zg026 (jr/d)
14/4/03
2:14 pm
Page 214
HEALTH POLICY AND PLANNING; 18(2): 214–224
doi: 10.1093/heapol/czg026
Health Policy and Planning 18(2),
© Oxford University Press 2003,
all rights reserved.
Implementing a new health management information system
in Uganda
J GLADWIN,1 RA DIXON2 AND TD WILSON3
1Health Services Research Unit, London School of Hygiene and Tropical Medicine, 2Sheffield Health and Social
Research Consortium and 3Department of Information Studies, University of Sheffield, UK
This paper reports on research investigating the health management information system (HMIS) implementation process in Uganda, utilizing the diffusion of innovation and dynamic equilibrium organizational change
models. Neither perspective guided the HMIS development process. Instead, technological issues, rather
than wider organizational issues, dominated the planned change. The need to consider the organizational
context when changing information systems arises because the process is more complex than some practitioners have realized, when attempting to understand the causes of information management problems
and developing HMIS in low-income countries.
In particular, information system developers had not acknowledged that they were promoting an informational approach to management when they promoted a change from a centralized reporting system to a
HMIS supporting use of information at the level of collection. Strategies to facilitate this approach were not
advocated.
Organizational theory can contribute to the diffusion of innovation framework. It has yielded an integration
of Rogers’s diffusion of innovation framework and Leavitt’s concept of organizational forces in equilibrium.
The diffusion framework describes the process, but the organizational model has given the context and
reason for aspects of the process. The diffusion model does not predict what needs to change within the
organization when a particular innovation is introduced, or how much. The addition of the organizational
model has helped.
These frameworks can facilitate the introduction of future information management innovations and allow
practitioners to perceive their introduction as a staged process needing to be managed. Implications for practice are identified.
Key words: health information, organisational change, diffusion of innovation, information use
Introduction
Until 1993 Uganda had a centralized health information
system (HIS) focusing on morbidity and mortality reporting,
with data flowing only from individual health units to the
district and national level. Since then, the Ministry of Health
(MOH) implemented a health management information
system (HMIS) that emphasizes use of information at the
point of collection.
Many other low-income countries are moving in the same
direction, with more skills demanded of primary health care
(PHC) managers, including data and information handling at
all levels of the health care system (AKF 1993). The World
Health Organization (WHO) identified district-oriented
health information systems as a priority (WHO 1988) and
noted that ‘weakness of information support is acknowledged by most member states as a persistent obstacle to
vigorous and objective management’, and that ‘efforts to
strengthen national information systems have often
produced little improvement and have sometimes made the
problems worse’ (WHO 1994a).
The Commission on Health Research for Development
(1990) identified a need for research on the development of
practical health information systems to guide policy and
management decisions, and HIS improvements were
identified as essential by countries such as Tanzania
(Research into Action 1999: 8) and the Caribbean (Research
into Action 2000: 5). International partnerships have mobilized funds and technical support as HIS improvements were
advocated as part of improving health in low-income countries (World Bank 1993). HIS improvements in many countries including Peru (World Bank 1999a), the Solomon
Islands (World Bank 1999b), Honduras (World Bank 2000b),
Lesotho (World Bank 2000a), Uttar Pradesh in India (World
Bank 2000b) and Kryragya Republic (World Bank 2001) are
supported.
WHO has taken a lead in helping low-income countries
develop HIS by providing technical and financial support to
assess, design and develop such systems. No standard
development strategy for WHO support is used, but specific
principles for guiding HIS development and technical
cooperation are encouraged (WHO 2000a). A review of the
zg026 (jr/d)
14/4/03
2:14 pm
Page 215
Implementing a health management information system
support has been undertaken (WHO 1999) and a strategy to
support HIS development in member states is being developed (WHO 2000b, 2001a). Guidance on needs assessments
has been detailed (Lippeveld et al. 2000; WHO 2000c) and
experiences in HIS development described (Lippeveld et al.
2000).
Developing and improving health information
systems
A detailed review of research and reports on information
management for health unit and district managers in lowincome countries found many problems (Gladwin 1999),
including: data collection and processing issues (Nabarro et
al. 1988; Feacham et al. 1989; VanNorren et al. 1989; Robey
and Lee 1990; Foltz 1993; Jayasuriya 1993; Van Hartevelt
1993; Sandiford 1994; WHO 1994b; Jayasuriya 1999;
Lippeveld et al. 2000); over-reliance on special purpose
surveys and epidemiological data (De-Kadt 1989; Keller
1991; Hussein et al. 1993; WHO 1994b, 2000a; Sapirie and
Orzeszyna 1995); poor information use (Smith et al. 1988;
Wilson et al. 1988; Feachem et al. 1989; Bekui 1991; CamposOutcalt 1991; Keller 1991; Finau 1994; Loevinsohn 1994;
WHO 1994b, 1999; Sapirie and Orzeszyna 1995; Braa et al.
1997); identifying information required at specific levels to
monitor equity, coverage, quality and efficiency (Smith et al.
1988; WHO 1994b); general organizational and management
problems (Newbrander et al. 1988; De-Kadt 1989; Robey and
Lee 1990; Sandiford et al. 1992; Foltz 1993; Husein et al. 1993;
WHO 1994b, 2000a; Campbell et al. 1996; Braa et al. 1997;
Azubuike and Ehiri 1999); lack of an overall organizational
information strategy in low-income countries (Robey and
Lee 1990; Van Hartevelt 1993).
Many of these problems indicate a need for information that
could inform various aspects of operational managers’ policy
implementation, monitoring, evaluation and planning role,
rather than contribute to profiling morbidity and mortality
status for national use, which was often the aim of central
reporting HIS. An underlying concept in many papers is the
need for information management strategies to promote an
informational approach to health unit and district level
health management.
Few publications describe the development of a new HMIS
for operational management in a low-income country,
although some describe the adoption of information technology at national or health unit level (Newbrander and
Thomason 1988; Singh et al. 1992; Jayasuriya 1999). Information system developers in Ghana (Campbell et al. 1996)
recognized the scarcity of research describing the process of
developing HMIS for operational managers. Van Hartevelt
(1993) discusses the need for an information management
approach when strengthening information systems in Ghana.
