Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Review - Intervention

Drugs for preventing malaria in pregnant women in endemic areas: any drug regimen versus placebo or no treatment

Authors' declarations of interest

Version published: 10 October 2014 Version history

https://doi.org/10.1002/14651858.CD000169.pub3
Collapse all Expand all

Abstract

available in

Background

Pregnancy increases the risk of malaria and this is associated with poor health outcomes for both the mother and the infant, especially during the first or second pregnancy. To reduce these effects, the World Health Organization recommends that pregnant women living in malaria endemic areas sleep under insecticide‐treated bednets, are treated for malaria illness and anaemia, and receive chemoprevention with an effective antimalarial drug during the second and third trimesters.

Objectives

To assess the effects of malaria chemoprevention given to pregnant women living in malaria endemic areas on substantive maternal and infant health outcomes. We also summarised the effects of intermittent preventive treatment with sulfadoxine‐pyrimethamine (SP) alone, and preventive regimens for Plasmodium vivax.

Search methods

We searched the Cochrane Infectious Diseases Group Specialized Register, CENTRAL, MEDLINE, EMBASE, LILACS, and reference lists up to 1 June 2014.

Selection criteria

Randomized controlled trials (RCTs) and quasi‐RCTs of any antimalarial drug regimen for preventing malaria in pregnant women living in malaria‐endemic areas compared to placebo or no intervention. In the mother, we sought outcomes that included mortality, severe anaemia, and severe malaria; anaemia, haemoglobin values, and malaria episodes; indicators of malaria infection, and adverse events. In the baby, we sought foetal loss, perinatal, neonatal and infant mortality; preterm birth and birthweight measures; and indicators of malaria infection. We included regimens that were known to be effective against the malaria parasite at the time but may no longer be used because of parasite drug resistance.

Data collection and analysis

Two review authors applied inclusion criteria, assessed risk of bias and extracted data. Dichotomous outcomes were compared using risk ratios (RR), and continuous outcomes using mean differences (MD); both are presented with 95% confidence intervals (CI). We assessed the quality of evidence using the GRADE approach.

Main results

Seventeen trials enrolling 14,481 pregnant women met our inclusion criteria. These trials were conducted between 1957 and 2008, in Nigeria (three trials), The Gambia (three trials), Kenya (three trials), Mozambique (two trials), Uganda (two trials), Cameroon (one trial), Burkina Faso (one trial), and Thailand (two trials). Six different antimalarials were evaluated against placebo or no intervention; chloroquine (given weekly), pyrimethamine (weekly or monthly), proguanil (daily), pyrimethamine‐dapsone (weekly or fortnightly), and mefloquine (weekly), or intermittent preventive therapy with SP (given twice, three times or monthly). Trials recruited women in their first or second pregnancy (eight trials); only multigravid women (one trial); or all women (eight trials). Only six trials had adequate allocation concealment.

For women in their first or second pregnancy, malaria chemoprevention reduces the risk of moderate to severe anaemia by around 40% (RR 0.60, 95% CI 0.47 to 0.75; three trials, 2503 participants, high quality evidence), and the risk of any anaemia by around 17% (RR 0.83, 95% CI 0.74 to 0.93; five trials,, 3662 participants, high quality evidence). Malaria chemoprevention reduces the risk of antenatal parasitaemia by around 61% (RR 0.39, 95% CI 0.26 to 0.58; seven trials, 3663 participants, high quality evidence), and two trials reported a reduction in febrile illness (low quality evidence). There were only 16 maternal deaths and these trials were underpowered to detect an effect on maternal mortality (very low quality evidence).

For infants of women in their first and second pregnancies, malaria chemoprevention probably increases mean birthweight by around 93 g (MD 92.72 g, 95% CI 62.05 to 123.39; nine trials, 3936 participants, moderate quality evidence), reduces low birthweight by around 27% (RR 0.73, 95% CI 0.61 to 0.87; eight trials, 3619 participants, moderate quality evidence), and reduces placental parasitaemia by around 46% (RR 0.54, 95% CI 0.43 to 0.69; seven trials, 2830 participants, high quality evidence). Fewer trials evaluated spontaneous abortions, still births, perinatal deaths, or neonatal deaths, and these analyses were underpowered to detect clinically important differences.

In multigravid women, chemoprevention has similar effects on antenatal parasitaemia (RR 0.38, 95% CI 0.28 to 0.50; three trials, 977 participants, high quality evidence)but there are too few trials to evaluate effects on other outcomes.

In trials giving chemoprevention to all pregnant women irrespective of parity, the average effects of chemoprevention measured in all women indicated it may prevent severe anaemia (defined by authors, but at least < 8 g/L: RR 0.19, 95% CI 0.05 to 0.75; two trials, 1327 participants, low quality evidence), but consistent benefits have not been shown for other outcomes.

In an analysis confined only to intermittent preventive therapy with SP, the estimates of effect and the quality of the evidence were similar.

A summary of a single trial in Thailand of prophylaxis against P. vivax showed chloroquine prevented vivax infection (RR 0.01, 95% CI 0.00 to 0.20; one trial, 942 participants).

Authors' conclusions

Routine chemoprevention to prevent malaria and its consequences has been extensively tested in RCTs, with clinically important benefits on anaemia and parasitaemia in the mother, and on birthweight in infants.

PICOs

Population
Intervention
Comparison
Outcome

The PICO model is widely used and taught in evidence-based health care as a strategy for formulating questions and search strategies and for characterizing clinical studies or meta-analyses. PICO stands for four different potential components of a clinical question: Patient, Population or Problem; Intervention; Comparison; Outcome.

See more on using PICO in the Cochrane Handbook.

The effect of taking antimalarial drugs routinely to prevent malaria in pregnancy

Pregnancy increases the risk of malaria and this is associated with poor health outcomes for both the mother and the infant, especially during the first or second pregnancy. For this reason, women are encouraged to try and prevent malaria infection during pregnancy by sleeping under mosquito bed‐nets, and by taking drugs effective against malaria throughout pregnancy as chemoprevention.

This Cochrane Review looked at all drug regimens compared to placebo. The review authors sought to summarise and quantify the overall effects of chemoprevention. Seventeen trials were included, all conducted between 1957 and 2008, and all but two in countries of Africa.

For women in their first or second pregnancy, malaria chemoprevention prevents moderate to severe anaemia (high quality evidence); and prevents malaria parasites being detected in the blood (high quality evidence). It may also prevent malaria illness. We don't know if it prevents maternal deaths, as this would require very large studies to detect an effect.

In their infants, malaria chemoprevention improves the average birthweight (moderate quality evidence), and reduces the number of low birthweight infants (moderate quality evidence). We are not sure if chemoprevention reduces mortality of babies in the first week, month and year, as again studies would need to be very large to show these effects.

Authors' conclusions

Implications for practice

Routine chemoprevention to prevent malaria and its consequences has been extensively tested in RCTs, with clinically important benefits on anaemia and parasitaemia in the mother, and on birthweight in infants.

The data also assists in showing the potential attribution of malaria towards key endpoints, and what can be achieved by successful prevention to assist in modelling studies examining the impact of malaria on pregnancy.

Implications for research

Identifying current effective chemoprevention regimens remains a challenge, especially with the spread of drug‐resistant malaria, in particular against SP which is the only antimalarial currently recommended for IPT in pregnant women. There is justification for assessing the safety and efficacy of effects of alternative drugs that can replace SP in areas with high SP resistance, or alternative strategies that could replace IPT during pregnancy, such as intermittent screen and treat (IST) approaches that focus on prompt accessible treatment for anaemia and asymptomatic parasitaemia (Tagbor 2010).

