Perennial Malaria Chemoprevention (PMC) Effect Study: Nigeria Operational Feasibility and Effectiveness

Study Design and Procedures The impact of PMC will be evaluated in a two-arm cluster-randomised controlled trial and nested case control study. The primary impact outcomes are a comparison of the incidence of clinical malaria in children 2-18 months of age visiting selected sentinel primary care facilities in intervention arm and comparison (standard care) arm wards in the study site, and the protective effectiveness of SP in the 28 days following a dose in intervention arm wards. A ward, the lowest level of health services administration for the State, will be the cluster unit of randomisation. A single facility meeting study eligibility criteria will be selected in each ward as the sentinel facility in which study outcomes will be evaluated (bi-monthly extraction of data on confirmed malaria cases by the study team). Strengthened age recording (full age in months) of visiting children will be implemented in each sentinel facility. An identical programme of data collection will take place in secondary and tertiary hospitals which serve the study site and will include data on admissions of children with an address in a study ward.

Cross sectional household 'malariometric' surveys will be conducted in catchment area communities of ;in both arms at baseline, at 18 months following start of PMC implementation, and at 24 months once implementation ends. Data on coverage of PMC doses (main indicator of feasibility) will be collected from child health record cards or caregiver report for children in the age groups eligible to receive the doses in question. Additional data on malaria infection (RDT test) and anaemia prevalence will be collected during the cross sectional surveys.

Collection of malaria data from sentinel facilities for children aged 19 months and older who were eligible to receive PMC during the implementation period (but too old at the end of implementation) will continue for an additional 12 months following the end of implementation period. The potential rebound effects of the intervention will be assessed in this period by a comparison of incidence rates in control and intervention arm wards.

Sample size (impact): The expected reduction in incidence of malaria in children aged 2-18 months after 18 months of implementation is conservatively estimated to be between 18%-32% contingent on the level of coverage of PMC doses in the intervention arm; enrolling 40 wards in each arm (for a total of 80) will provide 90% power to detect an impact of 20% in this age group after 18 months of implementation, and 95% after 24 months of implementation. We will also recruit children aged 2-18m with clinical malaria from 8 of the 40 intervention arm sentinel facilities and match these to 2-4 controls to have 95% power to detect a protective effectiveness of 50% reduction in malaria in the 28 days following a dose of SP.

Sample size (feasibility): To estimate coverage for any dose of 50% with a margin of error of 8%, assuming a design effect of 2.44 based on an implied intra cluster correlation coefficient of 0.18 (Nigeria DHS) to account for the cluster sample strategy, and allowing for 10% non-response, 25 children aged 2-36 months will be required in each cluster (ward), for a total of 25x40=1,000 in each arm of the study and with a total of 2000 across both arms.