University of Groningen
The current hospital transfusion practices and procedures in Uganda
Kajja, Isaac
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to
cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2010
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Kajja, I. (2010). The current hospital transfusion practices and procedures in Uganda Groningen: s.n.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the
author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the
number of authors shown on this cover page is limited to 10 maximum.
Download date: 15-06-2017
Isaac (thesis).qxp:Layout 2
01-11-2010
09:30
Pagina 107
Chapter 8
Bottlenecks of Blood Processing
in Uganda
Kajja I, Kyeyune D, Bimenya GS and Smit Sibinga CTh
Published in: Transf Medicine, 2010 doi:10.1111/j.1365-3148.2010.01015.x
The Current Hospital Transfusion Practices and Procedures in Uganda
107
Isaac (thesis).qxp:Layout 2
01-11-2010
09:30
Pagina 108
ABSTRACT
Aim: To identify where and why delays occur in Uganda blood banks.
Background: The timely provision and supply of safe and efficacious blood components to
hospitals depends on sound systems in the processing blood banks. Poorly managed systems lead to apparent blood shortages in hospitals and increases discard rates due to expiry before dispatch.
Materials and methods: We reviewed records of 4,126 units of whole blood delivered by
the mobile collection teams to a major regional blood bank in the period March 1st to June
30th 2009, to ascertain the time intervals between the critical steps in the processing
chain. This was followed by interviews with staff in two blood banks to establish the causes
of process delays.
Results: The average duration between blood collection and final labeling (release from
quarantine for final storage) was 15.4 (SD 10.8) days. In time-line, the step between matrix
generation and grouping was (median duration 8 days) while grouping to labeling was
shortest (median duration 2 days). Blood expiry had the highest discard rate (0.17%) among
the non Transfusion Transmissible Infection marker causes. A minimally facilitated small
staff contributed to the process flaws.
Conclusion: A considerable amount of blood does not reach hospitals because of process
delays between collection and ultimate dispatch. This is caused by a thin staff working
with inadequate materials, out-of date methods, and in an overcrowded environment. Provision of adequate staff and improved financial allocations to the Uganda Blood Transfusion Services will mitigate this situation.
108
Isaac Kajja
Isaac (thesis).qxp:Layout 2
01-11-2010
09:30
Pagina 109
INTRODUCTION
The World Health Organization started more than 30 years ago the recognition of blood
transfusion as an essential part of patient care with adoption of the Resolution WHA28.72
by the World Health Assembly (WHO, 1975). In this respect, and with technical and financial assistance from the European Union and the USA, Uganda has taken steps to mobilize
human and material resources to develop a nationally coordinated blood transfusion service. Like any other low human development index (LHDI) country, Uganda Blood Transfusion Service (UBTS) is challenged by a low availability of voluntary non remunerated blood
donors-VNRBD; insufficient transport and storage facilities; low capacities in testing of donated blood and quality assurance in testing laboratories, and the unregulated clinical use
of blood (WHO,2002).
In Uganda Blood is collected through fixed and mobile donor sites. These include schools,
donor clubs, massive community drives and camping sessions. It is then delivered to regional blood banks where it is screened for safety, grouped and stored before dispatch to
hospitals. Blood is usually supplied as whole blood. However, component production is
slowly taking root in most of the regional blood banks (Kajja et al. 2010). Timely provision
of safe and efficacious blood and blood components to hospitals depends on sound logistics
and administrative systems in the processing blood banks. Delays at any processing step
in the blood bank leads to poor supply of blood to the consumers and increases the discard
rates due to expiry before dispatch.
Over a period of four years, there has been a nationwide increase in the amount of blood collected annually in Uganda (110,000 units in 2005 to 170,000 units in 2009). However this
has not mitigated the blood shortages in the hospitals supplied by the regional banks. In
the period from July 2008 to June 2009, the national total blood requests amounted to
189,400 units. However, only 156,393 units (82.5%) were supplied. It is postulated that
blood is held for a long time in banks before distribution to hospitals hence the shortages.
This is followed by a reduction in the shelf life of blood released from quarantine within the
banks and hospital transfusion laboratories, therefore increasing the discard rates.
The purpose of this study was to identify where and why the delays occur in Uganda’s Regional blood banks. It is anticipated that such analysis will strengthen the transfusion
chain, improve the supply of blood to hospitals and enhance patient care.