All new trials should systematically and carefully collect adverse effects of regimens.

The data on the longer term impact on infants is poor and needs further study: currently the evidence mainly relates to effects on clinically important outcomes, such as preterm birth and birthweight.

There is a dearth of data from endemic areas outside of Africa, such as Asia and Latin America.

Summary of findings

Open in table viewer
Summary of findings for the main comparison. Summary of findings table 1

Malaria chemoprevention for pregnant women (parity 0‐1) living in endemic areas: maternal outcomes

Patient or population: Pregnant women (parity 0‐1)
Settings: Malaria‐endemic areas
Intervention: Malaria chemoprevention (any regimen)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

Chemoprevention

Mortality
All‐cause death

7 per 1000

8 per 1000
(3 to 20)

RR 1.15
(0.44 to 3.06)

2097
(3 trials)

⊕⊝⊝⊝
very low1,2

Severe anaemia
During the third trimester

145 per 1000

87 per 1000
(68 to 108)

RR 0.60
(0.47 to 0.75)

2503
(3 trials)

⊕⊕⊕⊕
high3,4,5,6

Anaemia

649 per 1000

539 per 1000
(480 to 604)

RR 0.83
(0.74 to 0.93)

3662
(5 trials)

⊕⊕⊕⊕
high3,6,7,8

Uncomplicated clinical malaria

173 per 1000

64 per 1000
(31 to 128)

RR 0.37
(0.18 to 0.74)

307
(2 trials)

⊕⊕⊝⊝
low4,9,10

Antenatal parasitaemia

286 per 1000

111 per 1000
(74 to 165)

RR 0.39
(0.26 to 0.58)

3663
(8 trials)

⊕⊕⊕⊕
high3,6,7,11

Severe adverse effects12

‐

‐

‐

‐

‐

*The basis for the assumed risk (eg, the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 Downgraded by 1 for risk of bias: Only one of these trials adequately described allocation concealment to be considered at low risk of selection bias.
2 Downgraded by 2 for imprecision: These trials were not adequately powered to detect a difference in mortality. Only 15 deaths occurred in these three trials.To confidently detect a 25% reduction in maternal mortality in a setting of 350 deaths/100,000 would require a sample size of over 100,000.
3 No serious risk of bias: Exclusion of the trials at high risk of bias did not change the statistical significance or clinical importance of the result.
4 No serious inconsistency: This finding was consistent across all trials and statistical heterogeneity was low.
5 No serious indirectness: These trials were conducted in Kenya and Mozambique between 1996 and 2005, all three trials administered IPT with SP. The definition of severe anaemia was variable; Hb < 8 g/dL, Hb < 7 g/dL, or PCV < 21%.
6 No serious imprecision: This result is statistically significant and the meta‐analysis is adequately powered to detect this effect.
7 No serious inconsistency: Although statistical heterogeneity was high, all trials favoured chemoprevention but there was variability in the size of the effect.
8 No serious indirectness: These trials were conducted in Nigeria, Kenya and Uganda between 1978 and 1999. Three trials administered IPT as SP, one gave weekly chloroquine, and one gave daily proguanil. The definition of anaemia was variable: Hb < 12 g/dL, Hb < 11 g/dL, Hb < 10 g/dL, PCV < 33% and PCV < 30%.
9 Downgraded by 1 for risk of bias. Both trials had high or unclear risk of selection bias and an attrition rate above 20%.
10 Downgraded by 1 for indirectness: Both these trials, from Cameroon 1993 and Mozambique 2002, measured fever history only as proxy for malaria illness.
11 Not downgraded for inconsistency. Despite substantive quantitative heterogeneity (I2 69% across six trials), all show at least a reduction of 23%, often more
11 No serious indirectness: These trials were conducted in The Gambia, Nigeria, Kenya and Mozambique between 1978 and 2005. Five trials gave IPT as SP, one gave pyrimethamine‐dapsone, one pyrimethamine, and one proguanil.
12 Reporting of adverse events was generally poor. No severe adverse events were reported.

Open in table viewer
Summary of findings 2. Summary of findings table 2

Malaria chemoprevention for pregnant women (parity 0‐1) living in endemic areas: infant outcomes

Patient or population: Pregnant women (parity 0‐1)
Settings: Malaria‐endemic areas
Intervention: Malaria chemoprevention (any regimen)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

Chemoprevention

Spontaneous abortion

33 per 1000

21 per 1000
(13 to 33)

RR 0.65
(0.41 to 1.02)

2876
(5 trials)

⊕⊕⊝⊝
low1,2,3,4

Stillbirth

33 per 1000

32 per 1000
(21 to 49)

RR 0.97
(0.64 to 1.49)

2703
(3 trials)

⊕⊕⊝⊝
low2,4,5,6,

Perinatal mortality

104 per 1000

76 per 1000
(56 to 104)

RR 0.73
(0.54 to 1.00)

1620
(2 trials)

⊕⊕⊝⊝
low2,4,5,7,

Neonatal mortality

37 per 1000

23 per 1000
(14 to 39)

RR 0.62
(0.37 to 1.05)

2156
(2 trials)

⊕⊕⊝⊝
low2,4,5,7,

Preterm birth

164 per 1000

140 per 1000
(108 to 181)

RR 0.85
(0.66 to 1.10)

1493
(2 trials)

⊕⊕⊝⊝
low1,2,4

Low birthweight

152 per 1000

110 per 1000
(92.7 to 132.2)

RR 0.73
(0.61 to 0.87)

3619
(8 trials)

⊕⊕⊕⊝
moderate9,10

Mean birthweight

The mean birthweight in the control groups ranged from

2723 g to 3079 g

The mean birthweight in the intervention groups was
92.72 g higher
(62.05 higher to 123.39 higher)

‐

3936
(9 trials)

⊕⊕⊕⊝
moderate5,10

Placental parasitaemia

307 per 1000

160 per 1000
(132 to 211)

RR 0.54
(0.43 to 0.69)

2830
(7 trials)

⊕⊕⊕⊕
high3,11,12

Cord blood haemoglobin

The mean haemoglobin in the control group was
15.8 g/dL

The mean haemoglobin in the intervention groups was
1.8 g/dL lower
(3.46 lower to 0.14 lower)

‐

64
(1 trial)

⊕⊝⊝⊝
very low1,13,14

*The basis for the assumed risk (eg the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 Downgraded by 1 for serious risk of bias: None of the trials described adequate measures to prevent selection bias.
2 No serious inconsistency: The effect is consistent across trials and statistical heterogeneity is low.
3 No serious indirectness: These trials were conducted in The Gambia, Cameroon, Kenya and Mozambique between 1990 and 2002. One gave chemoprevention as weekly chloroquine and four trials gave IPT with SP.
4 Downgraded by 1 for serious imprecision: The 95% CI is wide and sample remains underpowered to detect or rule out an effect.
5 Downgraded by 1 for serious risk of bias: Only one trial adequately described methods to prevent selection bias.
6 No serious indirectness: Trials were conducted in Cameroon and Kenya between 1993 and 1997. One trial gave weekly chloroquine and the others gave IPT as SP.
7 No serious indirectness: The trials were conducted in The Gambia and Kenya between 1984 and 1997. One trial used IPT with SP and one gave pyrimethamine‐dapsone which is no longer in use.
8 No serious indirectness: Both trials were conducted in Kenya and used IPT with SP.
9 Downgraded by 1 for serious risk of bias: Only two of these trials were at low risk of selection bias.
10 No serious indirectness: These trials were conducted in The Gambia, Cameroon, Kenya, Uganda and Mozambique between 1986 and 2005. The majority of trials used IPT with SP.
11 No serious inconsistency: Although statistical heterogeneity was high, all trials favoured chemoprevention but there was variability in the size of the effect.
12 No serious indirectness: These trials were conducted in The Gambia, Cameroon, Kenya, Uganda and Mozambique between 1990 and 2002. The majority of trials used IPT with SP.
13 Downgraded by 1 for serious indirectness: This single trial used a regimen that is no longer in use (proguanil).
14 Downgraded by 1 for serious imprecision: Only a single small trial has evaluated this comparison.

