South Africa's District Health Information System: Case Study from Eastern Cape Province

Randy Wilson¹, Jon Rohde², Ruben Puchert³

Abstract:

This case study describes efforts to support improvements to health services in South Africa's Eastern Cape province through the enhancement of the province's Health Information System. It describes the context of health systems reform in which these enhancements were made, key characteristics of the information system, the process undertaken by many players at different levels of the health systems, and identifies a variety of lessons learned that should prove useful for similar efforts in other countries.

One of the most interesting aspects of the work highlighted in this case study is the fact that the system was introduced into a rapidly changing health care system -- and the system's implementation has helped both to monitor that change and to shape the change process. During discussions to gather information for this case study many examples of this inter-relationship were discovered. To cite just a few: district health workers have been using the system to help monitor priority health problem trends and the coverage of services such a EPI, TB and STD contact tracing in specific geographic areas. This has allowed allocation of increased resources to those facilities lagging behind. Analysis of certain indicators, such as workload, have assisted in the reallocation of staff and the determination of new locales for introduction of health services. Tracking selected drug stock-outs monthly has resulted in dramatic improvement in essential drug availability at clinics. At the same time, the introduction of the system has led to the determination of new functions and staffing requirements to support data management at the district and regional levels.

For public health professionals and multi-disciplinary teams involved in developing health management information systems, the Eastern Cape experience is rich in lessons learned and innovation. In particular,

District-level involvement in HIS development is not only possible, but it is critical for the success of information systems aimed at enhancing health services at the district level. At the same time it is crucial to identify and involve all stakeholders or 'actors' in the process. This includes individuals at all levels of the health system (potential users) as well as significant events and changes that are likely to influence the development of the system. It was important to focus on high level staff as well. Only when the information officers started making reports and putting them on managers' desks did anyone pay much attention to what the data were telling them.
It is important to 'walk before you try to run with information systems.' Considerable effort was focused on helping staff use paper and pencil to master analytical skills before the system was computerised. This required a minimum data set that health workers could easily understand and analyse on the spot -- without the need of a computer. Similarly, the use of hand-drawn, catchment area maps and graphs was strongly encouraged as a key step in getting people to understand the populations they serve and to believe the numbers. Later the computer helped them do it faster and better.
Many lessons were learned in the process of developing and implementing the DHIS software, including: problems with the initial development efforts without professional software developer support, successes with links to off-the-shelf software like Excel, underestimating the appeal of the system and thereby the size of data files and hardware and software requirements, coming up with the right balance between precision (read complexity) and approximation, as well as some difficulties in supporting a large installed base of freely distributed, open-source software with limited resources.
The implementation of new software can become a vehicle for change, including: standardisation, integration, and more widespread dissemination of health data/information. This experience once again underlines the chicken/egg nature of technical innovation in a health system. Although this is somewhat the corollary of the 'walk don't run' principle cited above, in this case it was also true that the process of computerisation served as a catalyst for change. By requiring managers to prioritise what information was really necessary and to question standard operating procedures for data collection and use.
The indicator-based approach to determining information needs is an effective way to reduce the number of data elements collected and to ensure that data collected is relevant. A simplified indicator framework such as WHO's health problem, service and critical resource indicator categories, would have been useful to ensure a more even distribution of indicator types and to clarify the linkages between them.
Improvements in data recording procedures deserve as much attention as reporting and data processing. In the Eastern Cape, focus on this aspect came very late in the process. Problems of time spent recording data, poor data quality and limited use of the data at local level are likely to continue unless more emphasis is placed on this area.
In spite of the ease with which data can be entered and analysed with the support of the computer, it has been very hard to get people to make timely reports. They prefer annual reporting -- which some feel is HISTORY. The system is most useful if it can provide fast and up to date indicators of what the problems are and where they are localised.

1 Deputy Chief Information Officer, MSH Boston

2 Technical Advisor, EQUITY Project, MSH, Bisho, Eastern Cape Province

3 Provincial Health Information Systems Officer, Bisho, Eastern Cape Province

To complete paper