Effectiveness and Chemoprevention Efficacy of Implementing Seasonal Malaria Chemoprevention in Karamoja Region, Uganda

The study aims to test the effectiveness and chemoprevention efficacy of seasonal malaria chemoprevention (SMC) with sulfadoxine-pyrimethamine (SPAQ) and dihydroartemisinin-piperaquine (DP) in the Karamoja region of Uganda. Specific objectives are:

1. to determine the effectiveness of SMC using SPAQ and DP;
2. to determine the chemoprevention efficacy of SPAQ and DP when used for SMC;
3. to investigate the presence and change of SPAQ and DP resistance markers over time as a result of SMC implementation;
4. to evaluate the process of SMC implementation, including coverage, quality of SMC implementation, costing, feasibility and acceptability, as well as an assessment of the role of gender in SMC implementation.

Study design This type 2 hybrid effectiveness-implementation study uses a convergent mixed-methods approach. SMC will be implemented in five monthly cycles between June 2022 and October 2022 in five districts of Karamoja. Table 2 provides an overview of the study components, as well as where and when they will be conducted, as well as the implementing partner and the proposed funding source.

Table 2. Uganda SMC implementation study phase 2: study components. Study component Primary outcome Objective Location Timing Implementing partner Funding Cluster randomised controlled trial Clinical malaria 1 Amudat May 2022 (baseline) and October 2022 (endline) Malaria Consortium Philanthropic Malariometric survey Malaria prevalence 1 Amudat May 2022 (baseline) and October 2022 (endline) Malaria Consortium Philanthropic Chemoprevention efficacy cohort study Chemoprevention failure Drug concentrations and drug resistance genotypes among chemoprevention failures 2 Nakapiripirit May/June 2022 (28 days following distribution of the first monthly SMC cycle) IDRC Bill \& Melinda Gates Foundation Resistance markers study Prevalence of antimalarial resistance markers 3 Amudat, Kotido, Moroto, Nabilatuk, Nakapiripirit April 2022 (baseline) and November 2022 (endline) IDRC Bill \& Melinda Gates Foundation End-of-round household survey Coverage and quality of SMC implementation 4 Amudat, Kotido, Moroto, Nabilatuk October 2022 Malaria Consortium Bill \& Melinda Gates Foundation Cost-effectiveness analysis Cost-effectiveness 4 n/a November 2022 Malaria Consortium Philanthropic Key informant interviews and focus group discussions Feasibility and acceptability of SMC implementation 4 Amudat, Kotido, Moroto, Nabilatuk August-October 2022 Malaria Consortium Bill \& Melinda Gates Foundation Methods

Malaria Consortium seeks funding from the Bill \& Melinda Gates Foundation for the following study components:

Chemoprevention efficacy cohort study While the detailed methods of the chemoprevention efficacy cohort study are described in the related proposal submitted by IDRC as the main implementing partner of this study component, Malaria Consortium will supervise the distribution of SMC medicines and coordinate community sensitisation activities. Some costs relating to this study component have therefore been included in Malaria Consortium's grant request. Malaria Consortium will also manage the relationship with Mahidol-Oxford Research Unit (MORU) in Bangkok, Thailand, the laboratory that will process the samples collected for this study component. The costs relating to sample processing by MORU have therefore also been included in this grant request.

End-of-round household survey At the end of the SMC round, a comprehensive household survey based on a multi-stage sampling protocol will be conducted to obtain representative data on coverage across the intervention area, as well as quality aspects such as implementers' adherence to SMC guidelines. The survey will be conducted in all four districts (all except Nakapiripirit), as only 1,000 children enrolled in the chemoprevention cohort study will receive SMC in that district. It uses a cross-sectional cluster-randomized sampling approach to target caregivers of children aged 3-119 months who were resident in the study location during the SMC implementation period. The sampling protocol aims to achieve a self-weighted sample with sampling units selected with probability proportional to size. Only at the last stage of sampling, i.e., at the household level, will a constant number of eligible children (one child per household) be selected. The survey will be powered to give an estimate of SMC coverage for children aged 3-59 months with a margin of error of 5%, while also providing a representative sample of children aged 60-119 months (based on the assumption of an intraclass correlation coefficient of 0.23). This will require a sample of 120 clusters, each comprising 15 children (n=1,800), randomly selected from areas with SMC delivery across the four districts with probability of selection proportional to population size.