Open in table viewer
Summary of findings 3. Summary of findings table 3

Malaria chemoprevention for pregnant women (parity 2+) living in endemic areas: maternal outcomes

Patient or population: Pregnant women (parity 2+)
Settings: Malaria‐endemic areas
Intervention: Malaria chemoprevention (any regimen)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

Chemoprevention

Mortality
All‐cause death

5 per 1000

7 per 1000
(2 to 26

RR 1.47
(0.42 to 5.21)

2239
(1 trial)

⊕⊝⊝⊝
very low1,2,3

Severe anaemia
During the third trimester

68 per 1000

65 per 1000
(28 to 153)

RR 0.96
(0.41 to 2.25)

2682
(2 trials)

⊕⊕⊝⊝
low1,4,5

Anaemia

The mean PCV in the control group was
30.4 %

The mean PCV in the intervention group was
0.3 % higher
(0.7 lower to 1.3 higher)

‐

244
(1 trial)

⊕⊝⊝⊝
very low6,7,8

Uncomplicated clinical malaria

‐

‐

‐

‐
(0 trials)

‐

Antenatal parasitaemia

108 per 1000

41 per 1000
(30 to 54)

RR 0.38
(0.28 to 0.50)

3022
(4 trials)

⊕⊕⊕⊕
high9,10

Severe adverse events11

‐

‐

‐

‐

‐

*The basis for the assumed risk (eg the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 No serious risk of bias: These trials are at low risk of bias.
2 Downgraded by 1 for serious indirectness: This single trial was conducted in The Gambia between 2002 and 2004 and administered IPT as monthly SP. The findings may not be easily generalised to elsewhere.
3 Downgraded by 2 for very serious imprecision: Only ten deaths occurred in this trial. Much larger trials would be needed to detect or exclude effects on maternal mortality.
4 No serious indirectness: These two trials were conducted in The Gambia in 2002‐2004 and Mozambique between 2003 and 2005.
5 Downgraded by 2 for very serious imprecision: The 95% CI are very wide and include the possibility of both clinically important benefits and harms.
6 Downgraded by 1 for serious risk of bias: This single trial is at unclear risk of selection bias.
7 Downgraded by 1 for serious indirectness: This trial administered chemoprevention as pyrimethamine‐dapsone which is no longer in use.
8 Downgraded by 1 for serious imprecision: A much larger sample size is required to confidently detect or exclude an effect.
9 No serious risk of bias: Two of the four trials were at low risk of selection bias and exclusion of the other two trials did not change the size of the effect.
10 No serious indirectness: These three trials were conducted in The Gambia, Nigeria and Mozambique between 1986 and 2005. The biggest and most recent trial administered IPT with SP (two doses)

11 Reporting of adverse events was generally poor. No severe adverse events were reported.

Open in table viewer
Summary of findings 4. Summary of findings table 4

Malaria chemoprevention for pregnant women (parity 2+) living in endemic areas: infant outcomes

Patient or population: Pregnant women (parity 2+)
Settings: Malaria‐endemic areas
Intervention: Malaria chemoprevention (any regimen)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

Chemoprevention

Spontaneous abortion

‐

‐

‐

‐

(0 trials)

‐

Stillbirth

‐

‐

‐

‐

(0 trials)

‐

Perinatal deaths

‐

‐

‐

‐

(0 trials)

‐

Neonatal mortality

26 per 1000

38 per 1000
(23 to 62)

RR 1.46

(0.90 to 2.38)

2017

(1 trial)

⊕⊝⊝⊝
very low1,2,3

Preterm birth

‐

‐

‐

‐

(0 trials)

‐

Low birthweight

60 per 1000

63 per 1000
(46 to 85)

RR 0.86
(0.63 to 1.17)

2743
(3 trials)

⊕⊕⊝⊝
low3,4,5

Mean birthweight

‐

‐

‐

‐

(0 trials)

‐

Placental parasitaemia

‐

‐

‐

‐

(0 trials)

‐

Cord blood haemoglobin

‐

‐

‐

‐

(0 trials)

‐

*The basis for the assumed risk (eg the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 No serious risk of bias: This single trial was at low risk of selection bias.
2 Downgraded by 1 for serious indirectness: This single trial was conducted in The Gambia between 2002 and 2004 and administered IPT as monthly SP. The findings may not be easily generalised to elsewhere.
3 Downgraded by 2 for serious imprecision: The 95% CI is very wide and includes clinically important effects and no effect. A much larger sample size is required to confidently detect or exclude an effect.
4 No serious risk of bias: These trials are at low risk of selection bias.
5 No serious indirectness: These trials were conducted in The Gambia, Mozambique, and Uganda between 2002 and 2008.

Open in table viewer
Summary of findings 5. Summary of findings table 5

Malaria chemoprevention for all pregnant women (all parities) living in endemic areas: maternal outcomes

Patient or population: Pregnant women (all parities)
Settings: Malaria‐endemic areas
Intervention: Malaria chemoprevention (any regimen)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

Chemoprevention

Mortality
All‐cause death

1 per 1000

1 per 1000
(0 to 3)

RR 0.84
(0.25 to 2.74)

6026
(4 trials)

⊕⊕⊝⊝
low1,2,3

Severe anaemia
During the third trimester

26 per 1000

5 per 1000
(1 to 19)

RR 0.19
(0.05 to 0.75)

1327
(2 trials)

⊕⊕⊝⊝
low2,4,5,6

Anaemia

206 per 1000

212 per 1000
(179 to 253)

RR 1.03
(0.87 to 1.23)

3027
(3 trials)

⊕⊕⊕⊝
moderate1,2,7,8

Uncomplicated clinical malaria

114 per 1000

42 per 1000
(13 to 140)

RR 0.37
(0.11 to 1.23)

3455
(4 trials)

⊕⊕⊝⊝
low1,9,10

Antenatal parasitaemia

152 per 1000

106 per 1000
(67 to 172)

RR 0.70
(0.44 to 1.13)

3455
(4 trials)

⊕⊕⊝⊝
low1,8,11

Severe adverse effects12

‐

‐

‐

‐

(0 trials)