Foltz (1993) describes technology transfer in Chad to
improve a national reporting system, apparently not developed for operational management of health services, but
does not measure success in such use of information. A case
study of information system development in Niger (Mock et
al. 1993) describes the change from a centralized reporting
215
system to an HMIS, but only MOH central management is
facilitated. Duran-Arenas et al. (1998) describe the development of an information system in Mexico which they believe
has implications for low-income countries. Robey and Lee
(1990), describing the HIS redesign process in the Philippines, identify useful lessons, as does Jayasuriya (1999) in
implementing a computerized information system. Heywood
and Campbell (1997) identify the need for critical appraisal
of determinants of success and failure of HMIS. WHO
identifies the need to document common barriers to establishing and sustaining effective routine HMIS and identifies
strategies to minimize their effects (WHO 2001b).
This paper reports on a research study to understand the
information system implementation process in Uganda as it
moved from a centralized reporting HIS to a HMIS that
supports district and health unit management, using existing
theory and research to deepen our understanding.
Methodology
Uganda was chosen because it was developing the HMIS
when research funding was available. The qualitative
approach follows Morse (1994), with the research strategy
based not on conscious, prior consideration of philosophical
questions, but on study purpose, research question, skills and
resources available: if the question concerns the nature of the
phenomenon, then the answer is best obtained by using
ethnography.
We used participant observation, interviews, official
document examination, written field-notes and diaries. Two
country visits were made by JG, using the peripheral
observer role (Alder and Alder 1994: 380) during a 1-day
workshop and later during a 9-week period of HMIS training.
Twenty-nine in-depth interviews, 47 semi-structured interviews and 19 group discussions with health unit, district level
and national health officials, academics, district management
trainers and other health service providers complemented
this work. Theme development was aided by NUD.ist
software.
Using Phillips and Pugh’s (1994) classification of relationships between theory and empirical evidence, we use theory
only to elucidate and develop further understanding when
analyzing empirical data. Although initial data collection and
analysis were not guided by existing theory, we were theoretically aware (Glaser and Strauss 1967) beforehand.
The theoretical constructs of most value have been Rogers’s
(1995) diffusion of innovation framework and Leavitt’s
(1965) idea of an organization as forces in dynamic equilibrium, elaborated by Leavitt et al. (1973), where organizations
have a semi-permanent framework, an arrangement of the
processes, material resources and people in some sequence
and hierarchy: change in one part leads to changes elsewhere.
The diffusion of innovation framework is an approach to
identify problems and specify a solution, or identify in
advance issues that inhibit or facilitate adoption of a technological change. The decision regarding an innovation is seen
zg026 (jr/d)
14/4/03
216
2:14 pm
Page 216
J Gladwin et al.
as a developing process, and a staged model, the InnovationDecision Process, is proposed. Implementation does not
automatically follow the decision to adopt an innovation –
Rogers (1995) describes this as the Innovation-Process model
of innovations within organizations.
Background
Uganda’s health sector reform was part of overall public
sector reform (Villadsen 1996), involving a form of decentralization (Mills et al. 1990) called devolution (Jeppsson and
Okuonzi 2000). The district administration was strengthened
with some formal transfer of power to lower levels as well,
but, with insufficient financial means, the districts relied on
central government. The MOH expected the districts to
develop their own plans to reflect national polices and guidelines, but with district priorities and requirements (MOH
1993). Uganda had 45 districts in 1997, with no intermediate
level. Each district has counties, sub-counties, parishes and
villages.
Other major health policy changes taking place included
restoration of services to acceptable levels to match the
changing social, economic and political environment (MOH
1993b), and reorientation of health services to primary health
care (PHC).
There was a legacy of irregular and inadequate wages for
health workers and undue influence by international donors
– on health services and policy (Okuonzi and Macrae 1995)
– and by the nationally based vertical programmes. In 1987
a new HIS was introduced by the Health Planning Unit of
the MOH; this was replaced by the HMIS in 1996/7.
The HIS was geared toward central planning to produce
information on health unit activity to support international
donors’ reports to their headquarters. The information was
supplemented by ad hoc community-based surveys carried
out by non-government organizations. Unlike the HIS, the
HMIS information was to be for decision-making and
improving operational health services performance. All
health units (including non-government units) were to
collect, process and report routine data relevant both to
national policy and health programme objectives and to the
needs of health unit health professionals. The design
identified critical management questions that the information should answer and, to identify appropriate data
collection, processing and analysis, utilized the systems
framework of inputs, procedures, outputs and outcomes for
immediate management, rather than awaiting higher-level
feedback. The system was to be integrated by having one
data source and set of forms in the health facility, so that all
existing health programme and general administrative information would be brought together, instead of having parallel
and duplicate information.
Specified data and information flows included: the internal
flow of information amongst the health unit team; written
monthly reporting from the health unit to the District Health
Team (DHT); oral reporting of specific information to DHT
members on supervisory visits; and written feedback from
the DHT for comparison with other health facilities.
Data processing and analysis were intended to be primarily
conducted by health unit staff, processed into summary
values to show changes over time, and provide performance
indicators at health facility, district and national level. Aggregation of data only to the level where information could be
produced for decision-making was intended to make the data
meaningful. Graphs of routine information were to be
produced by health units for their own purposes.
The intended use of information in health units was not
always clear from the documentation, but implied: indicators
of low performance leading to examination of individual
records to provide insights into how to improve; collection of
specific information to trigger certain actions at health unit
and district level, such as specific disease notification leading
to investigation; service targets to be made using population
information, knowledge of attendance and available resources; and information to be used to answer specific
management questions and plan future health unit services.
Information use at district level was intended to include:
formulation, monitoring and evaluation of annual workplans;
monitoring and improvement of health unit service delivery
in coordination with support supervision visits; and reporting
of selected information to the District Health Committee for
planning, monitoring and evaluating progress towards
district and national objectives. Reports generated for
various national MOH and donor departments were to be
used for national planning and policy formulation.
Some features within the HMIS were not always made
explicit: certain management tools; the teaching and supervision role of the Extended DHT (EDHT); and the ‘informational approach’ to decision-making, encompassing the
‘rational health unit decision-maker’.
