Curtis 1998

Methods	Study design: cluster randomized controlled trial. Unit of allocation: village. Number of units: 4:4:4 Length of follow‐up: prevalence surveys and passive surveillance: 15 months (3rd & 4th quarter 95 to 4th quarter 96). Incidence surveys: 20 months (April 95 to Dec 1996). Incidence of re‐infection was monitored once before the interventions and four times after introduction of intervention (once in each quarter of 1996) by taking weekly blood slides. Cross‐sectional surveys were carried out monthly from April 1995 to December 1996. In addition, a passive surveillance system was set up. People feeling sick with fever were encouraged to visit a local research assistant, who was taking a blood slide from them, which were collected weekly. Different children were used for each cohort. Confidence intervals were not adjusted for clustering. We could retrospectively adjust for the prevalence, but not for the incidence data.
Participants	Number of participants: Incidence: 60:60:60 (control:ITNs:IRS), prevalence: 104:93:86, passive surveillance: 500:357:795. Inclusion criteria: incidence: children aged 1 to 6 with cleared pre‐existing parasitaemia; prevalence: children aged 1 to 6, passive surveillance: people of all ages feeling sick with fever. Exclusion criteria: incidence: children away from home, for having missed the blood slide for more than 1 week; prevalence: children already included in the incidence group, children which were selected in the previous month and children with parasitaemia >4000/μl. Passive surveillance: no specific exclusion criteria mentioned.
Interventions	IRS: Microencapsulated lambdacyhalothrin (ICON™) 10%; dosage: 30 mg/m². The wall and roof areas were sprayed with Hudson X‐Pert spray pumps. Re‐spraying in the villages was carried out seven to eight months after the initial spraying (July to August 1996). The spray coverage was not specifically mentioned but maximal coverage was aimed for. ITNs: Lambdacyhalothrin (ICON™); dosage: 10 mg/m² in 2 villages, and 20 mg/m² in the other 2 villages. Retreatment after seven months. The coverage rate was not specifically mentioned.
Outcomes	(1) Incidence of re‐infection after parasitological clearance with antimalarials (2) Incidence rates by passive case detection (3) Malaria prevalence (4) Haemoglobin levels
Notes	Study location: six villages near Muheza, Tanga Region and six villages near Hale, both in northeast Tanzania EIR: estimated to be above 300 Malaria endemicity: high endemicity with intense perennial transmission Transmission season: April to June Main vector: Anopheles gambiae, Anopheles funestus and Anopheles arabiensis Material of wall sprayed: Mud Insecticide resistance: None (bioassay test showed mortality of 80‐100%)
Risk of bias
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Unclear risk	Quote: "Random assignment of intervention" Comment: Insufficient information to permit a judgement
Allocation concealment?	Unclear risk	Quote: "Random assignment of intervention" Comment: Insufficient information to permit a judgement
Blinding? Malaria infections	High risk
Incomplete outcome data addressed? Malaria infections	Unclear risk	The study did not address this outcome
Free of selective reporting?	Unclear risk	There is insufficient information to permit a judgement
Free of other bias?	Low risk