‐

*The basis for the assumed risk (eg the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 No serious risk of bias: The two most recent trials adequately described allocation concealment to be considered at low risk of selection bias.
2 No serious inconsistency: This finding was consistent across all trials and statistical heterogeneity was low.
3 Downgraded by 2 for very serious imprecision: These trials were not adequately powered to detect a difference in mortality. Only nine deaths occurred in these four trials. To confidently detect a 25% reduction in maternal mortality in a setting of 350 deaths/100,000 would require a sample size of over 100,000.
4 No serious risk of bias: One of these two trials adequately described allocation concealment to be at low risk of bias.
5 Downgraded by 1 for serious indirectness: Only a single trial from Mozambique provides data on the currently used regimen of IPT as two doses of SP. The definition of severe anaemia was PCV <21%.
6 Downgraded by 1 for serious imprecision: The number of events is very low and the trials underpowered to be confident in these results.
7 No serious indirectness: These trials were conducted in Thailand, Mozambique and Uganda between 1988 and 2008. The two recent trials administered IPT as two doses of SP. The definition of anaemia was variable; Hb < 11 g/dL, PCV < 33% and PCV <30%.
8 Downgraded by 1 for serious imprecision: Although the finding is of no effect. The 95% CI includes what may be clinically important differences.
9 Downgraded by 1 for serious inconsistency: The two old trials from 1957 and 1988 suggest clinically important benefits with chemoprophylaxis ‐ however, the two recent trials providing two doses of SP find no evidence of an effect.
10 Downgraded by 1 for serious indirectness: The finding of no effect in the two recent trials may be due to the declining efficacy of two doses of SP.
11 Downgraded for by 1 for serious inconsistency. There is substantive heterogeneity between trials (I2 = 79%), and this finding of no effect is in contrast to findings of benefit in both women of low parity and multigravidae. The finding of no effect in two of the recent trials may reflect declining efficacy in the regimens used.
12 Reporting of adverse events was generally poor. No severe adverse events were reported.

Open in table viewer
Summary of findings 6. Summary of findings table 6

Malaria chemoprevention for pregnant women (all parities) living in endemic areas: infant outcomes

Patient or population: Pregnant women (all parities)
Settings: Malaria‐endemic areas
Intervention: Malaria chemoprevention (any regimen)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

Chemoprevention

Spontaneous abortion

12 per 1000

11 per 1000
(7 to 16)

RR 0.89
(0.58 to 1.36)

5767
(3 trials)

⊕⊕⊝⊝
low1,2,3,4

Stillbirth

22 per 1000

22 per 1000
(17 to 30)

RR 1.02
(0.76 to 1.36)

7130
(5 trials)

⊕⊕⊕⊝
moderate1,2,5

Perinatal mortality

33 per 1000

41 per 1000
(31 to 54)

RR 1.24
(0.94 to 1.63)

5216
(4 trials)

⊕⊕⊕⊝
moderate1,2,5

Neonatal mortality

62 per 1000

56 per 1000
(44 to 72)

RR 0.91
(0.71 to 1.16)

6313
(5 trials)

⊕⊕⊕⊝
moderate1,2,5

Preterm birth

85 per 1000

81 per 1000
(55 to 117)

RR 0.95
(0.65 to 1.38)

1174
(2 trials)

⊕⊕⊝⊝
low2,5,6,10

Low birthweight

119 per 1000

126 per 1000
(106 to 151)

RR 1.06
(0.89 to 1.27)

3644
(4 trials)

⊕⊕⊝⊝
low1,2,5,10

Mean birthweight

The mean birthweight in the control groups ranged from

2797 g to 3161 g

The mean birthweight in the intervention groups was
0.54 g lower
(24.6 g lower to 23.6 g higher)

‐

6007
(5 trials)

⊕⊕⊕⊝
moderate1,7,8,10

Placental parasitaemia

181 per 1000

80 per 1000
(27 to 233)

RR 0.44
(0.15 to 1.29)

3200
(4 trials)

⊕⊕⊝⊝
low1,9,10

*The basis for the assumed risk (eg the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 No serious risk of bias: The two most recent trials adequately described allocation concealment to be considered at low risk of selection bias.
2 No serious inconsistency: The finding of no difference is consistent across trials and statistical heterogeneity is low
3 No serious indirectness: These trials were conducted in the Burkina Faso, Mozambique and Uganda between 1988 and 2008. One gave chemoprevention as weekly chloroquine and two trials gave IPT with SP.
4 Downgraded by 2 for very serious imprecision: The 95% CI is wide and sample remains underpowered to detect or rule out an effect.
5 Downgraded by 1 for serious imprecision: The 95% CI is wide and sample remains underpowered to detect or rule out an effect.
6 No serious risk of bias: The most recent trial adequately described allocation concealment to be considered at low risk of selection bias.
7 No serious inconsistency: Although substantial statistical heterogeneity is present (I2 = 72%), this relates to the oldest trial which found a benefit with chemoprevention. The subsequent four trials have consistently found no clinically important difference.
8 No serious imprecision: The 95% CI probably excludes clinically important benefits.
9 Downgraded by 1 for serious inconsistency: The two old trials from 1957 and 1988 suggest clinically important benefits with chemoprophylaxis ‐ however, the two recent trials providing two doses of SP find no evidence of an effect.
10 Downgraded by 1 for serious indirectness: The finding of no effect in the recent trials may be due to the declining efficacy of two doses of SP which is no longer recommended.

Open in table viewer
Summary of findings 7. Summary of findings table 7

Intermittent preventive treatment with SP for pregnant women (parity 0‐1) living in malaria endemic areas: maternal outcomes

Patient or population: Pregnant women (parity 0‐1)
Settings: Malaria‐endemic areas
Intervention: Intermittent preventive treatment with SP (2 doses, 3 doses, or monthly dosing)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

IPT (SP)

Mortality
All‐cause death

7 per 1000

8 per 1000
(3 to 20)

RR 1.15
(0.44 to 3.06)

2097
(2 trials)

⊕⊝⊝⊝
very low1,2

Severe anaemia
During the third trimester

145 per 1000

87 per 1000
(68 to 108)

RR 0.60
(0.47 to 0.75)

2503
(3 trials)

⊕⊕⊕⊕
high3,4,5,6

Anaemia

617 per 1000

543 per 1000
(480 to 604)

RR 0.88
(0.81 to 0.96)

3291
(4 trials)

⊕⊕⊕⊝
moderate1,6,7,8

Uncomplicated clinical malaria

9 per 100

2 per 100
(0 to 10)

RR 0.24
(0.05 to 1.12)

174
(1 trial)

⊕⊝⊝⊝
very low9,10,11

Antenatal parasitaemia

286 per 1000

108 per 1000
(69 to 169)

RR 0.38
(0.24 to 0.59)

2832
(4 trials)

⊕⊕⊕⊕
high3,6,7,12

Severe adverse effects13

‐

‐

‐

‐

(0 trials)

‐

*The basis for the assumed risk (eg the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 Downgraded by 1 for risk of bias: Only one of these trials adequately described allocation concealment to be considered at low risk of selection bias.
2 Downgraded by 2 for imprecision: These trials were not adequately powered to detect a difference in mortality. Only 15 deaths occurred in these two trials. To confidently detect a 50% reduction in maternal mortality in a setting of 350 deaths/100,000 would require a sample size of over 100,000.
3 No serious risk of bias: Exclusion of the trials at high risk of bias did not change the statistical significance or clinical importance of the result.
4 No serious inconsistency: This finding was consistent across all trials and statistical heterogeneity was low.
5 No serious indirectness: These trials were conducted in Kenya and Mozambique between 1996 and 2005, all three trials administered IPT with SP. The definition of severe anaemia was variable; Hb < 8 g/dL, Hb < 7g/dL, or PCV < 21%.
6 No serious imprecision: This result is statistically significant and the meta‐analysis is adequately powered to detect this effect.
7 No serious inconsistency: Although statistical heterogeneity was high, all trials favoured IPT with SP but there was variability in the size of the effect.
8 No serious indirectness: These trials were conducted Kenya between 1996 and 1999. The definition of anaemia was variable; Hb < 11 g/dL, Hb < 10 g/dL.
9 Downgraded by 1 for risk of bias: This trial is at unclear risk of selection bias.
10 Downgraded by 1 for indirectness: This trial from Mozambique 2002, measured fever history only as proxy for malaria illness.
11 Downgraded by 1 for serious imprecision: The 95% CI is wide and includes clinically important benefits and no effect.
12 No serious indirectness: These trials were conducted in the Kenya and Mozambique between 1996 and 2005.
13Reporting of adverse events was generally poor. No severe adverse events were reported.

