97
3.5
194
376
1.9
1
4528
137
3.0
165
454
2.8
2
4966
173
3.5
215
507
2.4
3
6912
198
2.9
200
475
2.4
2005
2006
2007
2008
2009
1.2
57.2
32.5
8.4
0.6
100
1.5
50.0
35.1
12.9
0.5
100
0.6
50.9
32.7
15.2
0.6
100
2.3
55.3
32.6
8.8
0.9
100
3.0
52.0
33.0
9.5
1.0
1.5
100
Information about children who died
Demographics Age
0-28 days
28 days-1 year
1-5 years
5-13 years
13-18 years
Unknown
Total (% of all deaths)
Health context
NUTRITION Nutritional category
OWFA
Normal
UWFA
Marasmus
Kwashiorkor
Marasmic-Kwashiorkor
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
27.1
33.1
27.1
4.2
5.4
3.0
27.1
100
30.4
40.2
16.5
3.6
7.7
1.5
30.4
100
38.8
27.9
21.2
3.6
6.7
1.8
38.8
100
27.0
34.4
20.9
6.5
8.8
2.3
27.0
100
23.5
37.0
18.0
7.5
10.0
4.0
23.5
100
HIV&AIDS Laboratory category
Negative
Exposed
Infected
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
17.5
13.3
38.6
30.7
100
20.1
8.2
32.0
39.8
100
27.3
10.3
35.8
26.6
100
34.0
8.4
36.7
20.9
100
37.0
11.5
29.5
22.0
100
Clinical HIV staging
2005
2006
2007
2008
2009
Stage I
Stage II
Stage III
Stage IV
Not staged (± indicated)
Unknown
Total (% of all deaths)
0.6
1.8
9.0
38.6
37.9
12.0
100
0.5
1.0
5.7
33.5
45.3
13.9
100
3.6
0.6
4.2
33.3
49.7
8.5
100
1.4
0.5
2.8
33.0
51.6
10.7
100
0.5
1.0
6.5
27.0
51.0
14.0
100
2005
2006
2007
2008
2009
14.5
25.3
18.1
42.2
100
10.8
22.7
13.4
53.1
100
16.4
21.2
32.1
30.3
100
22.3
20.5
43.3
14.0
100
22.0
15.5
42.0
20.5
100
PMTCT Nevirapine (NVP) prophylaxis
NVP given
NVP not given
Mother negative
Unknown
Total (% of all deaths)
166
S A V I N G
FEEDING PRACTICE Infant feeding
Exclusive breast
No breast, ever
Mixed
Unknown
Total (% of all deaths)
ART (CHILD ART – child deaths
DEATHS) Current
Ever
Never (but indicated)
Never (not indicated)
Unknown
Total (% of all deaths)
MOTHER’S Mother’s wellbeing
WELLBEING Alive and well
Sick
Dead
Unknown
Total (% of all deaths)
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
2005
2006
2007
2008
2009
13.3
9.6
12.7
64.5
100
14.9
5.7
18.0
61.3
100
24.8
6.1
25.5
43.6
100
33.0
14.4
26.5
26.0
100
24.5
10.5
30.5
34.5
100
2005
2006
2007
2008
2009
7.8
0.6
45.2
9.6
36.7
100
5.7
32.5
46.9
14.9
100
7.9
0.6
33.9
48.5
9.1
100
10.2
0.9
28.4
48.8
11.6
100
9.0
1.0
24.0
50.0
16.0
100
2005
2006
2007
2008
2009
81.3
4.8
7.2
6.6
100
77.8
7.2
9.3
5.7
100
77.0
5.5
10.9
6.7
100
82.3
2.8
10.2
4.7
100
83.5
3.5
7.0
6.0
100
Causes of child Main cause of death: top 5
deaths Septicaemia
2005-2009: % of all deaths
33.7
16.9
9.8
8.5
6.8
Pneumonia, ARI
PCP (suspected and confirmed)
Diarrhoeal disease (acute and chronic)
TB (pulmonary, meningitis and miliary)
Information about quality of child healthcare
Records
2005
2006
2007
2008
2009
Folder not available
Folder incomplete / inadequate
Folder available: OK
Unknown
Total (% of all deaths)
3.6
25.9
69.9
0.6
100
2.1
21.6
72.2
4.1
100
3.0
18.1
78.2
0.6
100
6.0
10.7
83.3
100
3.5
10.5
86.0
100
Modifiable factors: where?
2005
2006
2007
2008
2009
Ward
Admission & emergency care
Referring facility and transit
Clinic/OPD
Home
Total (% of MFs)
20.7
12.4
1.2
25.4
40.2
100
9.6
4.5
1.6
21.5
62.8
100
4.6
3.5
0.4
28.4
63.0
100
14.0
7.7
0.8
22.3
55.2
100
18.9
8.0
6.7
22.1
44.2
100
Modifiable factors: who?
2005
2006
2007
2008
2009
Clinical personnel
Administrator
Caregiver and family
Total (MF rate per death)
1.0
0.2
0.8
1.9
0.7
0.1
1.2
1.9
0.9
0.2
1.7
2.8
0.9
0.2
1.2
2.4
1.2
0.2
0.9
2.4
167
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Western Cape
Baseline data
Western Cape
2005
2006
2007
2008
2009
No. of sites
Total admissions
Total deaths
In-hospital mortality rate (%)
Audited deaths
Total modifiable factors
Modifiable factor rate (per death)
2
529
6
1.1
27
60
2.2
4
5978
72
1.2
86
268
3.1
11
7702
87
1.1
158
309
2.0
11
28195
343
1.2
412
472
1.1
15
12285
62
0.5
198
256
1.3
2005
2006
2007
2008
2009
7.4
51.9
29.6
7.4
3.7
100
2.3
57.0
31.4
5.8
1.2
2.3
100
4.4
68.4
22.8
3.8
0.6
100
9.7
48.8
30.1
10.0
1.5
100
1.5
59.6
25.8
11.6
1.5
100
2005
2006
2007
2008
2009
44.4
40.7
100
26.7
32.6
24.4
7.0
4.7
4.7
100
33.5
31.0
20.3
12.7
0.6
1.9
100
1.9
42.7
34.2
10.7
5.6
1.5
3.4
100
2.0
37.4
30.8
9.6
12.6
5.1
2.5
100
Information about children who died
Demographics Age
0-28 days
28 days-1 year
1-5 years
5-13 years
13-18 years
Unknown
Total (% of all deaths)
Health context
NUTRITION Nutritional category
OWFA
Normal
UWFA
Marasmus
Kwashiorkor
Marasmic-Kwashiorkor
Unknown
Total (% of all deaths)
14.8
HIV&AIDS Laboratory category
Negative
Exposed
Infected
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
22.2
22.2
18.5
37.0
100
34.9
9.3
26.7
29.1
100
29.7
7.0
32.3
31.0
100
33.7
12.6
21.1
32.5
100
39.4
10.6
16.2
33.8
100
Clinical HIV staging
2005
2006
2007
2008
2009
0.5
4.0
10.6
70.2
14.6
100
Stage I
Stage II
Stage III
Stage IV
Not staged (± indicated)
Unknown
Total (% of all deaths)
PMTCT Nevirapine (NVP) prophylaxis
NVP given
NVP not given
Mother negative
Unknown
Total (% of all deaths)
25.9
7.4
59.2
7.4
100
3.5
5.8
17.4
68.6
4.7
100
5.1
8.9
20.3
57.0
8.9
100
0.7
0.5
3.2
18.0
49.5
28.2
100
2005
2006
2007
2008
2009
14.8
18.5
25.9
40.7
100
26.7
11.6
40.7
20.9
100
20.3
17.1
29.7
32.9
100
18.0
6.8
35.0
40.3
100
18.7
6.6
44.9
29.8
100
168
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
2005
2006
2007
2008
2009
3.7
7.4
33.3
55.6
100
7.0
17.4
36.0
39.5
100
10.8
17.7
31.0
40.5
100
9.5
16.0
17.2
57.3
100
11.6
19.7
21.7
47.0
100
ART (CHILD ART – child deaths
DEATHS) Current
Ever
Never (but indicated)
Never (not indicated)
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
3.7
7.6
1.3
24.7
50.0
16.5
100
8.3
1.2
13.1
51.5
26.0
100
6.6
18.5
70.4
7.4
100
2.3
3.5
26.7
55.8
11.6
100
10.6
61.6
21.2
100
MOTHER’S Mother’s wellbeing
WELLBEING Alive and well
Sick
Dead
Unknown
Total (% of all deaths)
2005
2006
2007
2008
2009
96.3
90.7
2.3
2.3
4.7
100
81.0
2.5