MATERIALS AND METHODS
Following ethical approval from The Uganda National Council of Science and Technology, a
mixed (quantitative and qualitative) study was conducted in two regional blood banks Nakasero and Mbale, respectively serving the central and eastern regions of Uganda. This
was done by analyzing the flow of blood from collection to hospital dispatch (Fig 1). First, it
involved the review of records in one bank. To ensure generalizability of findings, the
records review was followed by in-depth interviews with the technical and support staff in
the processing laboratories of the two banks.
The Current Hospital Transfusion Practices and Procedures in Uganda
109
Isaac (thesis).qxp:Layout 2
01-11-2010
09:30
Pagina 110
Figure 1 Critical steps in the production line.
Transportation of
collected blood to
blood bank
Matrix
generation by
IT department
TTI marker
testing
ABO/Rh
testing
Labeling/
quarantine
release
Storage of
finished
product
Table 1 Regional blood banks: Annual collections and demands from hospitals.
Regional blood bank
Population
Collections in
Demands in
Served x106
units@ of Blood (n)
units$ of Blood (n)
Nakasero
9.00
59,198
65,160
Mbarara
5.38
31,970
36,720
Mbale
5.50
25,024
27,840
Gulu
3.42
19,480
21,960
Fortportal
2.64
16,810
18,960
Kitovu
1.67
9,498
9,880
Arua
3.42
8,020
8,880
170,000
189,400
TOTAL
@
$
= units of donated whole blood.
= units of blood products (whole blood, Red cell concentrates, platelets, fresh frozen plasma) demanded by hospitals
served by a particular regional blood bank.
110
Isaac Kajja
Isaac (thesis).qxp:Layout 2
01-11-2010
09:30
Pagina 111
Study settings
Uganda Blood Transfusion Services runs a centralized administrative system. It has seven
regional blood banks serving a population of about 31 million (Table 1). All regional blood
banks collect a total of 170,000 units of whole blood annually and are able to fulfill about
83% of hospital blood demands. Nakasero regional blood bank serves a population of 9
million people, collects 59,198 units annually, (90.8%) of its 65,160 annual hospital demands. Mbale serves 5.5 million people, collects 25,024 units an equivalent of (89.8%) of
its 27,840 annual hospital demands. Each of the two banks has fixed and mobile donor
sites from where blood is collected. At the end of a donor session blood is either transported
to the bank or kept at designated collection centers awaiting transport to the bank in a day
or two. When blood is delivered to the bank, it is immediately quarantined for processing.
Each unit of blood is accompanied with two sample tubes; one for screening for Transfusion Transmissible Infections-TTI (HIV 1 and 2, Hepatitis B, Hepatitis C and Syphilis) and
the other for ABO/Rhesus grouping. It is also accompanied with a card bearing donor demographic and clinical details. A unique but similar number is stuck on a given unit of
blood with its test samples and donor card. The donor cards are forwarded to the information technology (IT) section for donor data entry and generation of a matrix (a computer
generated arrangement of donor numbers in rows and columns). The sequential arrangement of samples in racks for TTI screening and ABO/Rh grouping is a replica of the computer generated matrix. Results from screening and grouping are thereafter used first in
sorting out units of blood that are positive for any of the TTIs which are discarded, and second for the final labeling of the units with their ABO/Rh groups. After labeling the blood
leaves the quarantine and is stored ready for distribution to hospitals. In case of sample
tube shortages at the collection site, a given unit of blood will be accompanied by one instead of two tubes. This tube is first used by the TTI screening section then forwarded to
the ABO/Rh grouping section (Fig. 1). Moreover, both sections have to wait for the computer
generated matrix for this particular unit of blood. Quality control of the suitability of a unit
of blood is performed at the labeling and distributing steps. Here issues like expiry, volume
of blood in bag, hemolysis and quality of the bag containing blood are addressed.
Figure 2. Time intervals between the key preparatory activities.
The Current Hospital Transfusion Practices and Procedures in Uganda
111
Isaac (thesis).qxp:Layout 2
01-11-2010
09:30
Pagina 112
Table 2 Discard rates out of the 19,700 units collected from March 1st to June 30th 2009
Reason for discard
Rates n (%)
TTI marker reactivity:
Sero-positive for HBV
681 (3.46%)
Sero-positive for HCV
453 (2.30%)
Sero-positive for HIV
152 (0.77%)
Sero-positive for syphilis
71 (0.36%)
Others:
Expiry of blood
33 (0.17%)
Inadequate blood in collection bag
26 (0.13%)
Hemolysis
10 (0.05%)
Leakage of collection bag
TOTAL NUMBER OF UNITS DISCARDED
2 (0.008%)
1,428 (7.2%)
The records
We reviewed records of 4,126 of the 19,700 units of whole blood collected and delivered by
the mobile collection teams to Nakasero regional blood bank in the period from March 1st
to June 30th 2009, to ascertain the time intervals between the critical steps in the processing chain (figure 1). This allowed identification of major contributing step(s) – transportation, matrix formation (IT), testing (TTI markers and ABO/Rh), final labeling (quarantine
release) and final storage for dispatch to the hospitals - to the delays in blood processing.