Open in table viewer
Summary of findings 8. Summary of findings table 8

Intermittent preventive treatment with SP for pregnant women (parity 0‐1) living in malaria endemic areas: infant outcomes

Patient or population: Pregnant women (parity 0‐1)
Settings: Malaria‐endemic areas
Intervention: Intermittent preventive treatment with SP (2 doses, 3 doses, or monthly dosing)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

IPT (SP)

Spontaneous abortion

34 per 1000

21 per 1000
(13 to 33)

RR 0.61
(0.38 to 0.99)

2567
(3 trials)

⊕⊕⊝⊝
low1,2,3,4

Stillbirth

33 per 1000

32 per 1000
(21 to 49)

RR 0.97
(0.64 to 1.47)

2703
(3 trials)

⊕⊕⊝⊝
low2,4,5,6

Perinatal mortality

80 per 1000

62 per 1000
(42 to 94)

RR 0.78
(0.52 to 1.17)

1237
(1 trial)

⊕⊕⊝⊝
low7

Neonatal mortality

37 per 1000

23 per 1000
(14 to 39)

RR 0.62
(0.37 to 1.05)

2156
(2 trials)

⊕⊕⊝⊝
low2,4,5,6

Preterm birth

164 per 1000

140 per 1000
(108 to 181)

RR 0.85
(0.66 to 1.10)

1493
(2 trials)

⊕⊕⊝⊝
low1,2,4

Low birthweight

128 per 1000

104 per 1000
(86 to 127)

RR 0.81
(0.67 to 0.99)

3043
(4 trials)

⊕⊕⊕⊝
moderate8,9

Mean birthweight

The mean birthweight in the control groups ranged from

2908 g to 3079 g

The mean birthweight in the intervention groups was
84.18 g higher
(40.1 to 128.3 higher)

‐

2127
(3 trials)

⊕⊕⊕⊝
moderate5,9

Placental parasitaemia

225 per 1000

101 per 1000
(74 to 137)

RR 0.45
(0.33 to 0.61)

1633
(3 trials)

⊕⊕⊕⊝
moderate5,10

Cord blood haemoglobin

‐

‐

‐

‐

(0 trials)

‐

*The basis for the assumed risk (eg the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 Downgraded by 1 for serious risk of bias: None of the trials described adequate measures to prevent selection bias.
2 No serious inconsistency: The effect is consistent across trials and statistical heterogeneity is low
3 No serious indirectness: These trials were conducted in the Kenya and Mozambique between 1996 and 2002.
4 Downgraded by 1 for serious imprecision: The 95% CI is wide and sample remains underpowered to detect or rule out an effect.
5 Downgraded by 1 for serious risk of bias: Only one trial adequately described methods to prevent selection bias.
6 No serious indirectness: Trials were conducted in Kenya between 1996 and 1997.
7 Downgraded by 2 for serious imprecision: The 95% CI is wide and sample remains underpowered to detect or rule out an effect.
8 Downgraded by 1 for serious risk of bias: Only two of these trials were at low risk of selection bias.
9 No serious indirectness: These trials were conducted in the Kenya, Uganda and Mozambique between 1996 and 2008.
10 No serious inconsistency: Although statistical heterogeneity was high, all trials favoured chemoprevention but there was variability in the size of the effect.

Background

Description of the condition

Approximately 125 million women living in malaria‐endemic areas become pregnant each year (Dellicour 2010), and pregnancy is known to increase the risk of malaria infection and the severity of the illness compared to non‐pregnant women in the same age group (Desai 2007). Studies have also shown a strong association between malaria infection in pregnancy and consequent maternal anaemia, and low birthweight in infants, particularly in women in their first or second pregnancy (Desai 2007; Steketee 2001).

To reduce the burden and consequences of malaria in pregnancy, the World Health Organization (WHO) recommends that all pregnant women living in malaria‐endemic areas: i) sleep under a long lasting insecticide‐treated bednet (ITN; Gamble 2006; WHO 2012); ii) are treated when anaemic or when ill with malaria; and iii) receive some form of malaria chemoprevention. Currently the WHO recommends 'intermittent‐preventive therapy' with sulfadoxine‐pyrimethamine (SP) during the second and third trimesters in Africa (WHO 2013).

Description of the intervention

Over the years a variety of drugs have been evaluated for malaria chemoprevention in pregnancy, including amodiaquine, chloroquine, dapsone‐pyrimethamine, mefloquine, proguanil, pyrimethamine as monotherapy and as the fixed dose combination SP, and others. All have specific toxic and adverse effects, which are outlined in standard texts (WHO 2010), and these may be important factors influencing maternal adherence. For example, proguanil can cause mouth ulcers, chloroquine can cause itch, and mefloquine can cause dizziness and headaches.

How the intervention might work

Chemoprevention encompasses malaria chemoprophylaxis, and also the use of treatment courses given regularly to women. This is termed intermittent preventive treatment (IPT), defined as a full therapeutic course of antimalarial medicine given to pregnant women at routine prenatal visits, regardless of whether the recipient is infected with malaria. combines elements of a treatment effect through clearance or suppression of existing malaria infections in the placental and peripheral blood of mother, and a post‐treatment prophylactic effect by preventing new infections for several weeks after each dose (White 2005). Daily, weekly, or bi‐weekly malaria chemoprophylaxis is thought to work primarily through the prevention of new malaria infections. However, a reduction in malaria infections per se may be insufficient to justify the use of chemoprevention for widespread use without subsequent benefits on clinically important outcomes in the mother and her baby. These may include a reduction in clinical malaria episodes, a reduced risk of anaemia, improved birthweight, or more substantive outcomes such as a reduction in severe maternal illness, or fewer deaths in the mother and infant (see Figure 1).

Figure 1
Drugs for preventing malaria in pregnancy: conceptual framework.

Drugs for preventing malaria in pregnancy: conceptual framework.

The effects of malaria chemoprevention may differ between settings dependent on the local malaria epidemiology. In highly endemic areas with stable transmission, mothers may have partial immunity to malaria, and chronic subclinical placental infection are common leading to maternal anaemia and low birthweight, especially in primi‐ and secundigravidae. In contrast, where malaria transmission is low or unstable, the degree of life‐long acquired and pregnancy‐specific protective immunity may be lower and malaria infections are more likely to result in clinical episodes or severe illness, leading to low birthweight due to a preterm birth, foetal loss or maternal death.

