1. Objectives

To compare the effect of insecticide-treated nets to no nets or untreated nets on the prevention of malaria in pregnancy

2. How studies were identified

The following databases were searched to February 2009:

• The Cochrane Infectious Diseases Group Specialized Register
• CENTRAL (The Cochrane Library 2009, Issue 1)
• MEDLINE
• EMBASE
• LILACS

Reference lists were also handsearched and researchers working in the field were contacted

3. Criteria for including studies in the review

3.1 Study type

Randomized controlled trials, including cluster-randomized controlled trials

3.2 Study participants

Pregnant women in malaria endemic areas

3.3 Interventions

Insecticide-treated nets (ITN) compared to no nets or untreated nets

(Women in both arms may also have received malaria chemoprophylaxis or intermittent preventive therapy)

3.4 Primary outcomes

Maternal anaemia during pregnancy

• Severe maternal anaemia (haemoglobin <70 or 80 g/L, or haematocrit <21% or 25%)
• Any anaemia (haemoglobin <10 or 11 g/dL, or haematocrit <30%)
• Mean haemoglobin or haematocrit

Infant birth weight

• Low birth weight (<2.5 kg)
• Mean birth weight

Secondary outcomes included clinical malaria illness during pregnancy (parasitaemic and febrile, or a history of fever); peripheral parasitaemia (presence of malaria parasites determined by finger prick); parasite density; placental parasitaemia; preterm delivery (<37 weeks’ gestation); abortion or stillbirth; neonatal death (death within first 28 days after birth); infant death; maternal death; and illness warranting hospitalization

4. Main results

4.1 Included studies

Five trials enrolling a total of 6759 pregnant women were included in this review:

• Three trials were cluster-randomized by community and reported on pregnant women as a subgroup (5455 women), whereas the other two trials randomized individual pregnant women (1304 women)
• Three trials recruited women of any parity, one trial recruited women having their first or second baby, and one trial recruited women having their first baby
• Nets were distributed before the pregnancy (2 trials), before the start of the study (1 trial), or at the time of recruitment (2 trials)
• Frequency of blood tests to detect parasitaemia and measure haemoglobin and/or haematocrit levels varied between the trials
• Where parasitaemia and/or low haemoglobin/haematocrit were detected, women were treated with quinine, chloroquine, or sulfadoxine-pyrimethamine, and in some trials also received iron plus folic acid; in one trial half the women also received prophylactic intermittent preventive therapy with sulfadoxine-pyrimethamine, while the other half received placebo
• Four trials used no nets as a comparison, and one trial used untreated nets. Three trials employed permethrin-treated nets, while one used cyfluthrin-treated nets

4.2 Study settings

• Ghana, Kenya (3 trials), and the border between Thailand and Myanmar
• Three trials randomized entire communities in Africa, and one trial was conducted in refugee camps for displaced people of the Karen ethnic minority on the border between Thailand and Myanmar
• Malaria transmission rates varied between trials

4.3 Study settings

How the data were analysed
Two comparisons were made: i) ITN compared to no nets, and ii) ITN compared to untreated nets. A decision was made a priori not to combine no nets with untreated nets in meta-analysis. In one study with three treatment arms (Dolan et al., 1993), contamination occurred, whereby women in the no net group were given nets by women in the ITN or untreated net groups. Therefore, only data from the comparison of the ITN group versus the untreated net group were used in the current analyses. All three cluster-randomized trials adjusted for clustering appropriately and were thus included in meta-analysis. For dichotomous outcomes, odds ratios (OR) were considered to approximate risk ratios (RR) where the event rate was rare (<10%), and were pooled in meta-analysis using the generic inverse variance method, with corresponding 95% confidence intervals (CI) generated. Hazard ratios (HR) were not combined with RR in meta-analysis. For continuous outcomes, mean differences (MD) and 95% CI were generated. Where moderate heterogeneity was detected (I²>50%), random effects models were used, otherwise fixed effects models were employed. Subgroup analyses were performed by parity (< three children, ≥ three children); and a funnel plot was planned for investigation of potential publication bias

Results
Insecticide-treated nets compared to no nets
Severe anaemia during pregnancy (haemoglobin <70 or 80 g/L)
Pooled analysis of data from two cluster-randomized trials resulted in a non-statistically significant reduced likelihood of severe anaemia during the third trimester of pregnancy among women receiving ITN (OR 0.77, 95% CI [0.56 to 1.08], p=0.13). In a further cluster-randomized trial, the risk of severe anaemia at delivery did not differ between treatment groups (RR 0.98, 95% CI [0.63 to 1.52]).