7.0
9.5
100
87.6
1.9
6.6
3.9
100
90.4
1.5
4.0
4.0
100
FEEDING PRACTICE Infant feeding
Exclusive breast
No breast, ever
Mixed
Unknown
Total (% of all deaths)
3.7
100
Causes of child Main cause of death: top 5
deaths Septicaemia
2005-2009: % of all deaths
20.0
19.3
16.2
5.0
4.2
Pneumonia, ARI
Diarrhoeal disease (acute and chronic)
PCP (suspected and confirmed)
TB (pulmonary, meningitis and miliary)
Information about quality of child healthcare
Records
2005
2006
2007
2008
2009
Folder not available
Folder incomplete / inadequate
Folder available: OK
Unknown
Total (% of all deaths)
3.7
22.2
74.1
1.2
32.5
66.3
100
1.5
20.1
73.1
5.3
100
1.5
14.6
83.8
100
0.6
20.9
77.8
0.6
100
Modifiable factors: where?
2005
2006
2007
2008
2009
Ward
Admission & emergency care
Referring facility and transit
Clinic/OPD
Home
Total (% of MFs)
28.3
16.7
8.3
8.3
38.3
100
16.8
14.2
1.9
19.8
47.4
100
18.4
12.3
1.6
24.6
43.0
100
20.1
13.6
2.3
20.3
43.6
100
22.7
10.5
7.4
14.8
44.5
100
Modifiable factors: who?
2005
2006
2007
2008
2009
Clinical personnel
Administrator
Caregiver and family
Total (MF rate per death)
1.0
0.4
0.8
2.2
1.4
0.3
1.4
3.1
0.9
0.2
0.9
2.0
0.5
0.2
0.5
1.1
0.5
0.2
0.5
1.3
169
100
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
170
O F
D A T A
S A V I N G
C H I L D R E N
Appendix C:
Child PIP Data Capture Sheets
Monthly Tally
Death Data Capture Sheet
Child PIP Code Lists
Causes of Death
Modifiable Factors
171
2 0 0 9 :
F I V E
Y E A R S
O F
D A T A
Child Healthcare Problem Identification Programme
Child PIP
Monthly Tally Sheet
Child PIP v3.0.2
Saving lives
through death auditing
Hospital: ________________
Year: _______________
Ward: __________________
Month: ______________
Admissions1
Deaths2
IHMRs4,5
0 - < 28 days
≥ 28 days - < 1 yr
≥ 1 yr - < 5 yrs
≥ 5 yrs - < 13 yrs
Age
≥ 13 yrs - 18 yrs
Unknown
Totals
Complete information below for children < 5 years only
Admissions1
Deaths2
IHMRs4,5
Above or on 3rd centile
UWFA
Weight
Severe malnutrition3
Unknown
Totals (< 5 years)
Acute respiratory infections (ARI)
Diarrhoeal disease (DD)
Illness
Other
Totals (< 5 years)
Notes:
1.
2.
3.
4.
5.
6.
Include all children admitted to your institution’s paediatric/paediatric surgical/children’s service
If age categories are not known, enter zero in age boxes and the total number admissions/deaths in ‘Unknown’
Severe malnutrition includes marasmus, marasmic-kwashiorkor and kwashiorkor
Only enter one diagnosis per admission (choose the most appropriate if more than one applicable)
Shaded areas will be automatically calculated by the computer
deaths
In-hospital mortality rates can be manually calculated: IHMR =
X 100
admissions
Compiled by: _______________________ (Print name)
Date: ________________
___________________________(Sign)
Fax/Tel number: ____________________________
Child Healthcare Problem Identification Programme
Child__________________
PIP
Hospital:
Ward: _________________
Child Death Data Capture Sheet
Deaths Register No: _________
Entered on computer: _____
Child PIP v3.0.2
Saving lives
through death auditing
Confidential document
Patient name:
DoB
Residential
Subdistrict:
Folder no:
Age
yyyy-mm-dd
Date of Admission
pc
auto
Time of Admission
yyyy-mm-dd
Date of Death
Ⓜ/Ⓕ/Ⓤ
Gender
When death occurred
hh : mm
Time of Death
yyyy-mm-dd
Dead on
arrival
Ⓨ/Ⓝ/Ⓤ
Re-admission
Weekday
(07:00-19:00)
hh : mm
Ⓨ/Ⓝ/Ⓤ
Weekend / Public holiday
Weeknight
(19:00-07:00)
Unknown
Records (include RTHC assessment)
2. Folder present, records
incomplete
1. Folder not available
3. Folder present, notes inadequate (quality
of notes is poor)
4. Folder present, records
incomplete AND notes inadequate
5. Folder available,
records & notes OK
Referred
Name of referring
hospital/clinic:
Ⓨ/Ⓝ/Ⓤ
If yes, from:
1. Another hospital
2. A clinic
3. Private
practitioner
If yes, from:
1. Inside drainage area
2. Outside drainage area
Unknown
Unknown
Social
Mother
1. Alive and well
2. Dead
3. Sick
Unknown
Father
1. Alive and well
2. Dead
3. Sick
Unknown
4. Marasmus
5. Kwashiorkor
Primary
caregiver
1. Mother
2. Grandmother
3. Father
4. Other: ___________
Unknown
Nutrition (tick one category box, then fill in actual weight)
1. OWFA
2. Normal
3. UWFA
6. M-K
Weight
Unknown
________ kg
HIV & AIDS (enter status at time of admission, not at time of audit: this is NOT a post-mortem assessment)
Lab
Clinical
1. Negative
2. Exposed
3. Infected
4. No result
5. Not tested
(but indicated)
6. Not tested
(not indicated)
Unknown
1. Stage I
2. Stage II
3. Stage III
4. Stage IV
5. Not staged
(but indicated)
6. Not staged
(not indicated)
Unknown
Perinatal ARV
1. Prophylaxis given
2. Prophylaxis not given
3. Mother negative at delivery
Unknown
Feeding in
first 6 months
1. Exclusive breast for 6/12
2. No breast, ever
3. Mixed, from birth
Unknown
Cotrimoxazole
1. Current
2. Ever
3. Never (but indicated)
4. Never (not indicated)
Unknown
ART (child)
1. Current
2. Ever
3. Never (but indicated)
4. Never (not indicated)
Unknown
ART (mother)
1. Current
2. Ever
3. Never (but indicated)
4. Never (not indicated)
Unknown
Cause of Death (insert codes)
Main cause of death:
Underlying condition (if any):
Other important diagnoses (max 4):
Modifiable Factors (insert codes)
Code
Code
Ward: Hospital
Comments
Code
Referring Facility & Transit
Probable
Possible
Probable
Possible
Probable
Possible
Probable
Possible
Probable
Possible
Probable
Possible
Probable
Possible
Probable
Possible
Code
Admissions & Emergency: Hospital Comments
Probable
Possible
Code
Clinic/Outpatients
Probable
Possible
Probable
Possible
Home
Comments
Comments
Comments
Probable
Possible
Probable
Possible
Probable
Possible
Probable
Possible
Probable
Possible
Probable
Possible
In your opinion, had the process of caring been different, would this death have been avoidable?