The HMIS was initially designed and developed with expatriate consultant help and input from the MOH and other
health care providers and donors. It was piloted in two
districts and extended gradually to all districts. Two developers and 12 trainers trained the EDHT, who then trained
in-charges and senior health staff at health units.
Results: adoption of a HMIS innovation
The HMIS is viewed here as innovation diffusion, focusing
on the Implementation stage, when district personnel were
trained. Figure 1 illustrates our use of the Innovation-Process
model with concepts from the Innovation-Decision model.
Leavitt’s dynamic equilibrium framework was essential to
complete that understanding, as indicated by the diamond
shapes.
Redefining the innovation
Redefining the innovation to meet organizational needs and
structure took place as the HMIS definition and its purpose
varied. The developer and trainers did not adequately
zg026 (jr/d)
14/4/03
2:14 pm
Page 217
Implementing a health management information system
Adoption
Decision
217
Implementation
Redefining and Restructuring
Redefining of HMIS
Lack of management
Prior
tools necessary to
conditions
utilize innovation
Communication and social structure
impedes diffusion
Seeking knowledge
Organizational restructuring due to
HMIS
Using HMIS to bring in other changes
Attributes of innovation
Designer,
Developer,
Trainers,
Supervisors and
I/Cs lack full
Lack of power to make decisions
Innovation organizationally inappropriate
and constrains implementation
at health unit level
Other constraints to
implementation
knowledge of HMIS
Figure 1. Evidence for the Innovation-Process model: implementing HMIS
indicates where the dynamic equilibrium model is relevant.
explain the Principles Knowledge embedded in the HMIS,
namely the anticipated new approach to management and
decision-making based on information. Examination of the
specified management questions was not in the classroombased curriculum for supervisors or in-charges. Trainers
redefined the HMIS by focusing their training not on health
unit information use but on data collection and processing for
in-charges and by leaving use of information and much
processing for district-level staff.
New definitions were probably dependent upon individual
ability and role. Health workers saw the HMIS as: new and
integrated, with fewer forms; new centres for holding information; logistic and supplies data, morbidity data and data
produced through interaction with patients. The HMIS
affected data flows; monthly reports were going to district
offices, and duplicate reports were probably not sent to
national level. Health units did not process information as
expected; many health workers could not graph data and
district staff did this. Using information to inform decisions
proved too difficult for in-charges and often district or
national level staff set targets instead.
Innovation adoption at health units was only partial. The
HMIS was serving district needs more than health unit needs,
so redefining had taken place. One District Medical Officer
said he used the HMIS information “for the annual Workplan
we are doing now, for setting priorities, by knowing the most
recurring diseases, for resource calculations, to know the
numbers of patients. But I think the HMIS is of more benefit
to the district than to the health unit, especially with
decentralization.” District staff processed the data and new
forms to varying degrees, though some found the work too
difficult.
Reasons for redefining
Previous practice contributed to redefining and included:
existing information management problems; the HIS information management strategies; old ways of working;
concomitant changes; recent policy not enacted; management problems at district and health unit level; excessive
influence of international donors; and organizational and
cultural practices. Other reasons, also identified by Rogers
(1995), included: adopters lacking full knowledge of the innovation; the desire to simplify a complex and difficult to understand innovation; and the innovation being an abstract
concept and tool. Other reasons, not identified by Rogers,
were: ‘inventors’, change agents and aides lacked full knowledge of the innovation; lack of management tools to utilize
the innovation; and insufficient decision-making power.
The innovation’s perceived attributes and incompatibility
with management roles, ability, policy and organizational
situation were probably contributing to redefining. There
was a lack of tools to monitor and evaluate the innovation’s
implementation and use, and a lack of understanding of
changes needed with the HMIS. The management tools,
power and attributes of the innovation were not aligned with
the HMIS in accordance with the dynamic equilibrium
concept.
Restructuring
Organizational restructuring occurred since the innovation
was introduced but not aligned, including: the role of the
Medical Records Officer (MRO) at district and health unit
level; the power structure within health units and district
offices; administrative procedures; additional contractual
zg026 (jr/d)
14/4/03
2:14 pm
Page 218
218
J Gladwin et al.
relationships; and empowerment of some people to whom
the HMIS training brought extra finance.
Constraints on implementation
The hierarchy of power in health units and at district level
constrained HMIS implementation, as decentralization was
not fully enacted or understood. The previous hierarchy of
power placed the MRO in a lowly position, but the district
MRO was now expected to be an HMIS supervisor, even of
clinically trained staff, and fellow DHT members were
expected to produce reports the MRO could file, collate and
even interpret, which meant he had the right to demand
reports from these staff. While changes were expected in
health units, developers and trainers were unaware of these
constraints. Although face-to-face training of in-charges by
supervisors was probably an appropriate communications
strategy, insufficient time was allocated. Hence the social and
communication structures impeded diffusion.
Several constraints related to the HMIS’s perceived attributes. The intended training approach was not always undertaken and there was lack of understanding of changes needed
to accompany the innovation. Several unfounded assumptions were made about health unit staff and procedures when
identifying relevant information management strategies. For
example, HMIS data collection, processing and information
use assumes a certain level of general education and
specialist training amongst health workers, but this was not
available, especially in smaller health units. Too few support
supervision visits for HMIS training were made for health
unit personnel to grasp new skills, such as data processing,
compiling graphs and statistics. The trainers falsely assumed
that in-charges could easily be taught data processing
methods, but after 3 years many in-charges in pilot areas had
problems. Data collection was not always linked to diagnostic
ability as many health units did not have the equipment
and/or expertise to diagnose the diseases monitored. Many
nursing aides trained on-the-job had insufficient grasp of
English language concepts in the materials. Workers’ skills
were not aligned with the HMIS.
Organizational changes intended to be in place before HMIS
implementation, including decentralization and the extension of managerial responsibilities for health unit clinicians,
were incomplete. Thus there was some incompatibility of
information management strategies with management roles,
ability, policy and organizational situation.
After 3 years of engaging with the HMIS, people were
seeking innovation knowledge of various kinds. The developer lacked faith in the usefulness of the HMIS, and there
were too few tools to monitor and evaluate the innovation’s
implementation and use. These constraints may be due to the
HMIS definition or to the management of information
system development. Classifying these as incompatibility of
perceived attributes deepens understanding, as Rogers
(1995) suggests, but it could be interpreted as the innovation
upsetting the dynamic equilibrium of the organization.
