Epidemiology of Trypanosoma brucei rhodesiense sleeping sickness in Eastern Uganda

Sleeping sickness is now recognized as one of the recent re-emerging but neglected diseases. In Uganda and the rest of sub-Saharan Africa, old foci of sleeping sickness persist despite donor-funded projects and there is evidence of expansion of these foci. The World Health Organization acknowledges a high level of undiagnosed sleeping sickness cases and has identified surveillance and treatment as the major control strategy for sleeping sickness. However, to date there have been no clear epidemiological surveillance markers applied for monitoring the burden of the disease at control implementation levels. There is therefore a need to obtain markers for surveillance and to apply these to monitor and improve this important strategy of sleeping sickness control, hence, the objective of this study.Data of sleeping sickness patients from 1987 to 2002 were collected from Tororo, Busia and Soroti districts. Using the ratio of early to late stage patients in an affected population, data from (Livestock Health Research Institute) LIRI hospital in Tororo district were applied to develop a model for estimation of sleeping sickness underdetection. The model was applied to data from a recent outbreak of sleeping sickness in Soroti district to determine the burden of the disease. Geographical information systems (GIS) were applied to quantify the effect of distance from the health units on early detection of sleeping sickness. A semi-structured questionnaire was administered to prospective sleeping sickness patients to determine their treatment seeking behavior. Spatial and temporal trends of sleeping sickness with and without control are described using the SatScan method for spatial temporal clustering. The potential role of remote sensing in surveillance of sleeping sickness was explored using a processed Landsat 7 series satellite image to determine association of land cover classes, vegetation indices, population density, distance to wetlands, distance to the sleeping sickness hospital to the outcome of village scale disease geographical distribution generated using GIS.In the Tororo and Busia area, for every 3 reported patients, 2 went undetected. The Disability Adjusted Life Years (DALYs) for sleeping sickness in the Soroti outbreak are greatly increased when under-detection is taken into account. There is a strong effect of distance to health unit on the surveillance of sleeping sickness. Children and xviii adolescents are less likely to be detected in early stage. During times of low incidence, the proportion of late stage detected patients increases. Most of the delay to diagnosis is attributable to service provider delay and most patients are referred to the sleeping sickness referral hospital by the community and not the health system. Proximity to wetlands and to the sleeping sickness hospital, low population density, vegetation wetness and smaller area of short grassland and larger area of cropland were significantly associated with the distribution of sleeping sickness affected villages in Tororo and Busia districts.Surveillance of sleeping sickness can be monitored by applying the early to late stage ratio of sleeping sickness as an epidemiological tool to determine the level of underdetection and subsequently the burden of disease. Optimal distribution of fixed post diagnostic units will greatly improve rhodesiense sleeping sickness surveillance. Surveillance programmes should regularly update health workers in the primary health care units in the vicinity of sleeping sickness affected populations. Similarly, the affected communities can be resourcefully used to refer sleeping sickness suspects to the appropriate health care facilities. Timely responsive disease surveillance and intervention can prevent spread of sleeping sickness. Remotely sensed maps can assist in accurately describing sleeping sickness foci and therefore, the area requiring application of control interventions