Another potential effect modifier is HIV status. Many malaria‐endemic areas, especially in east and southern Africa, also have a high prevalence of HIV infection among pregnant women. Compared to HIV negative women, HIV positive women are more likely to carry malaria parasites in their blood, have higher parasite densities, and are more likely to have placental parasitaemia, anaemia, and malaria symptoms and deliver low birthweight babies (Nkhoma 2012a; Nkhoma 2012b; ter Kuile 2004).

Why it is important to do this review

This Cochrane Review aims to address the following questions:

  1. Does chemoprevention reduce mortality and substantive outcomes in the mother and infant?

  2. What is the potential reduction in the burden of malaria in pregnancy that can be achieved by successful malaria chemoprevention in pregnancy?

  3. Are the effects consistent in low parity and high parity women?

This review summarises the underpinning evidence of the protective efficacy achieved with antimalarial chemoprevention regimens on the effects on malaria and its consequences on the mother and baby when compared against placebo or no chemoprevention (case‐management strategies only). It does not compare different regimens. These were included in earlier editions of this Cochrane Review (Garner 2006); a more recent review has examined the effects of different IPT regimens in pregnant women (Kayentao 2013).

Objectives

In malaria‐endemic areas, to assess the effects in pregnant women of:

  1. Malaria chemoprevention versus no chemoprevention irrespective of the regimen;

  2. Malaria chemoprevention with SP (called intermittent preventive treatment) with no chemoprevention;

  3. Preventive regimens for Plasmodium vivax.

Methods

Criteria for considering studies for this review

Types of studies

Randomized controlled trials (RCTs) and quasi‐RCTs.

Types of participants

Pregnant women of any gravidity living in malaria‐endemic areas, defined as regions where transmission occurs and malaria is a characteristic of the region.

Types of interventions

Interventions

Any antimalarial drug chemoprevention regimen given to pregnant women.

Controls

Placebo or no intervention,

Types of outcome measures

For the conceptual framework, see Figure 1.

Maternal outcomes

  • Impact: maternal deaths (number of maternal deaths reported: death of a pregnant woman during pregnancy or within 42 days of termination of pregnancy).

  • Substantive outcomes: severe malaria, which includes severe anaemia (defined as Hb < 8 g/dL, < 7 g/dL, < 6 g/dL); severe adverse events.

  • Clinically important outcomes: anaemia (anaemia defined as Hb < between 10 and 12 g/dL); mean haemoglobin (g/dL) or mean PCV (%); clinical malaria (history of fever episodes prior to delivery); adverse events.

  • Indicators of malaria infection: parasitaemia (defined as the presence of asexual stage parasites in thick smears in peripheral, placental, or cord blood).

Infant outcomes

  • Impact: neonatal and Infant mortality.

  • Substantive outcomes: foetal loss (including spontaneous abortion (spontaneous expulsion of a fetus before it is able to survive independently); stillbirth (birth of a foetus with no vital signs, born after the 28th week of pregnancy); perinatal mortality; severe adverse events, including congenital anomalies (a defect that is present at birth).

  • Clinically important outcomes: preterm birth (delivery at < 37 weeks gestation); low birthweight (< 2500 g); mean birthweight; cord blood anaemia; adverse events.

  • Indicators of malaria infection: placental malaria; haemoglobin levels (infant), cord blood haemoglobin (g/dL), and cord blood PCV; cord blood parasitaemia.

Search methods for identification of studies

We attempted to identify all relevant trials regardless of language or publication status (published, unpublished, in press, and in progress).

Databases

We searched the following databases using the search terms and strategy described in Appendix 1: Cochrane Infectious Diseases Group Specialized Register (1 June 2014); Central Register of Controlled Trials (CENTRAL); MEDLINE (1966 to 1 June 2014); EMBASE (1974 to February 2012); and LILACS (1982 to February 2012).

Researchers

We contacted researchers working in the field for unpublished data, confidential reports, and raw data of published trials.

Reference lists

We also checked the citations of literature reviews, and of all trials identified by the above methods, and asked the referees to check the search strategy.

Data collection and analysis

Selection of studies

We applied inclusion criteria to all trials, including those in the previous edition of this Cochrane Review. DR‐P and PG independently screened all trials identified by the search strategy (Appendix 1). Using a form based on the inclusion criteria, DR‐P and PG assessed eligibility independently. FK checked the completeness of the included trials. We retrieved full text articles for all potentially relevant trials, applied the inclusion criteria, and then compared decisions. We resolved any differences by discussion and, when necessary, consulted with co‐authors.Trials identified in the initial abstract screening which did not meet the inclusion criteria are listed in the 'Characteristics of excluded studies'.

Data extraction and management

DR‐P and PG independently extracted data using a data extraction form. We extracted data on trial characteristics, including trial site, year, local malaria transmission and resistance, trial methods, participants, interventions, doses and outcomes and entered this data into Review Manager 5.1. The number of participants randomized and the number analysed in the experimental and control arms were extracted in each group for each outcome. For dichotomous outcomes, we recorded the number of participants experiencing the event and the number assessed in each treatment group. For continuous outcomes, we extracted the arithmetic means, standard deviations for each treatment group and the number of participants assessed in each group. We calculated and reported the loss to follow‐up in each group.

Assessment of risk of bias in included studies

We independently assessed the trials' methodological quality (risk of bias) of each trial, using the Cochrane Collaboration's tool for assessing the risk of bias (Higgins 2011). The following six components were assessed for each trial: generation of allocation sequence, allocation concealment, blinding (of participants, personnel, and outcome assessors), incomplete outcome data, selective outcome reporting, and other sources of bias. Each component was classified by 'yes' (low risk of bias), 'no' (high risk of bias), or 'unclear' to indicate level of bias. Where our judgement was 'unclear', we attempted to contact the trial authors for clarification.

Measures of treatment effect

We used the risk ratio (RR) to summarise dichotomous outcomes, reported the mean difference for continuous outcomes, and used the rate ratio for count outcomes. We presented all measures of effect with 95% confidence intervals (CI). One trial had four arms: one a comparison of IPT with nets, and a second comparison with no nets, and these were treated as separate comparisons (Njagi 2003i KEN; Njagi 2003ii KEN); a second trial had two intervention comparisons, so in meta‐analysis we split the control group in half for dichotomous outcomes. For continuous outcomes, we split the denominator of the control in half, but applied no correction to the standard deviation.

Unit of analysis issues

If the original trial analyses had not adjusted for clustering, we planned to adjust the results for clustering by multiplying the standard errors of the treatment effect by the square root of the design effect. The design effect would be calculated as 1+(m‐1)*ICC where m was the average cluster size and ICC was the intra‐cluster correlation coefficient. We planned to estimate the ICC from other trials included in the review or by contacting trial investigators. We also planned to include trials with multiple treatment arms if relevant to any of the comparisons. One trial randomized by compound in The Gambia (Greenwood 1989 GMB). However, we know that compounds are quite small, are grouped around families, and that, even if two women were pregnant at the same time in one family, this would not be quantitatively important in terms of overestimating the precision of the effect estimate.

Dealing with missing data

We planned to use intention‐to‐treat (ITT) data from the original trials, but it was more practical to use a complete‐case analysis, such that we excluded participants for whom no outcome was reported from the analysis. This analysis assumes that the participants for whom an outcome is available are representative of the original randomized patients. If data from the trial reports were insufficient, unclear, or missing, we attempted to contact the trial authors for additional information. In one trial with no standard deviation for birthweight, we used the average of the standard deviation for the other included trials.