A key perception of the implementers was that international
donors had excessive influence over health unit data collection and, because they were interested in particular countrywide programmes, national level MOH, rather than district
or health unit staff, often determined data collection.
The complexity of introducing the HMIS has been clarified
after developing a combined and expanded model from
Rogers’s two models. Our findings on HMIS implementation
do not fit neatly into the classical Innovation-Decision
Process model, particularly because the implementation
phase is too limiting.
The Innovation-Process model’s redefining and restructuring
concepts offered clarification, as did many InnovationDecision model concepts, including: prior conditions affecting implementation, perceived attributes, and lack of
knowledge constraining implementation. Unlike Rogers’s
(1995) Innovation-Decision model, it appears these concepts,
which would usually be important before the adoption
decision, were important in the implementation stage. Reinvention of the innovation during implementation is possible
(Rogers 1995).
These two models do not entirely explain the evidence.
Rogers’s (1995) concept of structural change within an
organization is too limited to understand how different
aspects of the organization change with innovation introduction. An adapted version of Leavitt’s dynamic equilibrium
model was developed viewing the HMIS as the technology.
Figure 2 illustrates the dynamic equilibrium model, with
suggestions of where non-alignment occurred in health units.
The earlier description indicated management roles or incharge ability were not always appropriate to the information
management strategies, and vice versa. Management roles
and the HMIS were not ‘aligned’; the HMIS assumes incharges will be managers, with monitoring, evaluating,
controlling and planning responsibilities, but these responsibilities have not been completely devolved. In-charges do not
have complete control over drug supply, which the rational
decision-maker idea assumes. The patchy implementation of
the cost-sharing policy and incomplete financial decentralization illustrate lack of alignment between structure and
information management strategies; cost analysis procedures
were sometimes redundant. If in-charges wanted to act as
rational managers and take an informational approach to
decision-making, lack of power prevented this.
The intended strategy of small health units is to provide
comprehensive PHC services with principles of equity and
community participation. In-charges have to monitor,
control and evaluate health centre services and resources,
manage staff and provide accommodation. For there to be
alignment of these two aspects, information to support the
policy and roles is needed, but this was often not the case.
Non-alignment of technology with organizational strategy
arose, as the data to monitor the strategies were lacking. Few
data were collected which could indicate which specific
groups within the health unit catchment area were in greater
need or accessing more services than other groups. Data
collection or management questions on the following were
zg026 (jr/d)
14/4/03
2:14 pm
Page 219
Implementing a health management information system
Structure:
*Insufficient decentralization;
*Poor communication
Management style, tools
Intended
HMIS:
and procedures
Strategy:
*Operational management of
primary and secondary units;
*Enact conceptual framework;
*Enact health policy;
*Increase market share;
*Strategy varies according to
unit type and staff ability;
*Manage health unit staff
219
Individuals and roles:
*Individuals not trained in using management
skills and tools;
*Role is often administration not management;
*Job changes taking place
*Different definitions include:
data collection and processing,
information use strategies,
rational/informational
approach;
*Inappropriate data collection
processing and information use
Key:
Indicates non-alignment
Figure 2. Application of dynamic equilibrium model of organizational change in health units: the HMIS is not aligned to Intended Strategy,
Structure, Individuals and Roles, or Management process
not incorporated into the HMIS: socioeconomic factors,
access to clean water and sanitation, and attitudes, practice
and knowledge of health-related behaviour – all components
of Comprehensive PHC (CPHC) vision.
Assuming management tools, procedures and style are an
additional organizational force (Scott Morton 1991), this case
study indicates non-alignment of health unit information
management strategies and this force. One key informant
said it was difficult to use information on the percentage of
children with protein-energy malnutrition “as there are no
standardized case definitions”, an essential management tool.
Information management strategies were not linked to
health unit decision rules, and there was non-alignment of
strategy and HMIS. ‘Individuals and Roles’ appear to be nonaligned with the HMIS, although some job changes were
taking place to facilitate alignment. Some administrative
procedures and information management strategies were not
aligned.
Figure 3 illustrates the dynamic equilibrium model of
organizational change in understanding the district level. The
intended district level strategies were not aligned to the
HMIS. One District Medical Officer felt the HMIS needed
to incorporate extra information related to aspects of CPHC,
in particular, the idea of socioeconomic factors influencing
health status. As some in-charges were not able to carry out
some intended management activities, district managers
were undertaking these. The informational or rational
management approach was not fully utilized at district level
as many DHT members were unable to use the management
tools, indicating non-alignment. There was non-alignment of
the HMIS and supervisors’ skills (Individuals and Roles),
especially as management training, teaching skills, improved
data processing skills and skills in using information are
needed. Some DHT management training was undertaken
and as one DHT officer said: “It’s only since the mid-level
managers’ course that I know the value of information for
planning.” He had only just realized the district MRO had
useful information. Some changes in the MRO role were also
taking place; the technology was pushing realignment of individuals and roles.
Non-alignment of organizational structure with the HMIS
was apparent to the developer, who felt lack of district level
financial control inhibited information use. In a devolved
situation, international donors should negotiate with district
and national level personnel, yet when donors held a district
meeting, they merely gave information. It was also felt the
organizational structure prevented implementation in other
ways. One District Health Volunteer believed the HMIS
needed a good communication and referral system in order
to work, but this was lacking, especially during the rainy
season. With decentralization there had been some redistribution and reinforcement of power at DHT level and District
Medical Officers became more powerful, but the teamworking idea imposed by national and international agencies
could conflict with this.
The exploration of the change from the HIS to the HMIS,
within Leavitt’s theoretical framework, reveals other changes
needed to ensure equilibrium and proper information system
functioning. It proved useful to change Task to Strategy and
add Management procedures, tools and style as an additional
force within the organization, though this organizational
change model has been useful only as an adjunct to the Innovation-Process model. Dynamic equilibrium diamonds are
placed within the Redefining and Restructuring Stage
(Figure 1) in order to complete understanding of the process.