1. Yes
2. Not sure
3. No
Unknown
Case Summary/Comments (write summary at time of death, if possible)
Child’s Details (age, weight, where from, admission date/time)
History of Presenting Complaint
Relevant Background History (including details of HIV and TB)
Examination
Problem List
Problem
1.
2.
3.
4.
5.
Comments
Investigations
Progress
Outcome
Child Healthcare Problem Identification Programme
Child PIP
Causes of Death
Child PIP v3.0.2
Saving lives
through death auditing
Please note: The nutritional categories and the clinical and laboratory classifications concerning HIV do not appear here. They have
to be captured in the relevant fields on the data sheet.
Category
Infections and Parasitic Diseases
Oncology, Haematology
Endocrine, Nutritional, Metabolic
Nervous System
Circulatory System
Respiratory System
Causes of Death
Code
Acute diarrhoea, hypovolaemic shock
101
Chronic diarrhoea
102
Dysentery
103
TB: Pulmonary
110
TB: Meningitis
111
TB: Miliary, other extra-pulmonary
112
Septicaemia, possible serious bacterial infection
120
Congenital Infections (not HIV)
130
Meningitis: Bacterial
140
Meningitis: Viral (meningo-encephalitis)
141
Other inflammatory disease of CNS (e.g. abscess)
142
Measles
150
Other possible serious infection (specify)
151
Malaria
170
Hospital-acquired infection
180
Leukaemias
201
Tumours
204
Anaemia
202
Other Oncology/Haematology (specify)
203
IDDM, DKA
301
Hypoglycaemia
304
Other Endocrine, Nutritional, Metabolic (specify)
305
Status epilepticus
401
Other Nervous System (specify)
402
RHD, Rheumatic fever
501
Heart failure, Pulmonary oedema
502
Myocarditis
503
Cardiomyopathy
504
Congenital Heart Disease
507
Endocarditis
505
Other Circulatory System (specify)
506
Croup
601
Pneumonia, ARI
602
PCP (suspected)
603
PCP (confirmed)
608
Pneumothorax, Pyothorax, Pleural effusion
604
Asthma
605
Congenital malformations of the respiratory system
606
Other Respiratory System (specify)
607
Category
Digestive System
Genito-urinary System
Ill-defined/Unknown Cause
Other Diagnosis
Burns
Poisoning
Bites and Stings, Toxic plants
Inhalation/Aspiration
Accidents
Non-accidental injury, Abuse
Homicide
Suicide
Causes of Death
Code
Cirrhosis, Portal Hypertension, Liver Failure, Hepatitis
701
Surgical (appendix, hernia, intestines, peritoneum)
702
Other Digestive System (specify)
703
Acute nephritis
801
Acute renal failure
802
Chronic renal disease
803
Other Genito-urinary System (specify)
804
Ill-defined/Unknown causes of mortality
900
Other diagnosis (specify)
901
Burns
1000
Paraffin
1101
Corrosives
1102
Other Poisoning (specify)
1103
Bites and stings, Toxic plants
1200
Inhalation of foreign body or gastric contents
1300
Transport-related accidents
1400
Other accidents (incl. Drowning; specify)
1500
Non-accidental injury, Abuse-related, Neglect
1600
Homicide
1700
Suicide
1800
Underlying Conditions
Code
Cerebral palsy
1
Hydrocephalus
2
Birth defect (preconception = chromosomal/genetic, or post conception e.g. foetal alcohol syndrome)
3
Ex-low birth weight/preterm infant
4
Twin/Multiple pregnancy
5
Other Underlying Condition (specify)
10
Child PIP
Child Healthcare Problem Identification Programme
Modifiable Factors
Saving lives
through death auditing
Child PIP v3.0.2
Categorising Modifiable Factors
Who is responsible
Clinical Personnel
Where they occur
Ward
Emergency &
Admission
Administrators
Clinical Methods
Assessment
Management
Monitoring
Clinical Methods
Assessment
Management
Monitoring
Infrastructure
Staff
Consumables
Infrastructure
Staff
Consumables
Referring Facility &
Transit
Pre-transit care in referring
facility
In-transit care
Pre-transit care in referring
facility
In-transit care
Clinic and
Outpatient Care
Clinical Methods
Assessment
Management
Monitoring
Infrastructure
Staff
Consumables
Home
Promotion
Prevention
Social support
Transport
Community development
Family/Caregiver
Growth & development
Disease prevention
Home treatment
Care seeking & compliance
Growth & development
Disease prevention
Home treatment
Care seeking & compliance
Growth & development
Disease prevention
Home treatment
Care seeking & compliance
Growth & development
Disease prevention
Home treatment
Care seeking & compliance
Growth & development
Disease prevention
Home treatment
Care seeking & compliance
Notes on codes:
The coding format has changed to enable more efficient data sorting and analysis. The modifiable factors are grouped against
priority conditions and activities as described in IMCI, ETAT, and the WHO Pocketbook on Hospital Care for Children.
