Reactive focal drug administration associated with decreased malaria transmission in an elimination setting: Serological evidence from the cluster-randomized CoRE study

Abstract

Efforts to eliminate malaria transmission need evidence-based strategies. However, accurately assessing end-game malaria elimination strategies is challenging due to the low level of transmission and the rarity of infections. We hypothesised that presumptively treating individuals during reactive case detection (RCD) would reduce transmission and that serology would more sensitively detect this change over standard approaches. We conducted a cluster randomised control trial (NCT02654912) of presumptive reactive focal drug administration (RFDA-intervention) compared to the standard of care, reactive focal test and treat (RFTAT-control) in Southern Province, Zambia-an area of low seasonal transmission (overall incidence of ~3 per 1,000). We measured routine malaria incidence from health facilities as well as PCR parasite prevalence / antimalarial seroprevalence in an endline cross-sectional population survey. No significant difference was identified from routine incidence data and endline prevalence by polymerase chain reaction (PCR) had insufficient numbers of malaria infections (i.e., 16 infections among 6,276 children) to assess the intervention. Comparing long-term serological markers, we found a 19% (95% CI = 4-32%) reduction in seropositivity for the RFDA intervention using a difference in differences approach incorporating serological positivity and age. We also found a 37% (95% CI = 2-59%) reduction in seropositivity to short-term serological markers in a post-only comparison. These serological analyses provide compelling evidence that RFDA both has an impact on malaria transmission and is an appropriate end-game malaria elimination strategy. Furthermore, serology provides a more sensitive approach to measure changes in transmission that other approaches miss, particularly in very low transmission settings. Trial Registration: Registered at www.clinicaltrials.gov (NCT02654912, 13/1/2016).

Fig 1. Study participant flow.

Sixteen clusters were selected from all eligible health facility catchment areas in Southern Province, Zambia.

Fig 2. Median confirmed population malaria incidence for the entire HFCA (total) or just those identified at the health facility (HF).

RFDA (blue) and RFTAT (orange) arms are shown.

Fig 3. Seropositivity by trial arm and age for any of AMA-1, GLURP-R2, or MSP1-19 antigens in a post-only simple random sample.

Data are fitted using a loess smoother function and 95% confidence intervals. RFTAT control (orange) and RFDA intervention (blue) arms are shown accordingly. Plots for individual catchments are shown in S3 Fig.

Fig 4

Seropositivity for each of the health facility catchment populations to long-term antigens, stratified by age (left and middle), and short-term antigens (right). Health facilities are ordered in each panel according to maximum seropositivity observed. RFTAT control (orange) and RFDA intervention (blue) arms are shown accordingly. Bars show 95% confidence intervals.