Barsosio 2024.

Study characteristics
Methods	Two‐arm multicentre, individually randomized, placebo‐controlled trial in 6 antenatal clinics in western Kenya (n = 3) and Malawi (n = 3) in areas with high‐grade S‐P resistance and perennial malaria transmission
Participants	904 women living with HIV Inclusion crieria Women living with HIV Eligible for (or on) daily ART consisting of tenofovir, lamivudine, and dolutegravir Had ultrasound confirmed viable singleton pregnancies between 16 and 28 weeks’ gestation Residents of study area Willing to adhere to scheduled and unscheduled study visit procedures and deliver in a study clinic Exclusion criteria Women with multiple pregnancies (e.g. twin pregnancies) Known heart conditions Advanced HIV disease at WHO clinical stages 3 and 4 Confirmed or suspected tuberculosis disease Known allergy or contraindication to dihydroartemisinin–piperaquine HIV‐negative or unknown HIV status
Interventions	Cotrimoxazole plus monthly dihydroartemisinin/piperaquine‐IPTp vs cotrimoxazole plus monthly placebo‐IPTp
Outcomes	The primary endpoint was the incidence of at least one Plasmodium infection detected in the peripheral (maternal) or placental (maternal) blood or tissue by PCR, microscopy, rapid diagnostic test, or placental histology (active infection) from 2 weeks after the first day of the first dose of the first course of dihydroartemisinin–piperaquine or placebo to delivery, inclusive. Key secondary efficacy endpoints included the individual components of the primary endpoint, clinical malaria, maternal haemoglobin concentrations, and anaemia measured in the third trimester and at delivery; maternal weight gain and mid‐upper arm circumference measured at each scheduled monthly visit; and adverse pregnancy outcome, defined as a composite of either foetal loss (miscarriage or stillbirth), small vulnerable newborn or with low birthweight (< 2500 g), or preterm (< 37 weeks’ gestation) or subsequent neonatal death by day 28, and the individual components of the composite adverse pregnancy outcome.
Notes	All participants received an LLIN.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Balanced randomization was done using computer‐generated permuted block randomization stratified by site and HIV status.
Allocation concealment (selection bias)	Low risk	An independent statistician, not involved in the study, generated the randomization list for the trial pharmacists in Kenya and Malawi, who prepared sequentially numbered, sealed, opaque envelopes for each participant with the randomization assignments. Contained in each opaque envelope were the pre‐packed investigational products for the entire study duration for that participant.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	All investigators, laboratory staff, data analysts, and participants were masked to treatment assignment.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	All investigators, laboratory staff, data analysts, and participants were masked to treatment assignment.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Very low number of losses in outcomes reported.
Selective reporting (reporting bias)	Low risk	None observed
Other bias	Low risk	No other biases identified