PLEASE note that although the categories of modifiable factors are presented beginning at home, the order of the code list has
been reversed, starting with the ward, to avoid undue caregiver-blaming.
•
•
•
•
•
First letter applies to “place”: W = Ward; E (admission and Emergency) = the place of entry to the hospital and
admission; T = Transit (Referring facility and ambulance); C = Clinic and Outpatient care (i.e. ambulatory); H = Home
Second letter applies to the “who”: P = clinical Personnel; A = administrator; F = Family/ caregiver
Third letter or digit applies to the priority condition or activity. R = Records; M = Clinical Methods; O = Other and has
“COMMENT” fields (If you use this code, you should add the modifiable factor in the ‘Comment’ box); X =Cross-cutting;
1 = Danger Signs; 2 = ARI; 3 = DD; 4 = Fever; 5 = Nutrition; 6 = PSBI (Sepsis); 7 = HIV; 8 = TB; 9 = Immunisation
Please note: a Referral hospital is a higher level hospital you refer your patients to; and a Referring hospital is a lower
level hospital where your patients come from
Acronyms used in the code list include:
A&E
APLS
ARI
BLS
CDG
CSG
DD
ETAT
FCG
HIV
Admissions and Emergency
Advanced Paediatric Life Support
Acute Respiratory Infection
Basic Life Support
Child Dependency Grant
Child Support Grant
Diarrhoeal Disease
Emergency Triage, Assessment and Treatment
Foster Care Grant
Human Immunodeficiency Virus
1
IV
LP
NGT
PSBI
OPD
RTHC
SIRS
TB
WHO
Intravenous
Lumbar puncture
Nasogastric tube
Possible Serious Bacterial Infection
Outpatients Department
Road to Health Chart
Systemic Inflammatory Response Syndrome
Tuberculosis
World Health Organisation
Child PIP
Child Healthcare Problem Identification Programme
Modifiable Factors
Saving lives
through death auditing
Child PIP v3.0.2
Ward
Priority
Condition or
Activity
Category
v3 Code
Modifiable factor
Ward - Clinical Personnel
Records
Clinical
methods
Emergency &
Priority
Conditions
ARI
Notes
WPR01
Insufficient notes on clinical care in ward (assess, manage, monitor)
Problem List
WPM01
Inadequate case assessment and management at previous admission to ward
History
WPM02
Inadequate history taken in ward
Examination
WPM03
Inadequate physical examination in ward
WPM04
Inadequate investigations in ward
Investigation
WPM05
Results of investigations inadequately documented (including x-rays) in ward
Problem List
WPM06
Inadequate daily 'Problem List' in ward
Plan
WPM07
Inadequate daily 'Care Plan' in ward
Assess
WP101
New danger signs inadequately identified while in ward
Manage
WP102
Inadequate response to new danger signs
Monitor
WP103
Danger signs missed due to inadequate monitoring in ward
Assess
WP201
Inadequate review of severe ARI in ward
Manage
WP202
Inadequate oxygen therapy in ward
WP203
Inadequate response to non-responding ARI/pneumonia
WP204
Inadequate monitoring of respiratory rate and/or oxygen saturation in ward
Assess
WP301
Manage
WP302
WP303
Inadequate review of child with severe dehydration
Inadequate revision of fluid management plan, despite child's changing condition
in ward
Inadequate blood chemistry review in child with shock and/or dehydration
WP304
Inadequate monitoring of shocked child in ward
WP401
Inadequate assessment of fitting and/or comatose child in ward
WP402
Convulsions not managed according to standardised protocol in ward
WP403
Inadequate referral to higher level of care for child with coma, from ward
WP404
Inadequate monitoring of blood glucose in ward
WP405
Children's coma score and/or 'neuro-obs' not done in ward
WP406
Inadequate monitoring of convulsions in ward
WP501
Child not categorised as having severe malnutrition in ward
WP502
Too much/too little, incorrect type of IV fluids given in ward
WP503
Inadequate prescription for IV fluids in ward
WP504
NGT feedings not prescribed when indicated in ward
WP505
WHO '10 Steps' not followed for child with severe malnutrition
WP506
Prescribed NGT feeds not given in ward
Monitor
DD
Monitor
Assess
Convulsions/
Coma
Manage
Monitor
Assess
Nutrition/
Intake
Manage
WP507
Prescribed feeds not given in ward
WP508
Inadequate NGT feeding technique causing problems, e.g. cough, cyanosis
WP509
Inadequate monitoring of IV fluids and/or drip sites
WP510
WP602
Inadequate intake-output charting in ward
Possible serious bacterial infection (including nosocomial infection) not considered
in ward
Inadequate 'septic workup' in ward
Manage
WP603
Inadequate antibiotics prescribed in ward
Monitor
WP604
Child with 'septic shock'/SIRS inadequately monitored in ward
Monitor
Assess
PSBI
WP601
2
Child PIP
Child Healthcare Problem Identification Programme
Modifiable Factors
Saving lives
through death auditing
Priority
Condition or
Activity
Child PIP v3.0.2
Category
Inadequate HIV assessment in ward
WP702
HIV-infected but not adequately screened for TB
WP703
Inadequate HIV review (testing and staging) in ward
Manage
WP704
Delayed initiation of ART in ward
Monitor
WP705
HIV results (including serology, PCR and viral load) not obtained in ward
WP801
Inadequate TB assessment in ward
WP802
TB-infected child, but not screened for HIV
HIV
Assess
TB
Manage
Monitor
Manage
Communication
Cross-cutting
Monitor
Manage
Other
Modifiable factor
WP701
Assess
Immunisation
v3 Code
Other
WP803
Inadequate TB review in ward
WP804
Inadequate initiation of TB treatment
WP805
Incorrect TB regimen prescribed in ward
WP806
TB not notified
WP901
Immunisations not brought up to date in ward
WPX01
Doctor not called for critically ill child in ward
WPX02
No hand-over of critically ill child in ward
WPX03
Doctor called, but did not respond and/or did not come
WPX06
Junior doctor did not call more senior doctor to ward
WPX07
Doctor at peripheral hospital did not call referral hospital
WPX08
Inadequate advice from higher level facility
WPX09
Unable to contact responsible doctor at higher level facility
WPX10
Critical clinical information inadequately communicated between ward staff
WPX11
Ward staff inadequately communicated with caregiver
WPX12
Critically ill child not reviewed by doctor during weekend/public holiday in ward
WPX04
Essential prescribed treatment not given in ward
WPX05
No team decision for terminal care
WPX13
Appropriate blood product not prescribed
WPX14
Essential treatment not prescribed
WPO
Other clinical personnel modifiable factor in ward (COMMENT)
Ward - Administrator
Records
Notes
Consumables
Emergency &
Priority
Conditions
ARI
Nutrition/
Intake
PSBI
Equipment
WAR01
Inadequate record keeping system for children in ward
WAR02
Inadequate notes on administrator problems in ward
WA101
Inadequate blood product supply to ward
WA102
No functioning pulse oxymeter in ward
WA103
Inadequate oxygen supply to ward
WA104
