Identification of factors associated with residual malaria transmission using school-based serological surveys in settings pursuing elimination.

BACKGROUND: Targeted research on residual malaria transmission is important to improve strategies in settings pursuing elimination, where transmission reductions prove challenging. This study aimed to detect and characterize spatial heterogeneity and factors associated with Plasmodium falciparum infections and exposure, P. falciparum apical membrane antigen 1 (PfAMA1) antibody (Ab) response, in the Central Highlands of Madagascar (CHL). METHODS: From May to July 2014, a cross-sectional school-based survey was carried out in 182 fokontany (villages) within 7 health districts of the CHL. Rapid diagnostic tests (RDTs) and a bead-based immunoassay including PfAMA1 antigen biomarker were used to estimate malaria prevalence and seroprevalence, respectively. Local Moran's I index was used to detect spatial "hotspots". Remotely sensed environmental data-temperature, vegetation indices, land covers, and elevation-were used in multivariable mixed-effects logistic regression models to characterize factors associated with malaria infection and cumulative exposure. RESULTS: Among 6,293 school-children ages 2-14 years surveyed, RDT prevalence was low at 0.8% (95% CI 0.6-1.1%), while PfAMA1 Ab seroprevalence was 7.0% (95% CI 6.4-7.7%). Hotspots of PfAMA1 Ab seroprevalence were observed in two districts (Ankazobe and Mandoto). Seroprevalence increased for children living > 5 km from a health centre (adjusted odds ratio (OR) = 1.6, 95% CI 1.2-2.2), and for those experiencing a fever episode in the previous 2 weeks (OR 1.7, 95% CI 1.2-2.4), but decreased at higher elevation (for each 100-m increase, OR = 0.7, 95% CI 0.6-0.8). A clear age pattern was observed whereby children 9-10 years old had an OR of 1.8 (95% CI 1.2-2.4), children 11-12 years an OR of 3.7 (95% CI 2.8-5.0), and children 13-14 years an OR of 5.7 (95% CI 4.0-8.0) for seropositivity, compared with younger children (2-8 years). CONCLUSION: The use of serology in this study provided a better understanding of malaria hotspots and associated factors, revealing a pattern of higher transmission linked to geographical barriers in health care access. The integration of antibody-assays into existing surveillance activities could improve exposure assessment, and may help to monitor the effectiveness of malaria control efforts and adapt elimination interventions.

School-Based Serosurveys to Assess the Validity of Using Routine Health Facility Data to Target Malaria Interventions in the Central Highlands of Madagascar.

BACKGROUND: In low-malaria-transmission areas of Madagascar, annual parasite incidence (API) from routine data has been used to target indoor residual spraying at subdistrict commune level. To assess validity of this approach, we conducted school-based serological surveys and health facility (HF) data quality assessments in 7 districts to compare API to gold-standard commune-level serological measures. METHODS: At 2 primary schools in each of 93 communes, 60 students were randomly selected with parents and teachers. Capillary blood was drawn for rapid diagnostic tests (RDTs) and serology. Multiplex bead-based immunoassays to detect antibodies to 5 Plasmodium falciparum antigens were conducted, and finite mixture models used to characterize seronegative and seropositive populations. Reversible catalytic models generated commune-level annual seroconversion rates (SCRs). HF register data were abstracted to assess completeness and accuracy. RESULTS: RDT positivity from 12 770 samples was 0.5%. Seroprevalence to tested antigens ranged from 17.9% (MSP-1) to 59.7% (PF13). Median commune-level SCR was 0.0108 (range, 0.001-0.075). Compared to SCRs, API identified 71% (95% confidence interval, 51%-87%) of the 30% highest-transmission communes; sensitivity declined at lower levels. Routine data accuracy did not substantially affect API performance. CONCLUSIONS: API performs reasonably well at identifying higher-transmission communes but sensitivity declined at lower transmission levels.