Fig. 2: Arm-differential and temporal patterns in between cluster heterogeneity in malaria CRTs.

a, b Difference in prevalence k (a) and ICC (b) by arm for each survey by months since intervention among trials. Positive k/ICC difference (above horizontal dashed line) indicates higher k/ICC value in the intervention arm compared to the control. Horizontal dashed line: Time of intervention implementation. Bar chart: number of surveys with k/ICC estimates higher in the intervention arm (>10% difference), the same between arms (<10% difference) and higher in the control arm (>10% difference). c Difference in incidence k between arms for each by study year among trials. Positive k difference (above horizontal dashed line) indicates higher k value in the intervention arm compared to the control. Bar chart: number of study years with k estimates higher in the intervention arm (>10% difference), the same (<10% difference) and higher in the control arm (>10% difference). d, e Temporal patterns in control-arm prevalence k (d) and ICC (e) values for each survey over months since intervention implementation, stratified by trial endemicity. Dots represent control-arm survey k estimates. Dashed lines represent individual trials. Error bars represent 95%CIs. Surveys conducted <0 months since intervention represent those conducted prior to intervention roll out.