The diffusion framework describes the process, but the
zg026 (jr/d)
14/4/03
2:14 pm
220
Page 220
J Gladwin et al.
Structure:
Intended Strategy:
*Technical support;
*Enact health policy;
*Coordinate monitoring,
evaluation and planning of
primary and secondary care;
*Management of communitybased and health unit health
workers;
*Lead coordination of district
ministries?;
*Enact conceptual framework
*Decentralization incomplete
*HMIS is health unit based but
additional structure exists
Management style, tools
and procedures
* Data collection and processing,
*Information use;
HMIS: *Rational manager/informational
approach;
*Inappropriate strategies
Individuals and roles:
*Some management trained;
*Lack of data processing skills;
*Management and technical support role;
*Some changes taking place
*Lack of teaching skills
Key:
Indicates non-alignment
Figure 3. District level application of dynamic equilibrium model of organizational change: the HMIS is not aligned to intended Strategy,
Structure, Individuals and Roles, or Management processes
organizational model has given the context and reason for
aspects of that process. The diffusion model does not predict
what needs to change within the organization when a particular innovation is introduced, or how much, but the addition
of the organizational model has helped.
Discussion
combination of organization and computer technology. The
developers and trainers did not fully recognize that the HMIS
was intended to promote an informational approach to
management. In Niger (Mock et al. 1993), information
system developers realized they were introducing not only
statistical techniques, but also a new management approach
with wider organizational consequences.
The extent of change needed to accompany the HMIS had
not been recognized and health workers were focusing upon
the small, rather than significant information system changes
in the absence of definitive information. Rogers (1995) did
not discuss this, but others distinguish radical change from
incremental change (Kaluzny et al. 1977; Greer 1981;
Onstrud and Pinto 1991; Orlikowski 1993).
The diffusion of innovation framework, in its application to
organizations, is predicated on the following idea: ‘An
organization is a stable system of individuals who work
together to achieve common goals . . .’ Rogers (1995: 403).
This idea was not borne out here. There was a conflict of
interest amongst individuals within the organization, as international donors requested extra data collection and processing to support their headquarters’ interest, rather than MOH
information management strategies. Using the diffusion of
innovation framework to deepen understanding of the
adoption of new information management strategies in a
management-training package, Gladwin et al. (2002)1
identified that individuals supporting the adoption were
pursuing personal career goals in addition to organizational
ones. Mock et al. (1993) found personal agendas affected
implementation when new information system strategies to
reform the HMIS were introduced in Niger.
The issue of an innovation bringing concomitant changes or
a cluster of innovations is mentioned by Bonair et al. (1989)
who reviewed the transfer of medical technologies to lowincome countries: ‘Transfer of foreign medical technology to
developing countries means not only transfer of drugs and
equipment, but also transfer of a foreign cultural perception
of disease, the so-called western medical paradigm.’ Foltz
(1993) recognizes the problem of defining the innovation,
a new MIS in Chad, when she says this is a complex
Greer (1977: 506) criticizes the diffusion of innovation framework because it does not take into account political theory.
Themes of a political nature here were not explained within
the framework, but they have been displayed graphically as
additional issues in Figure 1. That civil servants pursue
personal rather than organizational goals, constraining
rational decision-making, is not a new concept (Montgomery
1987: 914; Gyimah-Boadi and Rothchild 1990: 52). Waddington (1992) claims that recognizing differences between
The need to consider the organizational context when
changing information systems suggests the process is more
complex than some practitioners have realized when
attempting to understand the causes of information management problems and developing HMIS in low-income countries. Avgerou (1993) also criticizes national development
planning information-system developers for not seeing
organizational change as part of the systems development
process in low-income countries.
zg026 (jr/d)
14/4/03
2:14 pm
Page 221
Implementing a health management information system
private and publicly stated goals was crucial to understanding constraints to rational decision-making at district level in
Ghana.
The Innovation-Process model does not include the idea that
after the adoption decision, rejection or discontinuation
could take place during implementation, although this is
recognized in the Innovation-Decision Process. This research
has shown the relevance of the Innovation-Process model
and that social structure affects not only the decision to adopt
(as the Innovation-Decision process suggests) but also the
implementation process and consequences.
Information system developers and implementers had not
acknowledged they were promoting an informational
approach to management with the change from a centralized
reporting system to an MIS supporting use of information at
the level of collection. In Ghana, Campbell et al. (1996: 15),
reporting on HMIS development, acknowledge their desire
to increase the number of ‘informed decisions’. They
assumed that having more objective, locally collected information would lead to more effective and consistent health
management, but they did not acknowledge a different
management approach was needed. Although they acknowledge appropriate data analysis tools and some management
tools were necessary, there appears to be a lack of management questions or decisions associated with the data, and no
training in management tools. They do not provide the
conceptual framework to link information, management
tools and management. Lippeveld et al. (2000) call for
research which identifies ‘how one could design and conduct
training that effectively improves the actual use of information for health services planning and management’.
Conclusion
This case study followed the HMIS planning process and
explored the evidence by utilizing the diffusion of innovation
and dynamic equilibrium organizational change model,
which guided previous research and practice. Neither
perspective guided HMIS development and implementation
in Uganda. Instead, technological issues rather than wider
organizational issues dominated the planned change to the
HMIS.
The diffusion of innovation and dynamic equilibrium
organizational change models are applicable to the introduction of new information management strategies and management approaches in low-income countries. Some refinements
to the models described by Rogers and Leavitt have been
made, as detailed by Gladwin (1999) and Gladwin and
Wilson (2000). These frameworks can facilitate the practice
of introducing such innovations and enable practitioners to
see the introduction of innovations as a staged process to be
managed. Issues that may facilitate or inhibit adoption may
be identified in advance.
Some implications for practice from this study reinforce
existing guidelines while others are new. Rogers’ (1995)
diffusion framework is relevant to the introduction of new
information management strategies and management
221
approaches, and his 87 generalizations with implications for
practice are also relevant. Table 1 identifies some of the
practical implications of this work; more detail is given in
Gladwin (1999). These ideas can also be used as part of needs
assessment and evaluation as well as introducing information
management innovations. This case study has documented
common barriers to establishing and sustaining effective
routine HMIS and strategies to minimize their effects, a
recent WHO aim (WHO 2001b).