Inadequate suction in ward
WA105
Inadequate resuscitation area and/or trolley in ward
Buildings/Beds
WA106
Lack of High Care and/or ICU facilities for children in own and higher level facility
Equipment
WA201
Inadequate oxygen delivery equipment in ward
Care Seeking &
Compliance
WA501
Inadequate supply of food/milk to ward
Consumables
WA601
Inadequate antibiotic supply to ward
3
Child PIP
Child Healthcare Problem Identification Programme
Modifiable Factors
Saving lives
through death auditing
Priority
Condition or
Activity
Child PIP v3.0.2
Category
Staff
Cross-cutting
Consumables
Modifiable factor
WAX01
Inadequate number of doctors assigned to children's ward
WAX02
Doctors in children's ward inadequately supervised
WAX03
Inadequate number of nurses assigned to children's ward
WAX04
Inadequate supervision of nurses in children's ward
WAX05
Lack of professional nurse in children's ward 24 hours a day
WAX06
Lack of experienced doctors (post Community Service), for children's ward
WAX08
Inadequate ward stock of essential consumables
WAX09
Inadequate hospital stock of essential consumables
Laboratory
WAX10
Basic laboratory investigations not available to ward 24 hours a day
Buildings/Beds
WAX11
Lack of hospital beds and/or ward overcrowded
WAX12
Lack of standardised case management protocols in ward
WAX13
No policy or system for weekend and/or public holiday ward rounds
Policy
Other
v3 Code
Other
WAO
Other administrator modifiable factor in ward (COMMENT)
Ward - Family or Caregiver
Records
Danger signs
HIV
Other
Notes
WFR01
RTHC information not present in child's folder
Care Seeking &
Compliance
WFR02
Child's 'patient held records' left at home
WFR03
Previous folder number and/or discharge letter not available
WF101
Caregiver declined consent for life-saving intervention in ward
WF701
Caregiver declined HIV test in ward
Care Seeking &
Compliance
Care Seeking &
Compliance
Other
WFO
Other caregiver modifiable factor in ward (COMMENT)
4
Child PIP
Child Healthcare Problem Identification Programme
Modifiable Factors
Saving lives
through death auditing
Child PIP v3.0.2
Admissions and Emergency Care
Priority
Condition or
Activity
Category
v3 Code
Modifiable factor
Admissions and Emergency Care - Clinical Personnel
Records
Clinical
methods
Emergency &
Priority
Conditions
Airway
Breathing
EPR01
Inadequate notes on clinical care (assessment, management, monitoring at A&E)
EPR02
Admission records incomplete or inappropriate
History
EPM01
Inadequate history taken at A&E
Examination
EPM02
Inadequate physical examination at A&E
EPM03
Inadequate investigations (blood, x-ray, other) at A&E
EPM04
Results of urgent investigations not obtained at A&E
Problem List
EPM05
Inadequate problem list compiled at A&E
Plan
EPM06
Inadequate emergency care plan in A&E
EP101
Emergency signs not recognised at A&E
EP102
Priority signs not recognised at A&E
EP103
Not classified as critically ill despite presence of danger signs at A&E
Notes
Investigation
Assess
Manage
EP104
Child not triaged at A&E (spent time in a queue)
Assess
EPA01
Airway obstruction not recognised or correctly classified at A&E
Manage
EPA02
Inadequate management of airway obstruction in A&E
Monitor
EPA03
Critical airway not monitored at A&E
Assess
EPB01
Respiratory rate not taken, respiratory distress not noticed in A&E
Manage
EPB02
Correct oxygen therapy not prescribed or not given at A&E
EPB03
Oxygen saturation not monitored at A&E
Monitor
Circulation
DD
Convulsions/
Coma
EPB04
Respiratory rate not monitored at A&E
Assess
EPC01
Inadequate assessment of shock at A&E
Manage
EPC02
Inadequate treatment for shock in A&E (fluid type, amount, rate; intraosseus line)
Monitor
EPC03
Shock not monitored while awaiting admission, at A&E
Assess
EP301
Inadequate assessment of dehydration at A&E
Manage
EP302
Inadequate rehydration plan at A&E
Monitor
EP303
Hydration not reviewed at A&E
Assess
EP401
Convulsions not recognised at A&E
Manage
EP402
Convulsions not managed according to accepted protocol at A&E
Monitor
EP403
Convulsions not monitored at A&E
Assess
EP411
Inadequate assessment of level of consciousness at A&E
Manage
EP412
Inadequate management of child with depressed LOC at A&E
EP413
Blood glucose not monitored in child with danger signs at A&E
EP414
Level of consciousness not monitored at A&E
Assess
EP601
Possible serious bacterial infection not considered at A&E
Monitor
PSBI
HIV
Cross-cutting
Other
Manage
EP602
Appropriate antibiotics not prescribed at A&E
Investigation
EP603
Important cultures (blood, CSF, urine) not sent at A&E
Monitor
EP604
LP result not obtained at A&E
Assess
EP701
Inadequate HIV assessment at A&E
Communication
Other
EPX01
Inadequate communication by staff to caregiver at A&E
EPX02
Doctor not called for critically ill child at A&E
EPX03
Doctor called for A&E, but did not respond or did not come
EPX04
No hand-over of critically ill child from admitting doctor to ward doctor at A&E
EPO
Other clinical personnel modifiable factor at A&E (COMMENT)
5
Child PIP
Child Healthcare Problem Identification Programme
Modifiable Factors
Saving lives
through death auditing
Priority
Condition or
Activity
Child PIP v3.0.2
Category
v3 Code
Modifiable factor
Admissions and Emergency Care - Administrator
Records
EAR01
Notes
Policy
Emergency &
Priority
Conditions
Staff
Inadequate notes on administrator problems at A&E
EAR02
Inadequate record keeping system for A&E
EA101
Barriers to entry to A&E service
EA102
Insufficient professional nurses allocated to A&E
EA103
No A&E staff trained in ETAT/BLS/APLS
EA104
Inadequate blood product supply at A&E
EA105
Inadequate emergency drugs at A&E
Buildings/Beds
EA106
Inadequate paediatric resuscitation area in casualty/OPD
Equipment
EAA01
Inadequate suction capability in A&E
Equipment
EAB01
No pulse oxymeter at A&E
Consumables
EAB02
Inadequate oxygen supply and/or equipment at A&E
Consumables
EAC01
Inadequate IV fluid supply at A&E
Dehydration
Equipment
EA301
No mechanical intravenous flow controller available at A&E
Convulsions/
Coma
Consumables
EA401
Intravenous phenobarbitone not available at A&E
Equipment
EA402
No children's coma score sheet available at A&E
PSBI
Consumables
EA601
Inadequate antibiotic supply at A&E
Staff
EAX01
Lack of experienced doctors at A&E
Transport
EAX04
Inadequate transport from home to A&E
Laboratory
EAX05
24 hour emergency laboratory investigations not available to A&E
EAX06
Lack of ward beds, delaying movement out of Emergency Room
EAX07
Lack of beds in the resuscitation area/Emergency Room in A&E
EAX08
Lack of Intensive and High Care beds in own, or higher level hospital
EAX09
No formal, structured triage system for A&E
EAX10
Lack of standardised case management protocols at A&E
EAO