Any anaemia during pregnancy (haemoglobin <100 or 110 g/L)
For the outcome any anaemia during pregnancy, estimates of treatment effect were not statistically significant in three separate analyses from three separate trials (all p≥0.26). However, in subgroup analysis of women pregnant with up to their fourth child, a statistically significant reduction in any anaemia at birth was observed (HR 0.79, 95% CI [0.65 to 0.96], 1 cluster-randomized trial).

Mean haemoglobin during pregnancy (g/L)
The MD in haemoglobin between treatment groups was not statistically significantly different in three pooled analyses: among women in their third trimester with their first or second pregnancy (MD -0.04 g/L, 95% CI [-2.41 to 2.33]); among women in their third trimester with their third or greater pregnancy (MD 0.30 g/L, 95% CI [-3.38 to 3.97]); and among women at delivery with their first or second pregnancy (MD 1.95 g/L, 95% CI [-2.61 to 6.51]).

Low birth weight (<2.5 kg)
Treatment with ITN reduced the risk of delivering a low birth weight baby by 23% among women having their first or second child in pooled analysis (RR 0.77, 95% CI [0.61 to 0.98], p=0.03; 2 trials). This became non-statistically significant overall (RR 0.80, 95% CI [0.64 to 1.00], p=0.052; 2 trials), and the direction of effect was reversed among women having their fifth or greater baby (RR 1.12, 95% CI [0.56 to 2.24], 1 trial).

Birth weight (kg)
Meta-analysis of four trials revealed a statistically significant greater birth weight among infants born to ITN-treated mothers having their first or second child (MD 0.05 kg, 95% [0.02 to 0.09], p=0.0014), and overall (MD 0.06 kg, 95% CI [0.02 to 0.09], p=0.0011; 4 trials), but not for those having their third or greater child (MD 0.08 kg, 95% CI [-0.13 to 0.28], p=0.46; 1 trial).

Additional outcomes
In pooled analysis of two trials, the risk of peripheral parasitaemia at delivery was reduced with assignment to ITN (RR 0.76, 95% CI [0.67 to 0.86]), although clinical malaria during pregnancy and parasite density were not different between treatment groups. Overall, the risk of placental parasitaemia was reduced among the ITN group (RR 0.79, 95% CI [0.63 to 0.98], p=0.034; 3 trials), as was the risk of fetal loss (RR 0.68, 95% CI [0.48 to 0.98], p=0.036; 3 trials), although these effects were not significant in women having their fifth or greater child. Preterm delivery, neonatal death and maternal death were not different between groups.

Insecticide-treated nets compared to untreated nets
Any anaemia during pregnancy (haematocrit <30%)
The risk of any anaemia during pregnancy was reduced by 37% among women receiving ITN in the single trial comparing ITN to untreated nets (RR 0.63, 95% CI [0.42 to 0.93]).

Low birth weight (<2.5 kg)
No statistically significant difference was found in the risk of low birth weight between infants born to women assigned ITN and those born to women assigned untreated nets (RR 1.04, 95% CI [0.52 to 2.07]; 1 trial/179 women). No statistically significant difference was reported between groups for mean birth weight in the original trial (Dolan et al., 1993); however, this could not be re-analysed with the data provided.

Additional outcomes
The risk of peripheral parasitaemia was reduced among women using ITN at one refugee camp (RR 0.39, 95% CI [0.37 to 0.82]), but not at another (RR 1.10, 95% CI [0.57 to 2.12]) Overall, the risk of fetal loss was reduced by 79% among women using ITN in comparison to those using untreated nets (RR 0.21, 95% CI [0.05 to 0.92]). No other statistically significant differences were noted between groups for any other secondary outcomes analysed.

5. Additional author observations*

The methodological quality of the included trials was mixed, with three of the five studies describing adequate allocation concealment and one trial achieving double blinding (ITN versus untreated nets). Four of the five trials were conducted in Africa, limiting the generalizability of the results to other regions, such as Asia and Latin America. In addition, the intensity of detection and treatment of anaemia and parasitaemia during follow-up may have been greater within the trials than in routine clinical practice, potentially reducing the observed effect of ITN on these outcomes.

ITN have a beneficial impact on pregnancy outcomes in malaria-endemic regions of Africa when used by whole communities or by individual women. Consequently, the authors state that no further trials of ITN in pregnancy are required in sub-Saharan Africa, and that research efforts should instead focus on improving distribution of ITN to pregnant women, and on investigating the combined effects of ITN and intermittent preventive therapy for malaria. Further evaluation of ITN in Asia and Latin America is required before the recommendation to use ITN during pregnancy in these regions can be made.