Endnotes
1 This paper is a companion paper to the present one as it
follows the introduction of another information management innovation; the two yield complementary implications for practice.
References
Aga Khan Foundation. 1993. Primary Health Care Management
Advancement Programme. Washington DC/Geneva: Aga Khan
Foundation.
Alder PA, Alder P. 1994. Observational techniques. In: Denzin NK,
Lincoln YS (eds). Handbook of qualitative research. Thousand
Oaks, CA: Sage Publications, pp. 377–92.
Avgerou C. 1993. Information systems for development planning.
International Journal of Information Management 13: 260–73.
Azubuike MC, Ehiri JE. 1999. Health information systems in
developing countries: benefits, problems and prospects. Journal
of the Royal Society for the Promotion of Health 119: 180–4.
Bekui AM. 1990. A health management information system for
district health services in Ghana: improving the current system.
MSc Dissertation, Leeds University, UK.
Bonair A, Rosenfield P, Tengvald K. 1989. Medical technologies in
developing countries: issues of technology development,
transfer, diffusion and use. Social Science and Medicine 28:
769–81.
Braa J, Heywood A, Shung King M. 1997. District level information
systems: two cases from South Africa. Methods of Information
in Medicine 36: 115–21.
Campbell B, Adjei S, Heywood A. 1996. From data to decision
making in health: the evolution of a health management information system. Amsterdam: Royal Tropical Institute.
Campos-Outcalt D. 1991. Microcomputers and health information
in Papua New Guinea: a two-year follow-up evaluation. Health
Policy and Planning 6: 348–53.
Checkland P. 1981. Systems thinking, systems practice. Chichester:
John Wiley and Sons.
Commission on Health Research for Development. 1990. Health
Research: essential link to equity in development. New York:
Oxford University Press.
Davis GB, Olson MH. 1984. Management Information Systems:
conceptual foundations, structure and development. Second
edition. New York, London: McGraw-Hill.
De Kadt E. 1989. Making health policy management intersectoral:
issues of information analysis and use in less developed countries. Social Science and Medicine 29: 503–14.
Denzin NK, Lincoln YS. 1994. Introduction: Entering the field of
qualitative research. In: Denzin NK, Lincoln YS (eds).
Handbook of qualitative research. Thousand Oaks, CA: Sage
Publications, pp. 1–18.
Feacham RG, Graham WJ, Timaeus IM. 1989. Identifying health
problems and health research priorities in developing countries. Journal of Tropical Medicine and Hygiene 92: 132–91.
Finau SA. 1994. National health information systems in the Pacific
Islands – in search of a future. Health Policy and Planning 9:
161–70.
Foltz AM. 1993. Modelling technology transfer in health information systems: learning from the experience of Chad.
zg026 (jr/d)
14/4/03
2:14 pm
Page 222
222
J Gladwin et al.
Table 1. Key implications for introducing information management innovations
Recommendation
Strategies
Improve innovation definition
Develop implied Meaning or Principle Knowledge before introduction and
introduce it before, or at the same time as, Awareness and How-to Knowledge
Clarify whether a radical change or natural extension is implied
Be aware of organizational context and influencing factors
Address compatibility of innovation with existing practice
Ensure alignment of new IS technology by viewing introduction of IM
innovation as issue of organizational change and facilitate alignment of forces
within organization at all stages of process
Focus on information use as well as data collection and processing
Make explicit IM strategies to support informational management approach
and put support strategies in place
Diffusion of innovation framework and organizational change ideas will allow
practitioners to see innovation introduction as staged process to be managed
Efforts to improve HMIS should prioritize conceptual frameworks that
describe health workers’ understanding of factors affecting health status and
are utilized in planning and monitoring
Clarify links between information, management tools and management
Apply an Information System Development Methodology: Soft Systems
Methodology developed by Checkland (1981), and ETHICS, developed by
Mumford (1983) identify techniques for dealing with issues, such as
participation, bottom-up development, stakeholder conflict and powerful
stakeholder domination
Develop national Health Information Management Strategy
Needs assessment, evaluation and monitoring should focus upon data
collection, processing and information use, skill levels and roles performed,
organizational structure, organizational strategies, management tools and
management processes in operation, and health workers’ view of changes
All IM changes need to be monitored and evaluated; failure to adopt a
particular IM strategy may signal inappropriateness
Allow sufficient time for this long process, particularly at planning stage, as
HMIS implementation will proceed slowly, but chance of success will be
higher
Health workers’ training should focus on data processing, analysis and use
of information in their own written/spoken language
In-Charges and DHT must be trained in, and undertake their expected
management role at the same time, or before HMIS implementation
Teaching skills should be an integral part of supervisor’s training
Adequate finance should be made available for support supervision including
transportation
Strategies to overcome lack of IM skills amongst less well-educated health
workers need to be developed, eg DHT may need to temporarily undertake
data processing, and provide newsletter
Change agents should be aware of adoption stage the organization is at to
facilitate process
Opportunity to see the innovation in practice or clearly reported trial should
accompany innovation introduction
Innovation reinvention, redefining or adaptation occurs at different stages in
diffusion process; steps should be taken to facilitate or discourage this,
depending upon the change agents’ aim
Innovations are introduced to individuals after organizational adoption;
diffusion process should be conducted accordingly
Consider various stakeholders’ views, innovation facilitators personal
agendas, likely changes in status and power and other political factors when
designing, facilitating adoption and implementing IS
Conflicts between existing organizational culture and changes need to be
carefully negotiated
HMIS improvement should focus upon utilizing information as well as data
collection and processing
HMIS data and information is for operational management, not broader
evaluation Although some of that data may be of value for researchers,
research will need to conduct special surveys
Understand potential adopter’s situation
Understand IM innovations
involve organizational change
Adopt conceptual frameworks
Utilize existing expertise and research in IS
Broad needs assessment,
monitoring and evaluation
Training
Adopt staged process of innovation diffusion
Understand cultural issues
Develop appropriate IM strategies
and general features
IM = information management; IS = information system; HMIS = Health Management Information System; DHT = District Health Team.
zg026 (jr/d)
14/4/03
2:14 pm
Page 223
Implementing a health management information system
International Journal of Technology Assessment in Health Care
9: 346–59.
