1.2. Information Systems in District Health Systems

Recognizing the manifold responsibilities of the district health team and the need to carefully allocate the scarce resources, the necessity of health management based on adequate information use becomes obvious. It is therefore not surprising, that the "main constraint [for implementing the primary health care approach] reported by practically all countries is inadequate information for the managerial process" ([WHO (1987b)]).

1.2.1. Current Deficits of Information Systems in District Health Systems

The reasons for the lack of adequate information are many. Comprising the results of an international workshop on information systems in primary health care, WILSON lists four major problems in [WILSON (1989)]:

1. The utility of data is not appreciated.
2. Program strategies and therefore monitoring indicators as a basis for information systems are not sufficiently defined.
3. Data are inadequately and inefficiently integrated into the managerial process regarding collection, processing, analysis, and use of data.
4. Microcomputer and information technologies are not sufficiently used.

This synopsis is of course an abstraction of various practical problems: often stressed is the time consuming process of data collection at the health unit level. MOIDU found an average of nearly 50% working time spent on collecting and collating data in a multi-site study ([MOIDU, WIGERTZ and TRELL (1991)]). WILSON cites a scope of 40% up to 60% from different studies in [WILSON (1989)]. Also, data collection is often not coordinated with the result of a lot of different forms and registers at the health unit level, which often contain duplicated data ([MOIDU, WIGERTZ and TRELL (1991)]). As a natural consequence, the staff at the health unit level is tempted to do incomplete and incorrect data collection. The feeling that data are only collected for upward reporting and do not have any impact on the daily activities only support this tendency ([WILSON (1989)]). Under the impression of a study conducted in 1982 and 1983 in a not specified Latin American country, SCHWEFEL states that due to these factors "almost 80% of the forms and data cannot be used meaningfully even if they were to be processed" ([SCHWEFEL (no Year)]).

1.2.2. Definition of Information Systems in District Health Systems

The complex nature of a district health system brings along a complex nature of the involved information system. In this context, parallels to information systems in hospitals in industrialized countries can be drawn. Adapting the definition of a hospital information system from [WINTER and HAUX (1995], an information system in a district health system can be defined as the partial system of a district health system, which is dealing with the complete information processing and information storing of the district health system. The information system has to make available

• the right information and the right knowledge
• to the right persons and institutions
• in the right form
• at the right time
• in the right place. (adapted from [Winter and Haux (1995)])

This broad definition comprises the various expectations in information systems for district health systems. The scope of the needed information ranges from mortality and morbidity data over stock control data to knowledge on how to prevent certain diseases. The right persons and institutions include for example the members of the district health team, the ministry of health, the donor agencies which promote vertical programs and also the communities and individual clients of a health service. The right form for a morbidity report to the ministry of health might be just a computer printed list, while donor agencies require a much more complex impact evaluation. On the other hand, the right form for health education requires a very informal, but culturally embedded and didactic sound presentation of health issues. The right time for information about an upcoming cholera epidemic differs from the right time for sporadic oral health education seminars. Looking at the right places, it is evident that transport plays an essential role in information systems.

In literature, the terms "Health Information System" and "Health Management Information System" are used alternatively for information systems in district health systems, while the latter is becoming more popular. Both terms basically refer to the same thing. The term "Health Information System" tends to be used with more impact on the nationally organized health statistics. The term "Health Management Information System" is often used with the intention, to broaden the view towards the management impact of an information system. This term also values the need for resource management, logistics and staff information. Since in this thesis focus is put on information processing directly related to health, the term "Health Information System" shall be applied. This does not exclude the integration into the managerial process.

1.2.3. Sources of Information in District Health Systems

In order to guarantee the availability of information, an information system has to obtain information first. Pacey identified three possible information sources for health services management in [PACEY (1982)]:

• Service data and statistics of a routine reporting system traditionally provide most of the used data. This kind of information upraising is often criticized since it does not provide accurate information on the real target population, but only on the clients of a health service ([SMITH, HANSEN and KARIM (1988)]). Since these reporting systems were introduced from the national level, they also tend to be self-sufficient entities for upward reporting, which are rarely used at the health unit level ([REYNOLDS (1988)], [MOIDU (1993)]).
• Soundings and reconnaissance at the community level provide a cheap and broad based supplement to service data. They are a qualitative tool for a flexible observation of the health system. However, this informal approach is not suited for an automated routine processing of the results.
• Conventional surveys and research provide formalized qualitative results. They contribute little to service management and are expensive and difficult to conduct. As a simplified survey variant, rapid community surveys are proposed in [WILSON and SAPANUCHART (1993)]. Rapid surveys are a fast and inexpensive alternative to conventional surveys. But although being simplified, they still contain potential sources for biases and statistical inconsistence when they are not conducted properly.

In an information system for district health systems these different sources must be seen as complementary. Rapid surveys could for example be used to validate alarming figures from the routine reporting system, while soundings provide a measure to assess the causes.

