Assessment of the surveillance system for diseases of epidemic potential in Rukungiri District, Uganda.

Abstract

BACKGROUND: During a workshop held on 4th May 2000, DHT members in Rukungiri district expressed concern over the various problems facing the surveillance system. The magnitude of these problems and the factors responsible for them were unknown. This made it necessary to carry out an in-depth assessment of the district surveillance system. OBJECTIVES: The study aimed at assessing the performance of the surveillance system so as to design appropriate for improving it. Specifically the study assessed the availability of resources, the process of collection of surveillance data, its management and utilization for action and finally identified the factors affecting disease surveillance. MATERIALS AND METHODS: A descriptive cross sectional study was carried out at the district health office and 25 health units, selected by stratified random sampling. A total of 25 health unit in-charges, 11 records assistants, 21 laboratory workers and 102 other health workers were interviewed. Observations were made for surveillance resources and evidence of data analysis. Surveillance reports for 2000 were reviewed for their accuracy, timeliness and completeness. Ten key informant interviews and 4 focus group discussions were also conducted. RESULTS: Only 12% of all the in-charges of health units and 5% of other health workers had received some training specifically in surveillance of epidemic prone disease, while none of the records assistants had received any such training. Only 8% of the health units had standard case definitions and none had action threshold. There was 0-30% agreement between the monthly reported malaria cases and the actual cases in the registers. The type of person who carried out tallying significantly affected the accuracy of the reports (p fisher exact = 0.05). Evidence of data analysis was in 64% of the health units. The training status of an in-charge in HMIS had a significant effect on whether a health unit carried out data analysis (p fisher exact = 0.01) or utilized data for action (p fisher exact = 0.04). The major factor affecting disease surveillance was the poor knowledge and skills of the health workers and records assistants. CONCLUSIONS AND RECOMMENDATIONS: There were few personnel who were trained in disease surveillance. Detection of epidemic diseases was constrained by lack of standard case definitions in the majority of health units. There was a high level of inaccurate reporting by the majority of health units, however, they reported better with timeliness and completeness of reporting. Though the majority of health units analyzed surveillance data, this was limited in scope. The data was widely used at the collecting units but the lack of action thresholds limited its use in prediction or detection of epidemics. It is recommended that the DHT should organize refresher courses for all the personnel involved in disease surveillance and give them more frequent support supervision and feedback. The DDHS should provide standard case definitions and action thresholds to all health units in the district and train health workers and records assistants in their use.