Assessment of heterogeneity

We inspected the forest plots to detect overlapping CIs, applying the Chi2 test and a P value of 0.10 as the cut‐off value to determine statistical significance. We also estimated the I2‐statistic and categorized the degree of heterogeneity using standard cut‐offs (Higgins 2011).

Data synthesis

We used Review Manager 5.1 for the analysis.

Our primary analysis is stratified by parity, with results grouped into women of low parity (0‐1) and multigravidae (1+).

We included a category called 'all women'. This included trials that recruited women irrespective of parity. This analysis included the trials which had stratified the analysis by parity (and were therefore included in the primary analysis), and a second set of trials, which had not. This analysis provides information on the population effects of a policy of providing chemoprevention to all pregnant women.

We used RRs for dichotomous variables and mean differences (MD) for continuous variables; all results are presented with 95% CIs. In the absence of heterogeneity, we used a fixed‐effect model for the meta‐analysis, and where we detected heterogeneity we used a random‐effects model. Weighted averages were calculated where required. We converted Packed Cell Volume (PCV) values to haemoglobin values by dividing by three.

Subgroup analysis and investigation of heterogeneity

We grouped the analysis by parity. Although we intended to investigate heterogeneity by a variety of factors (including HIV status, risk of bias, geographical region, malaria transmission pattern, antimalarial resistance, ITN use, drug regimen), there were insufficient data to do this.

Results

Description of studies

Results of the search

The search was conducted up to 01 June 2014 for the time period 1964 to 2014, and identified 181 references of which two were duplicate trial reports. Out of 179, we retrieved 53 full‐text articles for eligibility screening (Figure 2).

Figure 2
Study flow diagram.

Study flow diagram.

Included studies

Seventeen chemoprevention trials, enrolling 20,256 pregnant women, met our inclusion criteria (see 'Characteristics of included studies'). These trials were conducted between 1957 and 2008, in Nigeria (three trials), The Gambia (three trials), Kenya (three trials), Mozambique (two trials), Uganda (two trials), Cameroon (one trial), Burkina Faso (one trial), and Thailand (two trials).

Six different antimalarials were evaluated against placebo or no preventive intervention (ie passive case detection and treatment of clinical cases only); chloroquine (given weekly), pyrimethamine (weekly or monthly), proguanil (daily), pyrimethamine‐dapsone (weekly or fortnightly), SP (given twice, monthly or intermittently for up to four doses at least one month apart), and mefloquine (weekly) (see Appendix 2). Fifteen trials reported that drug administration was supervised, and in two trials it was unsupervised (Fleming 1986 NGA; Ndyomugyenyi 2000 UGA).

Eight trials recruited women in all parity groups; four reported aggregate results, and four disaggregated by parity. The rest only recruited low parity women: six were parity 0, and two were women of parity 0‐1. One trial only recruited multigravidae (see Appendix 3).

In four trials, all women in both intervention and control groups received a long‐lasting ITNs at recruitment (Menendez 2008 MOZ; Ndyomugyenyi 2000 UGA; Ndyomugyenyi 2011 UGA; Njagi 2003i KEN). One additional trial mentioned that ITNs were in use in the area, with a use of 26% (Shulman 1999 KEN; ter Kuile 2007). In six trials iron and folic acid were routinely administered to all pregnant women (Fleming 1986 NGA; Mbaye 2006 GMB; Nahlen 1989 NGA; Njagi 2003i KEN; Njagi 2003ii KEN; Parise 1998i KEN; Parise 1998ii KEN; Villegas 2007 THA), in one trial only iron was administered (Shulman 1999 KEN), and in one trial both iron and folic acid were given to anaemic women (Nosten 1994 THA). The remaining trials did not comment on use of iron or folic acid.

One trial was randomized by compound, but for the analysis we assumed that it was individually randomized (Greenwood 1989 GMB). Two trials with multiple intervention arms were presented by individual arms, and the placebo patients split between the two arms where the treatment arms were both included in the meta‐analysis; Parise 1998i KEN compared two doses of SP versus no intervention while Parise 1998ii KEN compared monthly SP versus no intervention; Njagi 2003i KEN compared SP + ITNs versus placebo + ITNs; and Njagi 2003ii KEN compared SP alone versus placebo.

Excluded studies

We excluded 32 trials for the reasons given in the 'Characteristics of excluded studies' table. Also in this review update, we excluded one previously included trial (Hamilton 1972 UGA) as iron was administered to one of the control groups and folic acid to the other, but nothing was mentioned of iron and folates being administered to women in the intervention group (chloroquine).

Risk of bias in included studies

See Figure 3 for a summary of the risk of bias assessments. We have presented further details in the 'Characteristics of included studies' tables.

Figure 3
Risk of bias summary: review authors' judgements about each risk of bias item for each included trial.

Risk of bias summary: review authors' judgements about each risk of bias item for each included trial.

Allocation

Six trials adequately described methods of sequence generation and allocation concealment to be considered at low risk of selection bias (Fleming 1986 NGA; Mbaye 2006 GMB; Menendez 2008 MOZ; Ndyomugyenyi 2011 UGA; Shulman 1999 KEN; Villegas 2007 THA). Four trials were quasi‐RCT and so at high risk of selection bias (Cot 1992 BFA; Cot 1995 CMR; Morley 1964 NGA; Parise 1998i KEN; Parise 1998ii KEN), and in the remaining seven trials the risk was unclear.

Blinding

Eleven trials used placebo tablets, identical in taste and appearance to the active drug, and were assessed as having low risk of performance bias.

Four trials explicitly stated that outcome assessors were blinded and were assessed as having low risk of detection bias (Cot 1992 BFA; Morley 1964 NGA; Ndyomugyenyi 2011 UGA; Villegas 2007 THA). In the remaining included trials the risk was unclear.

Incomplete outcome data

Six trials had an attrition rate lower than 10% in both the intervention and control arm (Menendez 2008 MOZ; Morley 1964 NGA; Nahlen 1989 NGA; Ndyomugyenyi 2011 UGA; Nosten 1994 THA; Villegas 2007 THA). The remaining 11 trials were at high or unclear risk of attrition bias.

Selective reporting

Birthweight data were not available in one trial, but we obtained this data from a subsequent review (Njagi 2003i KEN; Njagi 2003ii KEN; ter Kuile 2007).

Other potential sources of bias

In one trial, 18 participants were replaced by others after randomization (Fleming 1986 NGA). We sought differences in baseline values with haemoglobin (Analysis 1.4) and detected no obvious difference.

Effects of interventions

See: Summary of findings for the main comparison Summary of findings table 1; Summary of findings 2 Summary of findings table 2; Summary of findings 3 Summary of findings table 3; Summary of findings 4 Summary of findings table 4; Summary of findings 5 Summary of findings table 5; Summary of findings 6 Summary of findings table 6; Summary of findings 7 Summary of findings table 7; Summary of findings 8 Summary of findings table 8

Comparison 1: Chemoprevention (any drug regimen) versus placebo/no chemoprevention

Chemoprevention for women in their first or second pregnancy

Maternal outcomes (see summary of findings Table for the main comparison)

Only 15 maternal deaths were reported across all trials with no difference between groups (three trials, 2097 participants, Analysis 1.1, very low quality evidence). Maternal death, even in these settings, is a relatively rare event occurring in less than five women per 1000 pregnancies. Consequently trials would need to enrol over 125,000 women to be adequately powered to detect or exclude effects as large as a 25% relative reduction (see Table 1).

