BACKGROUND: Artemisinin-based combination therapies (ACTs) are first-line treatment for uncomplicated malaria in sub-Saharan Africa, but emerging artemisinin partial resistance threatens efficacy. Multiple first-line treatments (MFTs) represent a proposed strategy to delay resistance emergence by deploying several ACTs simultaneously, sequentially, or using other strategies rather than relying on a single first-line therapy.
STUDY DESIGN: This 28-month implementation program (June 2020 - October 2022) with and extension only on Mfangano Island until January 2024, deployed adaptive cycling of four artemisinin-based combinations across health facilities in Western Kenya using geographic allocation:
INTERVENTION COUNTY - HOMA BAY:
* Homa Bay Mainland (n=approximately 40,000 patients treated): Sequential deployment with 8-month cycling using crowding-out approach: Baseline (AL) → DHA-PIP (8 months) → AS+AQ (8 months) → AL (8 months)
* Mfangano Island (n=approximately 21,000 patients treated): Extended single MFT deployment: Baseline (AL) → AS+PYD (39 months). This extension allowed longer follow-up surveillance of the most recent combination ACT in Kenya where data was not previously monitored.
COMPARISON COUNTY - MIGORI:
\- Continued artemether-lumefantrine (AL) monotherapy throughout study period (n=32,835 patients treated) PARTICIPANTS: Adults and children ≥5 years diagnosed with uncomplicated Plasmodium falciparum malaria at participating health facilities. Excluded: pregnant women and children \<5 years due to medication restrictions and lack of pediatric formulations.
OUTCOMES MEASURED:
Primary Outcomes (reported in Cole et al., Malaria Journal 2024):
1. Health systems feasibility was assessed through commodity management, human resources adequacy, information system functionality, and implementation barriers/facilitators analysis
2. Economic costs of MFT deployment including start-up and implementation phases, calculated per facility and per patient were sought after
3. Stakeholder acceptability among policymakers, healthcare workers, and patients assessed through key informant interviews and surveys was analyzed.
Secondary Outcomes (reported in companion manuscript recently submitted for peer-review):
1. Prevalence of antimalarial resistance markers in dhfr (N51I, C59R, S108N, I164L), dhps (S436H, A437G, K540E, A581G), mdr1 (N86Y, Y184F), and k13 (A578S, A675V) genes
2. Temporal trends in resistance marker prevalence over four timepoints (September 2020, August-October 2021, May-July 2022, November 2023-January 2024)
3. Geographic distribution of resistance patterns across study sites
4. Complexity of infection (COI) determined by molecular methods
METHODS:
* Sample collection: Dried blood spots (n=310) from malaria-positive patients at four timepoints
* Molecular analysis: Targeted amplicon deep sequencing using Oxford Nanopore Technology
* Health systems assessment: Semi-structured questionnaires, key informant interviews, exit interviews, cost analysis using activity-based costing
* Statistical analysis: Chi-square tests, Cochran-Armitage trend tests, machine learning models for resistance prediction SETTING: Government and faith-based health facilities in Homa Bay County facilities received MFTs and Migori County (control) where AL was the only ACT used in Western Kenya, a malaria-endemic region with high transmission intensity.
ETHICAL APPROVAL: Strathmore University Institutional Scientific and Ethics Review Committee (SU-ISERC 1730/20) and Kenya National Commission for Science, Technology and Innovation (NACOSTI).