Gladwin J. 1999. An informational approach to health management
in low-income countries. Ph.D. Thesis, University of Sheffield,
UK.
Gladwin J, Wilson T. 2000. Validation of a theoretical model linking
organisational fit and diffusion of innovation in information
systems development. Health Informatics Journal 6: 219–27.
Gladwin J, Dixon RA, Wilson TD. 2002. Rejection of an innovation:
health information management training materials in East
Africa. Health Policy and Planning 17: 354–61.
Glaser BG, Strauss AL. 1967. The discovery of grounded theory.
New York: Aldine de Gruyter.
Greer AL. 1977. Advances in the study of diffusion of innovation in
health care organizations. Millbank Memorial Fund Quarterly
55: 505–32.
Greer AL. 1981. Medical technology: assessment, adoption, and utilisation. Journal of Medical Systems 5: 129–45.
Gyimah-Boadi E, Rothchild D. 1990. Ghana. In: Subramaniam V
(ed). Public administration in the third world – an international
handbook. New York: Greenwood Press.
Hedberg B, Mumford E. 1975. The design of computer systems:
Man’s vision of man as an integral part of the system design
process. In: Mumford E, Sakman H (eds). Human choice and
computers. Amsterdam: North Holland, pp. 31–59.
Heywood AB, Campbell BC. 1997. Development of a primary
health care information system in Ghana: lessons learnt.
Methods of Information in Medicine 36: 63–8.
Husein K, Adeyi O, Bryant J, Cara NB. 1993. Developing a primary
health care management information system that supports the
pursuit of equity, effectiveness and affordability. Social Science
and Medicine 36: 585–96.
Iivari J. 1992. The organizational fit of information systems. Journal
of Information Systems 2: 3–29.
Jayasuiriya R. 1999. Managing information systems for health
services in a developing country: a case study using a contextualist framework. International Journal of Information
Management 19: 335–49.
Jeppsson A, Okuonzi SA. 2000. Vertical or holistic decentralization
of the health sector? Experiences from Zambia and Uganda.
International Journal of Health Planning and Management 15:
273–89.
Kaluzny AD, Veney JE. 1977. Types of change and hospital planning
strategies. American Journal of Health Planning 1: 13–19.
Keen PGW. 1981. Information systems and organizational change.
Communications of the ACM 24: 24–33.
Keller A. 1991. Management information systems in maternal and
child health/family planning programs: a multi-country
analysis. Studies in Family Planning 22: 19–30.
Leavitt HJ. 1965. Applied organisational change in industry: structural, technological and humanistic approaches. In: March JG
(ed). Handbook of organisations. Chicago: Rand McNally,
pp. 1144–70.
Leavitt HJ, Dill WR, Eyring HB. 1973. The organisational world.
Orlando, FL: Harcourt Brace Jovanovich, Inc.
Lippeveld T, Sauerborn R, Bodart C. 2000. Design and implementation of health information systems. Geneva: World Health
Organization.
Loevinsohn BP. 1994. Data utilization and analytical skills among
mid-level health programme managers in a developing country.
International Journal of Epidemiology 23: 194–200.
Mills A, Vaughan JP, Smith DL, Tabibzadeh I. 1990. Health system
decentralization: concepts, issues and country experience.
Geneva: World Health Organization.
Ministry of Health. 1993a. The Three-Year Health Plan Frame:
1993/94–1995/96, Uganda. Ministry of Health.
Ministry of Health. 1993b. White Paper on Health Policy Update
and Review (1993). Uganda: Ministry of Health.
Mock N, Setzer J, Sliney I, Hadizatou G, Bertrand W. 1993.
Development of information-based planning in Niger.
223
International Journal of Technology Assessment in Health Care
9: 360–8.
Montgomery JD. 1987. Probing managerial behaviour: image and
reality in Southern Africa. World Development Movement 15:
911–29.
Morse JE. 1994. Designing funded qualitative research. In: Denzin
NK, Lincoln YS (eds). Handbook of qualitative research.
Thousand Oaks, CA: Sage Publications Inc.
Mumford E. 1983. Designing participatively. Manchester: Manchester Business School.
Nabarro D, Annett H, Graham-Jones S, Nabeta E. 1988. Microcomputers in developing country programmes: valuable tools
or troublesome toys? Experience from Uganda and Nepal. In:
Wilson RG, Bryant JJ, Echols BE, Abrantes A (eds). Management information systems and microcomputers in primary
health care. Geneva: Aga Khan Foundation, pp. 41–52.
Newbrander WC, Thomason JA. 1988. Computerizing a national
health system in Papua New Guinea. Health Policy and
Planning 3: 255–9.
Newbrander WC, Thomason JA, Kolehmainen-Aitken R, CamposOutcalt D, Afo G, Davidson W. 1988. Management support for
provinces: a programme for developing provincial health
management capabilities in Papua New Guinea. International
Journal of Health Planning and Management 3: 45–55.
Okuonzi SA. 1994. Information systems development for tropical
disease control and monitoring the impact of MANTEAU.
Uganda Medical Bulletin 1: 1.
Okuonzi SA, Macrae J. 1995. Whose policy is it anyway? International and national influences on health policy development
in Uganda. Health Policy and Planning 10: 122–32.
Onstrud HJ, Pinto JK. 1991. Diffusion of geographic information
innovations. International Journal of Geographical Information
Systems 5: 447–67.
Orlikowski WJ. 1993. Case tools as organisational change: investigating incremental and radical changes in systems development. MIS Quarterly September: 309–40.
Phillips EM, Pugh DS. 1994. How to get a PhD: a handbook for
students and their supervisors. 2nd edition. Buckingham, UK:
Open University Press.
Robey JM, Lee SH. 1990. Information system development in
support of national health programme monitoring and evaluation: the case of the Philippines. World Health Statistical
Quarterly 43: 37–46.
Rogers EM. 1962. Diffusion of Innovations. New York: The Free
Press.
Rogers EM. 1995. Diffusion of Innovations. 4th edition. New York:
The Free Press.
Sandiford P, Annett H, Cibulskis R. 1992. What can information
systems do for primary health care? An international perspective. Social Science and Medicine 34: 1077–87.
