Defining malaria burden from morbidity and mortality records, self treatment practices and serological data in Magugu, Babati District, northern Tanzania

Abstract

Malaria morbidity and mortality data from clinical records provide essential information towards defining disease burden in the area and for planning control strategies, but should be augmented with data on transmission intensity and serological data as measures for exposure to malaria. The objective of this study was to estimate the malaria burden based on serological data and prevalence of malaria, and compare it with existing self-treatment practices in Magugu in Babati District of northern Tanzania. Prospectively, 470 individuals were selected for the study. Both microscopy and Rapid Diagnostic Test (RDT) were used for malaria diagnosis. Seroprevalence of antibodies to merozoite surface proteins (MSP-119) and apical membrane antigen (AMA-1) was performed and the entomological inoculation rate (EIR) was estimated. To complement this information, retrospective data on treatment history, prescriptions by physicians and use of bed nets were collected. Malaria prevalence in the area was 6.8% (32/470). Of 130 individuals treated with artemisinin combination therapy (ACT), 22.3 % (29/130) were slide confirmed while 75.3% (98/130) of them were blood smear negative. Three of the slides confirmed individuals were not treated with ACT. Fever was reported in 38.2% of individuals, of whom 48.8 % (88/180) were given ACT. Forty-two (32.3%) of those who received ACT had no history of fever. About half (51.1%) of those treated with ACT were children <10 years old. Immunoglobulin against MSP-119 was positive in 16.9% (74/437) while against AMA-1 was positive in 29.8 % (130/436). Transmission intensity was estimated at <0.2 infectious bites per person per year. The RDT was highly specific (96.3%) but with low sensitivity (15.6%).  Magugu is a low endemic area. There is substantial over diagnosis, over treatment and self treatment in the community.   The burden of malaria based on medical records is over estimated   as was mostly presumptive. The low sensitivity of RDT reflects the low number of immune individuals as well as the low parasite density.