Simultaneously, the rates and reasons for discarding blood during the same period were
noted.
The in-depth interviews
Based on the results of the average time a unit of blood spends at a given critical production
point as evidenced from record reviews, ten participants were purposefully selected for indepth interviews. These were people working at the critical production points of the blood
banks. They included drivers, laboratory technicians and administrators. The interviews
conducted at the participant work places lasted 50-60 minutes each. These were facilitated
by a trained research assistant in the presence of the principal investigator. Before the interview, the principal investigator reviewed the objectives of the study to the interviewee
and an oral consent was obtained from the interviewee. The interviews were guided by a set
of predetermined and standardized open-ended questions focusing on experiences and perceptions on staffing, quality of their working environment, methods and machines used,
and how these impact on the turnaround time that a unit of blood spends in their respective blood bank before distribution to hospitals. The proceedings of the interviews were
recorded using an audio recorder. Immediately after each session there was a debriefing
among the research team to ensure clarity of the recordings, documentation of any new
ideas or issues and plan for the next interview. The recordings were transcribed verbatim.
Data analysis
Continuous variables were represented by their respective median values. Statistical package for social sciences (SPSS, version 12.0; SPSS Inc. Chicago, IL, USA) program was used
for data analysis. A bivariate Pearson Product-Moment Correlation Index was performed to
find the relationship between the periods blood spends at different points in the blood
bank, with a significant correlation at the 0.01level (2-tailed).
112
Isaac Kajja
Isaac (thesis).qxp:Layout 2
01-11-2010
09:30
Pagina 113
Qualitative data analysis involved reading and re-reading of transcribed text in order to
identify key ideas. Sections of text carrying particular meanings were labeled and coded.
The coded units were then systematically compared and summarized into categories of text
with similar messages (Pope et al. 2000). Obvious and visible meanings of the categorized
text were noted (Manifest content analysis). This was followed by further synthesis and interpretation of the obvious meanings (latent content analysis) to generate themes (Graneheim & Lundman, 2004; Kondracki et al., 2002).
Figure 3 Example of codes, categories and themes from content analysis of narratives about delays in blood preparation.
The Current Hospital Transfusion Practices and Procedures in Uganda
113
Isaac (thesis).qxp:Layout 2
01-11-2010
09:30
Pagina 114
Figure 4 Features of the information technology system in use at Nakasero and Mbale regional blood banks.
) ! ) !
) & !
) ! $ ! " ) ! ! )! #& ) ! )" ) " "
) ! %! " $
RESULTS
Records
During the study period, 19,700 units of whole blood were collected and delivered to
Nakasero Blood bank. Out of these, 1,428 (7.2%) units were discarded. Among the non TTI
maker causes of blood discard at Nakasero regional blood bank, blood expiry contributed
the highest proportion while blood leakage from the collection bag was responsible for the
least proportion of discarded blood (0.17% against 0.007%) - (Table 2).
The 4.126 units of blood whose records were followed from collection to dispatch to hospitals, showed that the step, matrix generation till grouping, was longest (Median duration =
8 days). The steps, blood delivery to bank till matrix generation, and grouping till labeling,
took the least amount of time (median duration = 2days respectively) (Fig 2). There were
statistically significant but negative correlations between time spent by a unit of blood at a
given step in the chain and the time spent by the same unit at subsequent steps (all p values < 0.001). This means that if a unit of blood spent less time at step A (Fig 3), it would
spend more time at steps B, C, D, E, and vice-versa.
114
Isaac Kajja
Isaac (thesis).qxp:Layout 2
01-11-2010
09:30
Pagina 115
Responses from interviews explaining causes of process delays
Inadequate transport Facilities
On many occasions teams collecting blood from distant camping sites (more than 100kms
from the regional blood bank) spend two to four days. Camping sessions are a major
method of blood collection. Here blood is collected on daily basis, kept in cool boxes and delivered to the bank at the end of the camping session. One driver remarked:
“Each team has one vehicle so if we go camping there is no way we are going to move up and
down with a few units, we deliver the whole lot at the end of the camp.”
