The authors concluded that these data supported continued scale-up of these malaria prevention interventions in endemic settings that will prevent a considerable number of deaths due directly and indirectly to malaria. The review had some methodological weaknesses that limited the reliability of the authors' conclusions.

To estimate the effects of insecticide-treated nets and indoor-residual spraying on preventing malaria in children and the effects of intermittent preventive therapy in pregnant women (IPTp) for improving birth outcomes in Plasmodium falciparum endemic areas.

PubMed, WHOLIS, CAB abstracts via OVID Global Health, SIGLE and LILAC search engines were searched for published articles. Experts in the field were contacted for unpublished studies. Search terms were reported.

Three separate reviews were undertaken: one to estimate the efficacy of insecticide-treated nets in children aged one to 59 months; one to estimate the efficacy of indoor-residual spraying in children aged one to 59 months; and one to estimate the efficacy of IPTp on preventing malaria during pregnancy related to improvements in birth outcomes.

Individual and group randomised controlled trials (RCTs) that compared the effects of insecticide-treated nets with no nets and the effects of indoor-residual spraying with no spraying in children aged one to 59 months in malaria endemic areas were eligible for inclusion. Studies had to report malaria-attributable mortality, all-cause mortality, incidence of severe or uncomplicated malaria or the prevalence of malaria parasite infection. Studies that compared the effects of at least two doses of IPTp with sulphadoxine-pyrimethamine compared with placebo and studies that compared the effects of insecticide-treated nets compared with no nets in pregnant women in malaria endemic areas were also eligible for inclusion.

Studies had to report intrauterine growth retardation (IUGR), pre-term delivery, prematurity, low birth weight, perinatal mortality, neonatal mortality, infant mortality or child mortality. Studies that compared two doses of IPTp with another drug were excluded, as were studies in which the effects of IPTp could not be separated from insecticide-treated nets. Studies that compared untreated nets with insecticide-treated nets were excluded, as were studies in which the effects of insecticide-treated nets could not be separated from IPTp. Where there were insufficient data identified from RCTs, quasi-experimental and observational studies were identified.

The included studies assessed insecticide-treated nets in rural settings in children aged one month to less than 15 years and indoor-residual spraying in rural and mixed settings in households in malaria endemic areas. IPTp studies compared IPTp with placebo in pregnant women in routine antenatal clinics and compared insecticide-treated nets with no insecticide-treated nets in pregnant women (first two to four pregnancies) in a rural setting. Publication dates ranged from 1969 to 2008. The settings, where reported, included high or low transmission, seasonal or intense perennial transmission. A small number of studies also administered other interventions, such as mass administration of sulfalene-pyrimethamine.

The authors did not state how many reviewers undertook study selection, but stated that they followed Child Health Epidemiology Reference Group (CHERG) guidelines.

Trial quality assessment was based on the CHERG adaptation of the GRADE criteria. The authors did not state how many reviewers performed quality assessment.

Data were extracted on mortality, incidence of malaria and prevalence of malaria. These data were used to calculate relative risks (RRs) and 95% confidence intervals (CIs).

The pooled relative risks, together with 95% CIs, were calculated using a fixed-effect or random-effects meta-analysis, depending on the presence of statistical heterogeneity. Heterogeneity was assessed using the X² test and I² statistic. To take account of using estimates of the protective effects of insecticide-treated nets from post-mortem verbal autopsies, a pooled estimate of the protective effect of insecticide-treated nets for reducing malaria-attributable mortality and a non-pooled estimate were transformed to relative risks and then pooled.

Ten community RCTs were included in the review of insecticide-treated nets for children; study quality was deemed high outcome specific quality, although most were not blinded. Seven before and after studies were included in the review of indoor-residual spraying for children; study quality was deemed low outcome specific quality. Four randomised trials were included in the review of IPTp in pregnant women; study quality was deemed high outcome specific quality, although most of the studies did not use intention-to-treat analysis.

Insecticide-treated nets for children: Risk of all-cause child mortality was lower with insecticide-treated nets compared with no nets (RR 0.82, 95% CI 0.75 to 0.90; three RCTs, n=96,038). Risk of uncomplicated malaria incidence was lower with insecticide-treated nets compared with no nets (RR 0.49, 95% CI 0.44 to 0.54; four RCTs, n=59,918). Risk of prevalent malaria infection was lower with insecticide-treated nets compared with no nets (RR 0.83, 95% CI 0.64 to 0.88; six RCTs, n=8,014). There was evidence of statistical heterogeneity for uncomplicated malaria and prevalence of malaria parasite infection. The protective efficacy of insecticide-treated nets on reducing malaria attributable mortality at one to 59 months was approximately 55% (range 49-60%).

Indoor-residual spraying for children: Risk of all-cause child mortality was lower with indoor-residual spraying compared with no spraying (RR 0.53, 95% CI 0.40 to 0.71; two studies, n=4,000). Risk of all-cause infant mortality was lower with indoor-residual spraying compared with no spraying (0.47, 95% CI 0.40 to 0.54; three studies, n=18,622). Risk of prevalent malaria infection was lower with indoor-residual spraying compared with no spraying (0.16, 95% CI 0.03 to 0.94; five studies, n=unknown). There was no statistically significant difference in the risk of uncomplicated malaria incidence with indoor-residual spraying compared with no spraying (three studies, n=unknown).

IPTp in pregnant women: Compared with placebo, there was no statistically significant difference in risk of neonatal mortality (two RCTs, n=2,091) and perinatal mortality (one RCT, n=904) with IPTp in pregnant women.

IPTp and insecticide-treated nets during pregnancy: Risk of low birth weight was lower with insecticide-treated nets compared with no nets (RR 0.65, 95% CI 0.55 to 0.77; two studies, n=3,360). The protective efficacy of IPTp and insecticide-treated nets on reducing the prevalence of low birth weight in the first or second pregnancy was approximately 35% (95% CI 23% to 45%).

These data support the continued scale-up of these malaria prevention interventions in endemic settings that will prevent a considerable number of deaths due directly and indirectly to malaria.

Inclusion criteria for the review were clearly defined. Several relevant databases were searched for published articles; search dates were not reported. Publication bias was not assessed and could not be ruled out. The authors did not state how many reviewers performed study selection, data extraction and quality assessment, but did state that they adhered to CHERG guidelines, which should have minimised error and bias in the review. The quality assessment indicated the generally high quality of the included studies, apart from the analysis of indoor-residual spraying in children; however, none of the studies were blinded and several of the trials did not use intention-to-treat. There were limited details on interventions and participants, which made assessment of comparability of studies difficult. The estimates of malaria-attributable mortality were not direct estimates, but based on two statistical approaches of indirect estimation and must, therefore, be interpreted with caution. Studies were combined using meta-analysis and heterogeneity was taken into consideration; however, there was still evidence of statistical heterogeneity in several of the analyses and this was not investigated.

Overall, the review had some methodological weaknesses that limited the reliability of the authors' conclusions.

Practice: The authors stated that these data will allow the use of the Lives Saved Tool (LiST) model for estimating the likely impact of past intervention scale-up, as well as for predicting future impact of intervention scale-up by national malaria control programs.

Research: The authors did not state any implications for research.

Funded by the US Fund for UNICEF from the Bill & Melinda Gates Foundation (grant 43386).

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.