Other administrator modifiable factor at A&E (COMMENT)
Airway
Breathing
Circulation
Cross-cutting
Consumables
Buildings/Beds
Policy
Other
Other
Admissions and Emergency Care - Family or Caregiver
EF101
Did not arrive at A&E on day of referral
EF102
Declined consent for life saving procedure in A&E
Care Seeking &
Compliance
EF701
Caregiver declined HIV test in A&E
Records
Notes
EFR01
Caregiver did not bring RTHC and/or referral letter to A&E
Cross-cutting
Care Seeking &
Compliance
EFX01
Primary caregiver not present at A&E
EFX02
Accompanying caregiver knew little about the child at A&E
Other
Other
EFO
Danger signs
Care Seeking &
Compliance
HIV
Other caregiver modifiable factor at A&E (COMMENT)
6
Child PIP
Child Healthcare Problem Identification Programme
Modifiable Factors
Saving lives
through death auditing
Child PIP v3.0.2
Referring Facility and Transit Care
Priority
Condition or
Activity
Category
v3 Code
Modifiable factor
Referring Facility and Transit Care - Clinical Personnel
TPR01
Inadequate notes on transit care
TPR02
Inadequate referral letter from referring facility
TPP01
Severity of child’s condition incorrectly assessed at referring facility
TPP02
Child not re-assessed at time of departure from referring facility
TPP03
Emergency or priority care not provided at referring hospital
TPP04
Referring pathway and/or procedure not followed by referring facility
TPP05
TPP07
No or delayed referral to higher level
Child with life-threatening condition not monitored at referring facility while
awaiting ambulance
No plan for transporting caregiver to receiving facility
TPP08
Inappropriate care or late referral from private sector/GP
TPI01
Child not assessed properly by ambulance crew at time of entry into ambulance
TPI02
Major complications (e.g. blocked or dislodged ETT) in ambulance not identified
Manage
TPI03
Child not managed correctly in ambulance
Monitor
TPI04
Child not monitored correctly in ambulance
Notes
Assess
Referring
facility
Manage
TPP06
Monitor
Manage
Assess
In-transit care
Other
TPO
Other
Other clinical personnel modifiable factor in transit care (COMMENT)
Referring Facility and Transit Care - Administrator
Records
Notes
TAR01
Pre-transit care
Staff
TAP01
Referring
facility
Staff
TAP02
Consumables
TAP03
Equipment
TAP04
Buildings/Beds
TAP05
Policy
TAP06
Referring
facility
Transport
TAI01
Inadequate record keeping system for proper transit care
No nurse assigned responsibility for monitoring the child while awaiting the
ambulance
No doctor assigned responsibility for monitoring the child while awaiting the
ambulance
Inadequate critical care consumables (e.g. volume expander, ETT, ICD) in
referring facility
Inadequate monitoring and critical care equipment (e.g. ventilator) in referring
facility
No high care bed in referring facility for pre-transfer care of critically ill child
Referral pathways and/or procedures not clear to referring and/or receiving
facility
Inadequate ambulance service from health facility to receiving hospital
TAI02
No ambulance available for transfer from referring to receiving hospital
TAI03
Delayed arrival of ambulance at referring facility
TAI06
No or inappropriate grade of ambulance (i.e. vehicle) available
Staff
TAI04
Consumables
TAI05
Equipment
TAI07
Grade of ambulance crew (i.e. personnel) inappropriate for child’s condition
Inadequate in-transit consumables (e.g. volume expanders, dextrose,
anticonvulsant) in ambulance
Inadequate monitoring and critical care equipment in ambulance
Policy
TAI08
No known/available policy on child transfers in ambulance service (EMRS)
Other
Other
TAO
Danger signs
Care Seeking &
Compliance
TF101
Other
Other
TFO
In-transit care
Other administrator modifiable factor in transit care (COMMENT)
Referring Facility and Transit Care - Family or Caregiver
Caregiver not available to accompany child on transfer
Other caregiver modifiable factor in transit care (COMMENT)
7
Child PIP
Child Healthcare Problem Identification Programme
Modifiable Factors
Saving lives
through death auditing
Child PIP v3.0.2
Clinic and Outpatient Care
Priority
Condition or
Activity
Category
v3 Code
Modifiable factor
Clinic and Outpatient Care - Clinical Personnel
Records
Notes
Assess
Clinical
methods
Classify
Treat
Assess
Danger signs
ARI
DD
CPR02
RTHC inadequately documents child's health history
CPR03
Inadequate referral letter from clinic to hospital
CPM01
IMCI not used for patient assessment at clinic/OPD
CPM02
Incorrect IMCI assessment at clinic/OPD
CPM03
Insufficient assessment for chronic illness at clinic/OPD
CPM04
Insufficient investigations done at clinic/OPD
CPM05
Inadequate IMCI classification at clinic/OPD
CPM07
IMCI not used for case management at clinic/OPD
CPM08
IMCI not used to guide patient referral from clinic/OPD
CP101
Danger signs missed at clinic/OPD
Inadequate response to danger signs at clinic/OPD
CP103
Delayed referral of child with danger signs, from clinic/OPD
Monitor
CP104
Child with danger signs not monitored at clinic/OPD
Assess
CP201
Inadequate assessment for ARI at clinic/OPD
CP202
Oxygen not prescribed or given at clinic/OPD
CP203
Bronchodilator not given to child with wheeze, in clinic/OPD
Manage
CP204
Delayed referral for ARI from clinic/OPD
Monitor
CP205
Oxygen saturation not monitored at clinic/OPD
Assess
CP301
Severity of dehydration incorrectly assessed at clinic/OPD
CP302
Inadequate fluid management for diarrhoeal disease with dehydration
CP303
Delayed referral for child with severe dehydration from clinic/OPD
CP304
Delay in referring chronic diarrhoea from clinic/OPD
CP305
Inadequate review of child with dehydration at clinic/OPD
CP401
Meningitis not considered in child with fever at clinic/OPD
CP402
Malaria not considered in child with fever at clinic/OPD
Treat
Manage
Assess
Fever
Treat
CP403
Antipyretic measures not taken at clinic/OPD
CP404
Appropriate anti-meningitis treatment not initiated at clinic/OPD
CP405
Appropriate anti-malarial treatment not initiated at clinic/OPD
CP501
Monitor
CP505
Growth not plotted correctly on RTHC
Child's growth problem (severe malnutrition, not growing well) inadequately
identified or classified
Inadequate response to growth faltering or failure, at clinic/OPD
Delayed referral for severe malnutrition, weight loss, or growth faltering from
clinic/OPD
No follow up for child's nutritional problem at clinic/OPD
Assess
CP601
Possible serious bacterial infection (PSBI) not considered at clinic/OPD
Manage
CP602
Appropriate antibiotics not given for PSBI at clinic/OPD
CP701
No documentation of mother's antenatal HIV status
CP702
Inadequate assessment for HIV (IMCI not used) at clinic/OPD
CP703
No clear documentation of child's HIV status at clinic/OPD
CP704