“Sometimes we get hemolysed samples because of poor storage during transportation, if you are
using automated testing methods; you will not get results for these sample. So you do it manually. It is time consuming and may give erroneous results.”
The other causes of process delays as alluded to by the respondents to the interviews are
summarized in figure 2.
Staff shortages
Intrinsic problems within staff were considered key in prolonging the process. These include: 1. inadequate staff compared to the work load. 2. Despite availability of an IT system
with acceptable features (Fig 4), there is lack of enough trained staff to use this system.
During the study period, it was observed that two persons attend a given shift in the IT,
grouping and labeling sections. These shortages in staff are illustrated by the quotes:
“It is only the two of us who do this work, if I fall sick the section head may or may not call another person from another section to fill up the gap, but that section is also affected”
“We all need to be trained on the supplied soft ware for our IT section so that if something goes
wrong we don’t have to wait for the data managers to come and fix it.”
Manual processing methods
The manual or semi-automated equipment used in the processing of blood were identified
as another cause of delays. These are characterized by frequent mechanical breakdowns
and erratic power supplies. This is reflected in the following narratives:
“When we get power shortage in the washing section, the procedure becomes complicated and
time wasting.”
“We should get an automated system to assist in the rapid testing other than using this manual
method in some of our tests. This will save time and protect us since we shall not be exposed to
these samples”
“… at the grouping stage, we prepare the incubation aliquots manually.”
Limited work space
The limited working space was identified as another cause of delay in the work flow. This is
characterized by mix up of documents necessitating rechecks on results before submission
to the following step. Comment from IT and TTI sections clearly show this:
“The small working place affects our speed. Sometimes plenty of cards come, you have to sort
them out according to the teams and you may mix them up, you may miss one or two which fall
off your table.” “Then you have to re-check and re-sort to ensure that you don’t make mistakes,
one mistake affects the whole matrix so you must take care in a narrow place”
The Current Hospital Transfusion Practices and Procedures in Uganda
115
Isaac (thesis).qxp:Layout 2
01-11-2010
09:30
Pagina 116
Inadequate supplies
The inconsistent supply of consumables was reported as another cause of delay in the processing of blood. As a government institution, UBTS procures consumables according the
Public Procurement and Disposal of Public Assets guidelines (2003). This involves procedures like advertisement, bidding, contracts approval and award before supplying the
items. This takes time as has been alluded to by one senior laboratory technician:
“If we don’t have reagents for a particular test then we can’t produce results, we have to wait for
the supply”
“The public procurement processes are really time consuming. It takes approximately six weeks
from requisition to supply of an item.”
Other identified bottlenecks in the procurement of consumables are the delayed and incomplete submission of requirements for a particular period by the various laboratory sections to the designated procurement officer. This hinders prior planning and delays
initiation of a long process.
DISCUSSION
Donated blood must go through some process – though with time limits – that adds value to
it so that it can serve the recipient hemotherapy needs. In any transfusion organization,
this process involves a blend of people, machines and materials, methods and environment. The current study has identified weaknesses in the above blend for Uganda’s regional blood banks that limit provision of adequate and timely blood products to patients
when needed. The rate limiting steps in this setting are delivery of blood from collection
centers to the banks, matrix generation, testing for transfusion transmissible infections
and ABO/Rh blood grouping. The identified blood bank bottlenecks are compounded by
weaknesses at the clinical interface. Kajja et al. 2010 established that in Uganda, clinical
staff does not appreciate the quality and quantity of the available blood products in their respective regional banks. Likewise, the blood bank staff does not fully appreciate the clinical
transfusion needs and constraints. This fragile clinical interface falls short of motivating
the overseeing administrative structures in the regional blood banks to produce adequate
and timely blood products.
In a resource constrained Uganda, with limited numbers of voluntary non-remunerated
blood donors (VNRBDs) and transport facilities, it would be cost-effective to transport
blood in bulk from collection sites to the processing banks. However, this causes reduced
supply of blood to hospitals and reduces the shelf life and quality of blood. In related work
by Hassall et al 2009; 2010 at Coastal Province General Hospital Mombasa, Kenya, it was
found that blood waiting long in blood banks before transfusion losses quality in the following ways; it carries increased bacterial contamination rates of the small pediatric blood
units particularly with laboratory pathogens, has increased red blood cell hemolysis rates
and has high extracellular potassium levels. The combination of unclean working conditions, warm ambient temperatures (20-300C) in sub-saharan Africa, high relative humidity
(80%-90%) and unreliable refrigeration all combine to produce significant quality hazards
to blood products.