1.2.4. Computers as a Tool for Information Systems in District Health Systems

Promoting factor of information sciences in industrialized countries is undoubtedly the advancement of computer technology. The impact of computer technology on every day life as well as on the economic setup in many countries is great enough, that not only zeitgeist magazines talk about an transformation from industrial societies to information societies. The question necessarily arises, whether computer technology is appropriate for countries, which are still far from the border line between an agrarian and an industrial society, or to look at the very extremes, which contain peoples that do not even have the characteristics of an agrarian society jet, such as the Batwa and the Karimojong in Uganda.

It has to be acknowledged, that a few developing countries are already engaged in the production of computer hard- and software ([MANDIL (1993)]). Delegating software development to programmers in developing countries as a measure of cost reduction is already being practiced by western companies and discussed on congresses such as the joint annual conference of the Swiss and the German Informatics Societies ([GISI (1995)]). However, no matter whether these efforts provide a chance for sustainable development or form a new instance of exploitation, countries in sub-saharan Africa are not involved. Here the installation of satellite telephone systems and the upcoming of FM radio stations is already celebrated with headlines like "Ghana Races down Information Highway" ([AFRANI (1995)]).

In this context it is surprising, that the available literature on health informatics in developing countries without exception favours the introduction of computers into health information systems. Already in [HELFENBEIN (1987)] it is stressed, that a "full microcomputer system, including spare parts, programs and a maintenance contract, is cheaper than a landrover". Up to date, microcomputers have been installed in district health systems in a variety of countries. Based on his own experience with over 150 computers working under his general supervision, BERTRAND concludes that the "microcomputer is an appropriate technology whose benefits far outweigh its costs in information intensive activities" ([BERTRAND (1988)]).

While on the one hand it is often emphasized, that the introduction of microcomputers must be preceded by a careful analysis of information needs, on the other hand "the introduction of a microcomputer can itself be a powerful motivator for change and improvement in the information system, partly on the basis of its technical demands for clarity of thought and detailed specification of procedures, and partly because of its glamour and motivational value" ([SMITH, HANSEN and KARIM (1988)]).

However, despite the enthusiastic optimism which is reflected in all the publications cited in this section, it may not be forgotten, that the introduction of computers has to recognize some limiting factors. According to [SMITH, HANSEN and KARIM (1988)], the following factors have to be considered:

• The state of development of the country and its health services, including the availability and use of microcomputers at higher levels within the health system.
• The general level of computer use as well as computer literacy and availability within the country.
• The availability of hardware maintenance as well as training and support capability within the district.
• The capability to support recurrent costs, such as paper, supplies, diskettes, etc., at an adequate level to maintain operations. This also includes the availability of electrical power supply.

For use on a wider scale:

• The availability of foreign exchange within the country and national priorities for its use.
• The questions of creating long-term dependency on external suppliers for both hard- and software.

The reasons for "computer graveyards", which were informally reported to the author of this thesis can probably be seen in a neglection of these factors.

1.2.5. Approaches towards an Improvement of Information Systems in District Health Systems

The 1987 report of the World Health Organization ([WHO (1987b)]) triggered a series of single standing publications on health information systems and related topics. Some of these are [VAN NORREN, TIES BOERMA and SEMPEBWA (1989)], [VILNIUS and DANDOY (1990)], [BANOUB (1990)], [GARNER, HARPHAM and ANNETT (1992)] and [HUSEIN et al. (1993)].

A driving force in this area is undoubtedly the Aga Khan Foundation. Being the co-initiator of UNICEF for the development of [HELFENBEIN (1987)], the Aga Khan Foundation was one of the organizers of the 1987 workshop on management information systems in Lisbon ([WILSON et al. (1988)]). This conference was a melting pot for various practical experiences in the area of information systems with an impact on computer support. The results of this conference formed the basis for the development of the Primary Health Care Management Advancement Programme ([WILSON and SAPANUCHART (1993)]). The program is a series of nine modules and some example computer programs as a practical guideline on how to improve information support in a district health system.

MOIDU signs responsible for a series of studies, which aim at the construction of a common application software design, specialized for the implementation at a primary health care center ([MOIDU, WIGERTZ and TRELL (1991)], [MOIDU, WIGERTZ and TRELL (1992)], [SINGH et al. (1992)]). He points out that the construction of such an application software is feasible.

A meeting ground for health information experts in the African context was established by the first international conference on health informatics in Africa, HELINA 93 ([MANDIL et al. (1993)]), which was mainly promoted by the International Medical Informatics Association (IMIA). This conference did not only cover the primary health care sector, but also the broad scope of health informatics. The IMIA also hosts a working group on primary health care informatics.

In addition to these more or less experience sharing conferences, there are also some efforts from various governments to streamline the national reporting system. The author of this thesis was surprised, that not only in Uganda, where the methodology of this thesis was realized, the national health information system is being restructured. On an informal visit at the Statistics and Health Services Research Team of the Ethiopian Ministry of Health, similar activities were found to simplify the currently very complex national reporting system. Reynolds reports the same from Burma, Bangladesh and Thailand ([REYNOLDS (1988)]).

Despite these efforts, a commonly available and user-friendly software tool for handling routine data, which can be adapted to the needs of a district health team, is not existent yet.