Inadequate response to HIV classification at clinic/OPD
CP705
Not referred for ART from clinic/OPD, though indicated
CP706
Referred for ART but ART not initiated
CP707
HIV result not obtained/documented at clinic/OPD
Assess
PSBI
Inadequate notes on clinical care (assess, classify, treat) at clinic
CP102
Treat
Monitor
Nutrition/
Intake
CPR01
CP502
CP503
Manage
Assess
HIV
Manage
Monitor
CP504
8
Child PIP
Child Healthcare Problem Identification Programme
Modifiable Factors
Saving lives
through death auditing
Priority
Condition or
Activity
TB
Immunisation
Child PIP v3.0.2
Category
v3 Code
Assess
CP801
Inadequate assessment for household TB contact at clinic/OPD
CP802
No response to/delayed referral of chronic cough (>2 weeks) at clinic/OPD
CP803
INH prophylaxis not initiated in child with household TB contact
CP804
Child had household TB contact, but no contact tracing was done
CP901
Missed vaccines despite clinic/OPD attendance
CPX01
Communication problems: Staff to caregiver
Manage
Disease
Prevention
Communication
Modifiable factor
CPX02
Staff to staff communication problem at clinic or between clinic & hospital
Assess
CPX03
Insufficient assessment for non-IMCI condition at clinic/OPD
Manage
CPX04
Delay in referring other acute problem from clinic/OPD
Other
Other
CPO
Records
Notes
CAR01
Inadequate record keeping system in clinic
Clinical
methods
Policy
CAM01
Lack of standardised case management protocols in clinic/OPD
CAM02
Inadequate IMCI implementation at clinic/OPD
CA101
No transport from home to clinic
CA102
No emergency transport from clinic to hospital
CA103
No pulse oxymeter at clinic/OPD
CA104
No suction at clinic/OPD
Cross-cutting
Other clinical personnel modifiable factor at clinic/OPD (COMMENT)
Clinic and Outpatient Care - Administrator
Transport
Equipment
Danger signs
Buildings/Beds
Policy
Laboratory
ARI
Consumables
Equipment
CA105
Lack of high care beds and/or resuscitation area in clinic/OPD
CA106
No clinic within reach of child's home or limited opening times
CA107
No policy on short-stay for paediatric patients at clinic/OPD
CA108
Barriers to entry to clinic/OPD
CA109
Basic laboratory investigation not available (e.g. blood glucose)
CA201
No oxygen or oxygen delivery system at clinic/OPD
CA202
No bronchodilators at clinic/OPD
CA203
No spacer or nebuliser for bronchodilators at clinic/OPD
CA301
No oral rehydration solution at clinic/OPD
CA302
Inadequate intravenous sets or solutions at clinic/OPD (incl. intra-osseous)
Inadequate antibiotics at clinic (as per IMCI/EDL)
DD
Consumables
PSBI
Consumables
CA601
Consumables
CA701
No ART drugs available at clinic/OPD
CA702
Buildings/Beds
CA704
Delayed or lost laboratory results (especially HIV) at clinic/OPD
Initiation of ART at clinic/OPD delayed due to lost or delayed laboratory
investigations
No ART service provided at clinic/OPD
Cross-cutting
Staff
CAX01
No professional nurse at clinic/OPD
Other
Other
CAO
HIV
Laboratory
CA703
Other administrator modifiable factor at clinic/OPD (COMMENT)
Clinic and Outpatient Care - Family or Caregiver
Records
Danger Signs
Care Seeking &
Compliance
Cross-cutting
Other
Other
CFR01
Caregiver did not bring RTHC and/or referral letter to clinic
CF101
Caregiver refused treatment at clinic
CFX01
Did not arrive at clinic/OPD on day of referral/did not keep appointment
CFO
Other caregiver modifiable factor at clinic/OPD (COMMENT)
9
Child PIP
Child Healthcare Problem Identification Programme
Modifiable Factors
Saving lives
through death auditing
Child PIP v3.0.2
Home Care
Priority
Condition or
Activity
Category
v3 Code
Records
Notes
HPR01
Insufficient notes on home circumstances or child's health history
Danger signs
Disease
Prevention
HP101
Caregiver not advised about danger signs at previous visit
HPX02
Caregiver not advised about home treatment at previous visit
Nutrition/
Intake
Growth &
Development
HP501
Never referred to integrated nutrition programme (INP)
Assess
HP701
Caregiver not assessed and managed for HIV&AIDS
Disease
Prevention
HP702
Sibling of child with HIV&AIDS, but never traced and assessed
HP801
TB not notified
HP802
No TB contact tracing or treatment at home
Not referred for social grant, though eligible
Modifiable factor
Home Care - Clinical Personnel
HIV
TB
Disease
Prevention
Cross-cutting
Growth &
Development
HPX01
Other
Other
HPO
Records
Notes
HAR01
Lost RTHC not replaced (facility 'policy' not to replace cards)
Danger signs
Transport
HA101
Inadequate transport from home to nearest health facility
DD
Disease
Prevention
HA301
No tap water at home
Other clinical personnel modifiable factor in home/community (COMMENT)
Home Care - Administrator
Cross-cutting
Growth &
Development
HA302
No electricity
HAX01
Referred for grant (CSG, CDG, FCG) but never received
HAX02
Primary caregiver unemployed, or no household breadwinner
HAX03
Child came from child-headed household
No home/community IMCI in health subdistrict
Policy
HAX05
Other
Other
HAO
Records
Disease
Prevention
Care Seeking &
Compliance
Other administrator modifiable factor at home (COMMENT)
Home Care - Family or Caregiver
Danger signs
Care Seeking &
Compliance
Home Treatment
Nutrition/
Intake
HIV
Immunisation
Growth &
Development
Care Seeking &
Compliance
Care Seeking &
Compliance
HFR01
RTHC not used or lost by caregiver
HFR02
Caregiver unaware of child's health history
HF101
Caregiver did not recognise danger signs/severity of illness
HF102
'Traditional remedy' given from traditional healer, with negative effect on child
HF104
Caregiver delayed seeking care
Inappropriate treatment given at home with negative effect on the child, e.g.
enema
HF103
HF501
Child not provided with adequate (quality and/or quantity) food at home
HF701
Caregiver declined HIV test for the child
HF901
Caregiver did not take child to clinic for vaccines as scheduled
HFA01
Child accessed poison/drug
HFA02
Unsafe home environment (e.g. open flames)
Accidents
Home Treatment
HFA03
No adult supervision at home
Cross-cutting
Care Seeking &
Compliance
HFX01
Caregiver took child to clinic infrequently
Other
Other
HFO
Other caregiver modifiable factor at home/in community (COMMENT)
10
S A V I N G
Appendix D:
C H I L D R E N
2 0 0 9 :
F I V E
Additional Tools
Child PIP Mortality Review Process Guideline
187
Y E A R S
O F
D A T A
Child Healthcare Problem Identification Programme
Child PIP
The Child PIP Mortality Review Process
Child PIP v3.0.2
Saving lives
through death auditing
“Saving lives through death auditing”
It is the structured clinical audit of all children dying in hospital (including in
casualty/outpatients, and those who are ‘dead on arrival’) that enables a thorough
assessment of the quality of care that children receive in the health system.