Sandiford P, Kanga GJ, Ahmed AM. 1994. The management of
health services in Tanzania – a plea for health sector reform.
International Journal of Health Planning and Management 9:
295–308.
Sapirie SA, Orzeszyna S. 1995. Selecting and defining national health
indicators. Strengthening Country Health Information Unit,
Division of Epidemiological Surveillance and Health Situation
and Trend Assessment. Geneva: World Health Organization.
Accessed 1 January 2002 at [http://www.who.int/healthservices-delivery/information/20000629d.htm].
Scott Morton MS. 1991. Introduction. In: Scott Morton MS (ed). The
corporation of the 1990s. New York: Oxford University Press.
Smith DL, Hansen H, Karim MS. 1988. Management information
support for district health systems based on primary health
care. In: Wilson RG, Bryant JH, Echols BE, Abrantes A (eds).
Management information systems and microcomputers in
primary health care. Geneva: Aga Khan Foundation,
pp. 89–110.
Van Hartevelt JHW. 1993. Information management in international development as an area for information services with
zg026 (jr/d)
224
14/4/03
2:14 pm
Page 224
J Gladwin et al.
a case in the field of health care in Ghana. International Forum
on Information and Documentation 18: 32–6.
VanNorren B, Boerma JT, Sempebwa EK. 1989. Simplifying the
evaluation of primary health care programmes. Social Science
and Medicine 28: 1091–7.
Villadsen S. 1996. Decentralization and governance. In: Villadsen S,
Lubanga F (eds). Democratic Decentralisation in Uganda.
Kampala: Fountain Publishers.
Waddington CJ. 1992. Health economics in an irrational world – the
view from a regional health administration in Ghana. Ph.D.
Thesis, Liverpool School of Tropical Medicine, UK.
WHO. 1988. The challenge of implementation: district health systems
for primary health care. Geneva: World Health Organization.
WHO. 1994a. Implementation of the global strategy for health for all
by the year 2000. Vol.2, Africa Region. Second evaluation. Eight
report on the world health situation. WHO Regional Office for
Africa. Brazzaville: World Health Organization.
WHO. 1994b. Information support for new public health action at
district level. Report of a WHO Expert Committee. WHO Technical Report Series 845. Geneva: World Health Organization.
WHO. 1999. Health services delivery. Interregional consultation on
WHO support to health information system development as
part of health systems development: Kuala Lumpur, Malaysia,
23–27 August 1999. Geneva: World Health Organization.
Accessed 1 January 2002 at [http://www.who.int/healthservices-delivery/information/WHO_EIP_OSD_99.2.htm].
WHO. 2000a. Health services delivery: WHO cooperation in
strengthening national health information systems. Geneva:
World Health Organization. Accessed 30 December 2001 at
[http://www.who.int/health-services-delivery/information/
20000629a.htm].
WHO. 2000b. Investing in evidence and information base: a WHO
strategy to support member states. Health Information System
Development as part of Health Systems Development: a report
of an informal consultative meeting 11–12 May 2000 WHOHQ, Geneva. Geneva: World Health Organization. Accessed
30 December 2001 at [http://www.who.int/health-servicesdelivery/information/20000629b.htm].
WHO. 2000c. Health information systems development and
strengthening: guidance for national health information
systems development. Geneva: World Health Organization.
Accessed 30 December 2001 at [http://whqlibdoc.who.int/hq/
2000/WHO_eEIP_OSD_00.6.pdf].
WHO. 2001a. Health services delivery: the development of health
information systems. Geneva: World Health Organization.
Accessed 30 December 2001 at: [http://www.who.int/healthservices-delivery/WHO/information/his_development.htm].
WHO. 2001b. Health services delivery: health information systems
development recent activities. Geneva: World Health Organization. Accessed 30 December 2001 at [http://www.who.int/
health-services-delivery/information/recent.htm].
Wilson R, Echols BE, Smith D, Bryant JH. 1988. Fresh approaches
and new tools to improve management of health information
systems based on primary health care. In: Wilson RG, Bryant
JH, Echols BE, Abrantes A (eds). Management information
systems and microcomputers in primary health care. Geneva:
Aga Khan Foundation, pp. 147–62.
World Bank. 1993. World Development Report 1993: Investing in
health. New York: Oxford University Press.
Acknowledgements
The researchers would like to express their thanks to MOH personnel and other health workers in Uganda for their time during this
research. We are also grateful to the British Council for funding the
visit of one of the researchers (JG) as part of a link programme for
strengthening district management.
Biographies
Jean Gladwin, PGCE, MSc, RPHNutr, Ph.D., completed her doctorate at the Centre for Health Management Information Research,
University of Sheffield. The thesis (from which this case study is
taken) investigated the introduction of new information management strategies intended to promote an informational approach to
management at the operational health service level in low-income
countries. She is also a Public Health Nutritionist having worked in
Europe, East Africa and South East Asia as a practitioner and
researcher. She is presently focusing on the use of death and mortality information for management and policy in the UK.
Robert A Dixon, BSc, Ph.D., is a NHS Research Manager and independent consultant in International Health, formerly Senior
Lecturer in Public Health at the University of Sheffield. Doctoral
work was on local hospital and primary care health information
systems. He has worked in India, Indonesia, Cambodia and Bosnia,
and is currently evaluating the health programmes of Africa Inland
Mission. [Address: Sheffield Health and Social Research Consortium, Fulwood House, Old Fulwood Road, Sheffield S10 3TH, UK]
Professor TD Wilson, Ph.D., is Professor Emeritus in Information
Management in the Department of Information Studies, University
of Sheffield, where he has been since 1972 and was Head of Department for 15 years. He is also Visiting Professor at the Hogskolan I
Boras, Sweden and Academic Director of the International Centre
for Information Management Systems and Services at Nicholas
Copernicus University, Torun, Poland. His research interests are in
information behaviour, electronic scholarly communication, and the
strategic aspects of applications of information technology in
organizations. [Address: Department of Information Studies,
University of Sheffield, Western Bank, Sheffield S10 2TN, UK]
Correspondence: J Gladwin, HSRU, London School of Hygiene and
Tropical Medicine, Keppel Street, London WC1E 7HT, UK. Email:
[email protected]