The workload at matrix generation, testing and grouping sections overwhelms the available staff - two people attending a shift per section. This leads to the observed stalling in
the process flow at the banks. Noteworthy however, the existing staff is dedicated to its
work. They attempt to speedup activities at a given step following delays in the previous
116
Isaac Kajja
Isaac (thesis).qxp:Layout 2
01-11-2010
09:30
Pagina 117
preparatory steps. This is shown by the significant inverse proportionality (p<0.001) of
time spent by a unit of blood at a given step to the time it spends at the subsequent steps.
The UBTS is a semi-autonomous institution. It has no control over recruitment of employees. Its staff is recruited by the public service commission through the Ministry of Health.
As a National policy, there are limits to recruitments. These bureaucracies directly impact
on the number of staff that can be available to the transfusion services. Additionally, still
some people do not take up offered jobs at the bank due to poor remunerations. This leaves
most of the work to few staff. Additionally, all staff in information technology (IT) section is
not fully trained on the software currently in use. This leaves the work of matrix generation
to a few staff therefore delaying testing and grouping steps. The narrow job structures and
training gaps are not unique to Uganda blood banks. According to the World Health Organization Global Database on Blood safety (2002), and Kaseje’s report (2006) on Health care in
Africa, Low Human Development Index (L-HDI) countries are still challenged by inadequacy of the human resource in the health sector due to remuneration constrains and the
rampant brain drain to greener pastures.
In Uganda, transfusion services are 100% centrally budget funded. The UBTS does not
run any cost recovery programs from either the supplied hospitals or the transfused patients. As pointed out by Hensher and Jeffrey (2000), this single income stream has an inherent weakness in situations of delayed or limited funding. Here the supply of
consumables and other recurrent utilities will be affected and this impacts directly on flow
rates of processes using those items. This weakness has been clearly demonstrated in this
study.
A Blood Transfusion Service is unique in the sense that it handles a precious and perishable product. The procedures of procuring its operational utilities should not be equated to
those of any other organization as is the case in Uganda – (according to Public Procurement
and Disposal of Public Assets Regulations). Therefore, we recommend an urgent review of
the rules and regulations governing the procurement of consumables for blood bank laboratories in Uganda. We also recommend administrative emphasis to staff on the importance of timely and comprehensive submission of annual requirements to the concerned
procurement sections. This will ensure adequate supply of consumables throughout the
year.
The low level of revenue available to the Ministry of Health to cater for the entire national
health budget does not allow UBTS to access or use fully automated machines in their operations. Therefore, they resort to the slow manual methods in some procedures like testing for syphilis and blood grouping. Unlike manual method, automated methods have been
found to carry the following advantages; they give accurate, consistent and reliable results
irrespective of work load, are faster than manual methods therefore reduce the turnaround
time a unit of blood spends in quarantine, and offer a faster storage and transfer of data between sections of the laboratory. Additionally, but in the long run, automated methods are
cheaper than the manual methods. The initial capital expenses are out-weighed by the running costs of the manual methods which use more antiserum, require more staff and carry
a compromised patient safety outlook due to operator fatigue especially towards the end of
busy shifts (Gershon et al.,1996; Sazama, 1996)
Douglas (1982), recommended that the most important step in improvement of quality of
compatibility testing is the management of stressful conditions in transfusion laboratories. This has been emphasized in the American Association of Blood Bank (AABB) standards for blood banks and transfusion services (2009). The current design of the work
places in the IT, blood grouping and testing sections does not provide adequate space for
The Current Hospital Transfusion Practices and Procedures in Uganda
117
Isaac (thesis).qxp:Layout 2
01-11-2010
09:30
Pagina 118
easy and fast execution of tasks. We postulate that the psychological stress of working in
such an environment could impact strongly on the work flow rate in these areas hence the
observed delays.
The major limitation of this study was failure to find an isolated room to conduct the interviews in the congested laboratories. This could have affected free and spontaneity of flow of
information from interviewees. However, it reaffirms the urgent necessity to address the
unfriendly work environment in the blood banks.