For a clinical audit/mortality review to be successfully implemented there are two vital requirements:
1) Dedicated individuals willing to spend time and effort to make the process happen
2) A carefully structured system where roles and responsibilities are well-defined
Thus, the mortality review process in a paediatric/children’s ward consists of two main activities:
1) The data collection process
2) The actual mortality review process
Data collection
To conduct a mortality review, two data sources are needed:
1) The ward admissions, discharges and deaths register
2) The individual clinical records of the children who die
Keep a separate register of children who die for tracing their medical records, using the
monthly deaths register. Admission and deaths counts should be captured on monthly
tally sheets. Detailed information on each death should be captured on the death data
capture sheet.
To organise and keep track of the data it is helpful to compile a lever arch file, clearly labelled Child PIP. It is helpful to
order the contents in each section as follows:
1) Laminated copies of code lists (Cause of death and Modifiable factors), and growth charts
2) Monthly dividers for each month followed by a Monthly Tally Sheet for that month as well as Monthly Deaths
Register and a Death Data Capture Sheet completed for each death that occurred during that month
3) Spare data capture forms
The review process
Follow the four components of the mortality review process in your hospital:
Component
When
Who
Each death should be
The attending doctor or
1. 24 hour review reviewed and summarised
nurse at the time of the
within 24 hours
death
2. Preparatory
meeting
Before the Mortality
Review Meeting
The doctor and nurse in
charge of the ward/unit
Purpose
ƒ Ensure all necessary information is
captured at a time when information is
available
ƒ A detailed analysis of all deaths, with
case selection for presentation at the
Mortality Review Meeting
ƒ Compilation of monthly statistics for
presentation at the meeting
3. Mortality
review/Child PIP
meeting
(see below)
Weekly to monthly
depending on load
Whole paediatric
department (doctors
and nurses) as well as
clinic staff
ƒ Presentation of statistics, case
discussions and task reviews
ƒ Assign new tasks based on each
meeting’s discussion
ƒ Ensure all data capture sheets have
been completely completed
4. Epidemiology
& Analysis
6 monthly/annually
Managers and clinical
personnel
ƒ Broader problem identification with
trend assessment, and with proposed
solutions/recommendations
The 24 hour review
Every single death occurring in your hospital should be summarised using the Child PIP Death Data Capture sheet at the
time of death. The person best placed to do this is either the on-duty doctor or by way of handover, the daytime team
responsible for the long-term care of the child. The death summary should be regarded as no more burdensome, and no
less important, than the discharge summary for other children leaving the ward/unit.
It is still best to have a single person in the ward/unit making sure that this process
happens.
This can be a doctor or a nurse.
The preparatory meeting
This meeting is crucial. All data capture sheets must be completely completed, to the stage of readiness for entry onto
the computer. This means that all fields must be filled in, and codes must be entered where required. This makes data
entry onto the computer efficient and accurate, and allows for any category of employee to enter data.
Careful selection of cases for presentation will enhance learning opportunities, and facilitate problem identification, and
task definition and allocation.
The preparatory meeting is the responsibility of the most senior doctor and most senior
nurse in the ward/unit.
The mortality review meeting
Mortality meetings must be well organised and managed by the nurse and doctor responsible for the paediatric/children’s
ward.
1) Meetings should be held weekly to monthly depending on the number of deaths.
2) A suitable time and venue is needed.
3) All staff involved with child care should be invited (doctors, nurses, allied healthworkers and administrators). Staff
must understand that mortality meetings are very important. It is especially helpful to invite staff from clinics referring
to the hospital.
4) Case presentations should be concise and professional. Discussion is encouraged if the presenter does not provide
the cause of death and modifiable factors. This is best done by the group.
5) The meeting should by consensus establish the main cause of death and then look carefully for modifiable factors.
The meeting must never become a “witch hunt”, and should be confidential. The meeting should NOT be dominated
by senior doctors. The thoughts and insights of all participants make the meeting worthwhile.
6) All decisions (causes and modifiable factors) made must be recorded on the mortality sheets (death data capture
sheets) for entry later onto a computer.
7) Problems with the process of caring for children in the hospital, the referring clinics and in communities
must be identified and prioritised, and plans should be made and documented for addressing each problem.
8) Tasks arising out of discussions around cases should be assigned to team members, and minuted. Progress with
the tasks should be reviewed at the start of the next meeting.
The meeting agenda
A typical mortality review agenda is a follows:
1) Welcome and introductions, and identification of a minute taker
2) Review of tasks set at last meeting
3) Summary of last meeting’s statistics
4) Summary of this meeting’s statistics
5) Case presentations
6) Task identification and allocation
7) Closure and date of next meeting
Epidemiology and Analysis
The power of Child PIP lies in its ability to provide instant feedback on child death and quality of care information to
ward/unit staff. Simply by initiating this systematic review process, change will happen.
It is, however, important both for the identification of broader system problems and for monitoring change that 6-monthly
or annual reviews are performed.
These reviews should be compiled into reports, which document both findings and recommendations arising out of the
review. This is the point at which the power of Child PIP can be used for communicating problems to managers. Once
the process of mortality review is established in your site, the report will also look at success of implementation of, and
response to, previous recommendations.
You can use the Child PIP Report pro forma for guiding your report writing.
Making change happen
When making recommendations, it is important to link each recommendation clearly to specific information arising out of
your Child PIP review process. It is then useful to clearly define its requirements for implementation at each of the
following levels:
1) Policy
2) Administration
3) Clinical practice
4) Education
Finally, responsibility for implementation at each level should be assigned, so that at the next review, implementation (or
lack thereof) can be accounted for (for an example of this see Saving Children 2005).
By conducting mortality reviews in this systematic way, we will save lives and improve
quality of care, through death auditing.
(Adapted from Philpott and Voce: “4 Key Components of a Successful Perinatal Audit Process”, Kwikskwiz #29, 2001)
S A V I N G
C H I L D R E N
2 0 0 9 :
F I V E
Y E A R S
190
O F
D A T A