The main objective of the survey is to determine if children received the full three-day course of SPAQ or DP during each cycle of the SMC round. Other coverage indicators the survey will measure for each cycle include:

• Proportion of households with eligible children visited by a community distributor in each cycle
• Proportion of Day 1 SPAQ and DP administered by community distributors to eligible children
• Proportion of eligible children who received a full three-day course of SPAQ or DP
• Proportion of Day 1 SPAQ and DP administered observing DOT The survey will also explore indicators relating to quality of SMC implementation and adherence to the SMC protocol, such as the proportion of children above the age of five who receive SMC despite not being eligible and the proportion of children who are correctly re-dosed if they vomit or spit out the medicines within 30 minutes of first administration. In addition, the survey will gather data on gendered intra-household dynamics and other gender issues, including household composition (e.g., female-headed/male-headed household, sex of the primary caregiver), as well as roles in health-related decision-making.

Data will be collected using a survey questionnaire developed by Malaria Consortium and translated into relevant local languages, using the SurveyCTO software application that will enable direct, field-based computer-aided personal interview (CAPI) and remote capture of the data and transfer to a netbook computer. All survey tools will be pre-tested. Daily quality assurance checks will be carried out during the field work. Data will be analysed using Stata 16. All indicators of interest will be calculated in proportion by district and an average across all districts. The 95% confidence interval, with correction for survey design effect, will be used to provide a range of values around the estimate within which selected result will be expected to fall. Data analysis will also involve testing the association of gender-related variables with SMC outcomes measured by the surveys (e.g., adherence to the three-day SPAQ regimen).

Key informant interviews and focus group discussions

To understand perceptions of the feasibility and acceptability of the intervention among a range of stakeholders, key informant interviews (KIIs) and focus group discussions (FGDs) will be conducted during the later stages of SMC implementation and immediately after the end of the annual SMC round. Participants will fall into four main categories:

1. Caregivers of children eligible for SMC
2. VHTs and health workers involved in SMC delivery
3. Community members in areas where SMC is implemented
4. Programme managers, policy makers and civil society organisations, including gender experts Purposive sampling will ensure that both female and male participants, as well as participants from all study districts are selected, except Nakapiripirit where only a small number of children will receive SMC as part of the chemoprevention efficacy study. Other social determinants such as rural/urban and participants' age will also be taken into account. A mix of KIIs and FGDs will be conducted until data saturation is reached, i.e. until further data does no longer generate new findings relating to the topic of investigation.[8] The total sample size can therefore not be determined at this stage. We estimate that around five participants per relevant category will be recruited for KIIs and that at least four FGDs comprising six to eight participants each will be conducted with participants from categories a, b and c. Typically, separate FGDs will be conducted for male and female participants.

Qualitative data will be collected by trained research assistants who are fluent in relevant local languages. Data collection will be facilitated by topic guides to explore views about SMC and reflections on the implementation experience. Topic guides for different categories of participants will focus on different areas of interest that are expected to be most relevant for the different types of respondents. The role of gender and the experiences of participants of different genders will be explored in detail with all respondent categories. This will include discussion of gender barriers and gender gaps, as well as health-related decision-making power. A gender analysis framework developed by Jhpiego will be adopted for this purpose.

Topic guides will be developed for all participant categories. However, the guides will only serve as prompts for the data collector. KIIs and FGDs will generally be participant-led, providing opportunities to discuss topics not covered by the topic guide. All KIIs and FGDs will be audio-recorded and research assistants will take notes, including on non-verbal cues. Data analysis will be conducted iteratively, beginning at the point of data generation, and with participant recruitment and topic focus being adapted as data collection progresses to further test emerging concepts, as well as potential discrepancies from majority themes.