In summary, this study has established that blood looses a substantial amount of its shelf
life in processing blood banks due to; inadequate staffing, shortages in materials and
equipment, and poor work environments. In order to cater for the unique nature of blood
processing as a service, we recommend the following: First, the Government of Uganda
should review the staff ceiling on UBTS accompanied by modest staff remuneration and
appropriate training. Secondary, government of Uganda should develop a procurement
model unique to transfusion services. This will ensure continuous and timely provision of
the recurrent blood bank supplies and utilities. Thirdly, there should be improved funding
towards procurement of fixed assets for the transfusion services. This will lead to a shift
from use of manual to automated equipment in the banks. It will also ensure a constant
upgrading of the IT software so that it is compatible with any newly acquired equipment.
Fourth, and as a long term plan, there should be expansion of the existing facilities (blood
bank buildings) to cater for the current lack of work space and furniture in many regional
blood banks. This should be closely followed by training of blood bank staff on standard
laboratory hygiene and safety principles.
ACKNOWLEDGEMENTS
The authors would like to thank all staff of the participating Regional Blood banks for their
loyal and valuable inputs. They also acknowledge the valuable contribution of Mr. Abwooli
Gilbert, Laboratory technologist Nakasero regional blood bank, for helping in data collection, Mr. APM Los, Medical Sociologist Sanquin Consulting Services, The Netherlands, for
designing and standardizing the guide questions of interviews and advising on the statistical evaluation of the collected materials.
118
Isaac Kajja
Isaac (thesis).qxp:Layout 2
01-11-2010
09:30
Pagina 119
REFERENCES
1 Douglas R. The difficult crossmatch: A consideration of staff not methods. Transfusion,
1982;22:333-336.
2 Facilities and safety. In Standards for Blood Banks and Transfusion Services, 2009 Ed 26;
page 84.AABB Press. Bethesda, Maryland.
3 Gershon, H.G., Galenza, J., Heather, M., Whitehead, R.& Godolphin, W. The implementation and use of automated group and screen procedures in a hospital transfusion laboratory. Transf Med Reviews 1996 ;10:144-151.
4 Graneheim, UH and Lundman BQualitative content analysis in nursing research: concepts, procedures and measures to achieve trustworthiness
Nurse Educ Today .2004;24:105-112
5 Hassall,O., Maitland, K., Pole,L., Mwarumba,S., Denje,D.,Wambua,K.,Lowe,B.,
Parry,C., Mandaliya,K.& Bates,I. Bacterial contamination of pediatric whole blood
transfusions in a Kenyan hospital. Transfusion 2009;49:2594-2598.
6 Hassall,O., Maitland, K., Fegan, G., Thitiri, J., Pole, L., Mwakesi, R., Denje, D., Wambua,
K., Mandaliya, K.& Imelda Bates, I. The quality of stored umbilical cord and adult-donated whole blood in Mombasa, Kenya Transfusion 2010;50:611-616.
7 Hensher, M & Jeffrey, E. Financing blood transfusion services in sub-Saharan Africa: a
role of user fees. Health Policy and Planning; 2000;15:287-295. Oxford University Press
8 Kajja,I., Bimenya, G.S., Smit Sibinga, C.T. The interface between blood preparation and
use in Uganda. Vox Sang 2010;98: e257–e262
9 Kaseje D. Health care in Africa: Challenges, opportunities and an emerging model for
improvement. 2006 Available at http://wilsoncenter.org/topics/docs/Kaseje2.pdf accessed 2.10.2009.
10 Kondracki. N.L., Wellman, N.S., Amundson, D.R. Content analysis: review of methods
and their applications in nutrition education. J Nutr Educ Behaviour 2002;34;224-230.
11 Pope, C., Ziebland, S., May, N. Qualitative research in health care .Analyzing qualitative
data. Brit Med Jl 2000;320;114-116.
12 Resolution WHA28.72. Utilization and supply of human blood and blood products.
Twenty-eighth World Health Assembly, 1975 Geneva, World Health Organization.
13 Sazama KCurrent Good Manufacturing Practices for transfusion medicine. Transf Med
Reviews, .1996;10:286-295.
14 The Republic of Uganda- Public Procurement and Disposal of Public Assets (PPDA)
Regulations No. 70 of 2003. Available on www.ppda.go.ug Accessed 28th January, 2010.
15 World Health Organization Global database on Blood safety, 2002. Available at
www.who.int/bloodsafety/GDBS_Report_2001-2002.pdf. Accessed 3rd August 2009
The Current Hospital Transfusion Practices and Procedures in Uganda
119
Isaac (thesis).qxp:Layout 2
120
01-11-2010
09:30
Pagina 120
Isaac Kajja