Misra 1999

Methods	Study design: cluster randomized controlled trial. Baseline: Unit of allocation: village. Number of units: 15:15:15. Length of follow‐up: two weeks (16th to 30th September 1996). Mass surveys in September 1996 (peak transmission season) in 45 villages within the three ecological zones. Intervention: Unit of allocation: groups of three nearby villages (but at least one km apart) formed a cluster for random assignment of IRS, ITNs or control). The villages were not evenly distributed within the three ecological zones, 30 in coastal zone, 51 in the plains zone and 45 within the foothill eco zone. The distribution of the randomized villages was equal within the three groups. Number of units: 42:42:42. Length of follow‐up: 18 months. Active case detection by home visits twice a week by collecting blood smears from all fever cases. On Sundays, treatment was provided to any sick person calling on the health worker. Monitoring from May 1997 until May 1999 (only data collected until 31.12.1998 were evaluated). Cross‐sectional mass surveys once per year within the second half of September in 1997 and 1998. Drop out rates were 6.0% for IRS, 5.2% for ITN and 5.6% for the control group. The retrospective adjustment of incidence confidence intervals was not possible. Neither was the adjustment for prevalence data possible.
Participants	Baseline mass survey: Number of participants: 34,292. No explicit inclusion/exclusion criteria (all ages). Intervention: Number of participants: 93,210 (IRS: 30,989; ITNs: 31,168; control: 31,053). No explicit inclusion/exclusion criteria.
Interventions	The first intervention round took place from 26.5. to 14.6.1997; the second round from 23.5 to 14.6.1998. IRS: deltamethrin 2.5% WP; dosage 20 mg/m². Indoor surfaces of the walls, ceiling, back of cupboards, cots, eaves and cattle sheds were sprayed. Overall spray coverage was 92.2% in 1997 and 95.1% in 1998. In 1997 spray coverage was least in the irrigated plain eco zone (87%) and over 95% in the foothill and coastal eco zone. In 1998, coverage was 95% in all 3 zones. 1.36% and 19.3% of the houses which have received IRS were re‐plastered after three and six months, respectively. ITNs: deltamethrin 2.5% SC; dosage 25 mg/m². Overall net coverage of the whole population was very high, with 99.3% in 1997 and 85.4% in 1998, respectively. Overall, 86.8% of the nets were retreated, with the retreatment taking place one year after the distribution (May 1998). In the irrigated plain eco zone, 88.3% were retreated, whereas 95.7% of the nets in the foothill eco zone were retreated. The coverage rate for the coastal area was not mentioned. Only 4% of the nets were washed after nine months of treatment.
Outcomes	(1) Malaria prevalence. (2) Malaria incidence.
Notes	Study location: Surat District in Gujarat State, India. 3.7 million population with 54% distributed in 1281 villages. The district is divided into three ecological zones: (1) Foothill: Eastern tract of hilly land, largely deforested, summer hot and dry, maximal rainfall (2) Irrigated plain: cultivation of paddy, sugar cane and plantains, dam which irrigates the entire plain area (3) Coastal: Western coast belt with sandy soil and heavy industries. EIR: <1. Malaria endemicity: coastal area: hypo‐endemic; getting hyperendemic towards eastern hilly tracts. Transmission season: perennial transmission with peak from June to September. Main vector: Anopheles culicifacies (zoophilic and endophilic). Material of wall sprayed: mud or cement. Insecticide resistance: A. culicifacies is resistant to DDT and malathion, but highly susceptible to deltamethrin, the insecticide used within the study (mortality range for susceptibility tests: 74.4 to 96.5) Net ownership prior to distribution of nets for the trial was significant higher in the IRS group (16.5%) than in the other two groups (ITN:9.5%; control: 15.5%; Chi2 test=144.69, df=2, P=0.001)
Risk of bias
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Low risk	Quote:"Public drawing, witnessed by elected leaders, community members, and project government officials". Comment: Valid as randomization procedure.
Allocation concealment?	Low risk	Quote: "Public drawing, witnessed by elected leaders, community members, and project government officials". Comment: As there were many witnesses, the adherence to the randomized allocation is secured.
Blinding? Malaria infections	High risk
Incomplete outcome data addressed? Malaria infections	Low risk	Losses to follow‐up: Control: 1750/31053 losses to follow‐up (344/31053 due to death; 1215/31053 due to emigration and 191/31053 married and moved away). IRS: 1866/30989 losses to follow‐up (319/30989 due to death; 1332/30989 due to emigration and 215/30989 married and moved away). ITN: 1611/31168 losses to follow‐up (346/31168 due to death; 1034/31168 due to emigration and 231/31168 married and moved away). Additions to the study population: Control: 1035/31053 additions (603/31053 due to birth, 249/31053 due to marriage and 183/31053 due to immigration. IRS: 1250/30989 additions (713/30989 due to birth; 305/30989 due to marriage and 232/30989 due to immigration). ITNs: 1424/31168 additions (704/31168 due to birth; 320/31168 due to marriage and 400//31168 due to immigration).
Free of selective reporting?	Unclear risk	There is insufficient information for judgement
Free of other bias?	High risk	Quote: "a high incidence (24.3%) of plastering mud on the walls of houses..." (Bhatia et al. 2004) Comment: the re‐plastering of the walls after spraying most likely reduced the effectiveness of IRS