Changing malaria intervention coverage, transmission and hospitalization in Kenya

Emelda A Okiro 0 1 Victor A Alegana 0 Abdisalan M Noor 0 1 Robert W Snow 0 1 0 Malaria Public Health & Epidemiology Group, Centre for Geographic Medicine Research - Coast, Kenya Medical Research Institute/Wellcome Trust Research Programme , P.O. Box 43640, 00100 GPO, Nairobi , Kenya 1 Centre for Tropical Medicine, Nuffield Department of Clinical Medicine, University of Oxford, CCVTM , Oxford OX3 7LJ , UK Background: Reports of declining incidence of malaria disease burden across several countries in Africa suggest that the epidemiology of malaria across the continent is in transition. Whether this transition is directly related to the scaling of intervention coverage remains a moot point. Methods: Paediatric admission data from eight Kenyan hospitals and their catchments have been assembled across two three-year time periods: September 2003 to August 2006 (pre-scaled intervention) and September 2006 to August 2009 (post-scaled intervention). Interrupted time series (ITS) models were developed adjusting for variations in rainfall and hospital use by surrounding communities to show changes in malaria hospitalization over the two periods. The temporal changes in factors that might explain changes in disease incidence were examined sequentially for each hospital setting, compared between hospital settings and ranked according to plausible explanatory factors. Results: In six out of eight sites there was a decline in Malaria admission rates with declines between 18% and 69%. At two sites malaria admissions rates increased by 55% and 35%. Results from the ITS models indicate that before scaled intervention in September 2006, there was a significant month-to-month decline in the mean malaria admission rates at four hospitals (trend P < 0.05). At the point of scaled intervention, the estimated mean admission rates for malaria was significantly less at four sites compared to the pre-scaled period baseline. Following scaled intervention there was a significant change in the month-to-month trend in the mean malaria admission rates in some but not all of the sites. Plausibility assessment of possible drivers of change pre- versus post-scaled intervention showed inconsistent patterns however, allowing for the increase in rainfall in the second period, there is a suggestion that starting transmission intensity and the scale of change in ITN coverage might explain some but not all of the variation in effect size. At most sites where declines between observation periods were documented admission rates were changing before free mass ITN distribution and prior to the implementation of ACT across Kenya. Conclusion: This study provides evidence of significant within and between location heterogeneity in temporal trends of malaria disease burden. Plausible drivers for changing disease incidence suggest a complex combination of mechanisms, not easily measured retrospectively. - Background Reports of declining incidence of malaria hospitalization, deaths and prevalence in several, diverse areas of Africa suggest that the epidemiology of malaria is in transition across the continent [1-17]. This transition has been coincidental with the scaling of intervention coverage and increased international funding for the control of malaria. However, precise attribution to intervention coverage alone remains circumstantial, not least because at most sites where declining disease incidence and prevalence have been reported, declines started before significant increase in donor assistance and scaled intervention coverage [2,8,9,12-14]. Adjusted trends in inpatient paediatric malaria case burden over 10 years in a sample of 17 hospitals with varied malaria transmission ecologies across Kenya were recently described using population-adjusted clinical data from defined hospital catchment areas [17]. The results showed divergent temporal patterns of disease incidence between sites. Importantly these data signalled that all was not equal across a single country. Here the possible mechanisms underlying the differences between these sites are explored in more detail in an effort to explain plausible drivers for changing disease incidence. Methods and Results Overview of sample selection, time-periods and methodological approach For the purposes of evaluating mechanisms for change, hospital settings with adequate temporal data on insecticide-treated net (ITN) coverage (intervention change) and parasite prevalence (transmission intensity change) within defined spatial areas around hospitals that serve as the catchment to these hospitals were selected. Of the 17 original hospital settings the final selection of eight hospitals covered the dominant malaria ecologies that characterize Kenya including: three hospitals Bungoma, Kisumu, and Siaya District General Hospitals (DGH) in the Western/lakeside high transmission areas; Kericho and Kisii DGH located in the Highlands with typically epidemic transmission; and the three sites in Coastal Kenya: Malindi, Kilifi and Msambweni DGH. Data have been assembled across two three-year time periods: September 2003 to August 2006 and September 2006 to August 2009: corresponding to important timelines for malaria intervention and drug policy change in Kenya. A range of plausible drivers and effect modifiers of changes in disease incidence were identified including intervention coverage, rainfall, service use and malaria transmission intensity. These data have been assembled at different spatial resolutions to match the communities served by the hospitals used to define disease incidence. Defining attribution is fraught with many challenges and plausibility designs are often regarded as the only feasible option to evaluate the impact of nationally promoted large-scale intervention or changing risks [18,19]. These approaches have been recently promoted as a means of examining changes consequent upon scaled malaria control effort in sub-Saharan Africa [20]. Initially, temporal changes in factors that might explain changes in disease incidence were sequentially examined for each hospital setting, and were compared between hospital settings and plausible explanatory factors were ranked relative to proportional changes in disease incidence. An analytical procedure called intervention analysis was then used, this allows some exogenous event, in this case increasing ownership of ITNs and or a change in the first-line treatment policy to ACT, to occur that would affect the behaviour of the time-series of disease incidence being modelled as an ARMA process using the segmented regression [21,22]. Defining the geographic scope of attribute data - hospital catchment populations High resolution census data were used to produce population distribution estimates around each hospital and combined with a random sample of admissions where residence was defined to compute distance travelled from their homes to the hospital. Thematic maps were created in ARCGIS 9.1 (ESRI, Inc., Redland, CA (...truncated)