Rowland 2000

Methods	Study design: cluster‐randomized controlled trial. Unit of allocation: sectors; the study area comprising 60 villages was divided into nine sectors of approximately equal population size and surface area and then each sector was assigned at random to control, Wettable Powder (WP), or suspension concentrate (SC) formulation. Number of units: 3:3:3 sectors. During analysis the two insecticide groups (WP and SC) were merged into a single group because there was no evidence of difference between them. Length of follow‐up: 2 months before intervention and 7 months after spraying done in June 1997 (one season). Active case detection by home visits every fortnight. Blood slides were taken from any member of a household reporting to having had fever during the previous three days. Monitoring from April 1997 to January 1998, covering the entire malaria transmission season. Two cross‐sectional surveys were carried out in April‐May and September 1997, i.e. before and after the spraying, which was done in June 1997 (one survey within and one survey outside the malaria season, which runs from June to November). To assess the prevalence rate, blood slides were taken from children of one or two schools selected from sentinel villages in each sector. Drop‐out rates unknown. Confidence intervals were not adjusted for clustering by authors. We could adjust the incidence and prevalence data retrospectively. See Data collection and analysis for more details.The rate ratio (RR) of IRS vs no IRS was estimated by a generalized linear model with negative binomial mean and variance functions. This model sowed to best fit the observed cluster‐level incidence rates (Generalized Pearson statistics=1.29).
Participants	(1) Active case detection: Number enrolled:18,000 (2000 in each of the 9 sectors). Inclusion criteria into active surveillance group: any member of a household who reported having had fever during the previous 3 days. Exclusion criteria: No explicit exclusion (all ages). (2) Cross‐sectional surveys: Inclusion criteria: School children aged 5 to 15 years present in school on the day of the survey. Number enrolled: 200 to 300 children per sector. Exclusion criteria: none.
Interventions	Alpha‐cypermethrin WP and SC; dosage 25 mg/m²; living quarters, storage rooms and animals shelters were sprayed with Hudson X‐pert spray pumps Spray coverage: WP: 96%, SC: 97%.
Outcomes	(1) Malaria incidence through active case detection (P. falciparum and P. vivax). (2) Malaria prevalence through cross‐sectional surveys (P. falciparum and P. vivax).
Notes	Study location: 3 Union Councils, covering 180 km² of Sheikhupura District, Punjab Province, approximately 60 km west of Lahore, Pakistan. EIR: < 1 Malaria endemicity: not known (annual incidence of 50 episodes per 1000 person years). Malaria season: June to November. Main vector: Anopheles stephensi. Material of wall sprayed: mud and brick. Insecticide resistance: no detected resistance, 100% mortality of laboratory‐reared and wild‐caught A. stephensi.
Risk of bias
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Unclear risk	Quote: "..each sector was assigned at random to untreated, WP, or SC spraying..." Comment: Insufficient information for judgement
Allocation concealment?	Unclear risk	Insufficient information for judgement
Blinding? Malaria infections	High risk
Incomplete outcome data addressed? Malaria infections	Unclear risk	This outcome was not addressed by the study
Free of selective reporting?	Unclear risk	There is insufficient information to permit a judgement
Free of other bias?	Low risk