Open in table viewer
Table 1. Optimal information size calculations: Chemoprevention versus placebo

Outcome

Assumed risk

Source

Clinically important relative reduction

Sample size required1,2

Maternal mortality

350/100,000

Analysis 1.1

25%

125228

Severe anaemia

150/1000

Analysis 1.2

25%

2540

Anaemia

650/1000

Analysis 1.3

25%

284

Malaria

170/1000

Analysis 1.5

25%

2194

Parasitaemia

290/1000

Analysis 1.6

25%

1124

Spontaneous abortions

32/1000

Analysis 1.9

25%

13348

Still births

33/1000

Analysis 1.10

25%

12932

Neonatal deaths

37/1000

Analysis 1.12

25%

11492

Preterm birth

160/1000

Analysis 1.13

25%

2356

Low birthweight

150/1000

Analysis 1.14

25%

2540

Placental parasitaemia

300/1000

Analysis 1.18

25%

1074

1 All calculations are based on: 2‐sided tests, with a ratio of 1:1, power of 0.8, and confidence level of 0.05.
2 All calculations were performed using: http://www.sealedenvelope.com/power/binary‐superiority

No trials reported on episodes of severe malaria, but three trials reported moderate to severe anaemia (defined as Hb < 7/8 g/dL or PCV < 21%). Overall, chemoprevention was associated with a 40% reduction in the risk of moderate to severe anaemia in the third trimester (RR 0.60, 95% CI 0.47 to 0.75; three trials, 2503 participants, Analysis 1.2, high quality evidence). This effect was consistent despite variation in doses, and differences in the definition and timing of assessment for severe anaemia (I2 =0); Parise 1998ii KEN recorded severe anaemia at delivery (after three doses of SP); Shulman 1999 KEN at 34 weeks (after three doses of SP); Menendez 2008 MOZ at delivery (after two doses of SP), and Parise 1998i KEN at the beginning of the third trimester clinic visit (when the second dose of SP was due, and these women had only had one SP dose).

Chemoprevention was also associated with a reduction in the risk of any anaemia (defined as Hb < 10/11/12 g/dL or PCV < 33%/30%), although this reduction was generally of smaller magnitude (RR 0.83, 95% CI 0.74 to 0.93; five trials, 3662 participants, Analysis 1.3, high quality evidence). In addition, measures of mean haemoglobin in the third trimester were higher in those receiving chemoprevention (MD 0.41 g/dL, 95% CI 0.29 to 0.54; five trials, 3363 participants, Analysis 1.4).

Chemoprevention was associated with fewer episodes of presumed clinical malaria (history of fever), but this outcome was only reported in two small trials (RR 0.37, 95% CI 0.18 to 0.74; two trials, 307 participants, Analysis 1.5, low quality evidence). Instead most trials reported antenatal parasitaemia, defined as either parasitaemia at delivery or parasitaemia at 34 to 36 weeks, with most trials showing benefits but wide variation in the size of the reduction (RR 0.39, 95% CI 0.26 to 0.58; eight trials, 3663 participants, I2 = 82; Analysis 1.6, high quality evidence) This heterogeneity is probably not unexpected given the differences in chemoprevention regimens and malaria endemicity.

Infant outcomes (see summary of findings Table 2).

The trials and the meta‐analyses are underpowered to confidently detect or exclude effects on spontaneous abortion, perinatal deaths, or neonatal deaths (see Table 1). The CIs range from important benefits to no evidence of any harm in four outcomes: spontaneous abortions (RR 0.65, 95% CI 0.41 to 1.02; five trials, 2876 participants, Analysis 1.9, low quality evidence); perinatal deaths (RR 0.73, 95% CI 0.54 to 1.00; two trials, 1620 participants, Analysis 1.11, low quality evidence); neonatal deaths (RR 0.62, 95% CI 0.37 to 1.05; two trials, 2156 participants, Analysis 1.12, low quality evidence). The preterm births analysis was (RR 0.85, 95% CI 0.66 to 1.10; two trials, 1493 participants, Analysis 1.13, low quality evidence).

Chemoprevention was associated with fewer low birthweight infants (RR 0.73, 95% CI 0.61 to 0.87; eight trials, 3619 participants, Analysis 1.14, moderate quality evidence). and mean birthweight was higher with chemoprevention (MD 92.72 g, 95% CI 62.05 to 123.39; nine trials, 3936 participants, Analysis 1.15, moderate quality evidence).

One very small trial reported no difference in the prevalence of cord blood anaemia (64 participants, Analysis 1.16), and a lower cord blood haemoglobin in babies born to women receiving chemoprevention (MD ‐1.80 g/dL, 95% CI ‐3.46 to ‐0.14; one trial, 64 participants, Analysis 1.17, very low quality evidence).

Chemoprevention resulted in fewer cases of placental parasitaemia (RR 0.54, 95% CI 0.43 to 0.69; seven trials, 2830 participants, Analysis 1.17, high quality evidence). Only one trial examined cord blood parasitaemia, but there were too few events to be confident of the result (RR 0.47, 95% CI 0.22 to 1.01; one trial, 1335 participants, Analysis 1.19). The children born to mothers receiving monthly SP had reduced cord parasitaemia, whereas those born to mothers receiving two doses of SP did not (Parise 1998i KEN).

Chemoprevention for multigravidae

Maternal outcomes (see summary of findings Table 3).

Four trials provided data on multigravidae women. Only one trial assessed mortality with six deaths in the chemoprevention group and four in the control group (RR 1.47, 95% CI 0.42 to 5.21; one trial, 2239 participants, Analysis 1.1, very low quality evidence).

No trials reported episodes of severe malaria, but two reported severe anaemia. In one trial more women had severe anaemia in the chemoprevention group (RR 1.20, 95% CI 0.91 to 1.57; one trial, 1954 participants), and the second trial had few events and consequently very wide CIs (RR 0.41, 95% CI 0.08 to 2.09; one trial, 728 participants). The 95% CIs of the overall meta‐analysis does not exclude effects as large as those seen in women in their first or second pregnancy but this is probably unlikely (RR 0.96, 95% CI 0.41 to 2.25; two trials, 2682 participants, Analysis 1.2).

No trials reported the risk of mild anaemia, but two trials reported mean haemoglobin at delivery without clinically important differences between groups (MD 0.01 g/dL, 95% CI ‐0.23 to 0.24; two trials, 676 participants, Analysis 1.4).

No trial measured malaria or febrile episodes in the mother. Four trials reported antenatal parasitaemia, and all four trials report large effects of a similar magnitude to those seen in women in their first or second pregnancy (RR 0.38, 95% CI 0.28 to 0.50; four trials, 3022 participants, Analysis 1.6, high quality evidence).

Infant outcomes (see summary of findings Table 4).

Two trials included information on infant outcomes after chemoprevention given to multigravid women.

Spontaneous abortions, stillbirths and perinatal deaths were not reported. One trial reported deaths in the first six weeks of life with slightly higher deaths following chemoprevention, but with wide CIs including the possibility of no difference between groups (RR 1.46, 95% CI 0.90 to 2.38; one trial, 2017 participants, Analysis 1.12).

No trials reported mean birthweight in infants born to multigravid women, but three reported the risk of low birthweight. The trend is in favour of chemoprevention but neither the trials, or the meta‐analysis reached standard levels of statistical significance (RR 0.86, 95% CI 0.64 to 1.17; three trials, 2743 participants, Analysis 1.14, very low quality evidence).

No trials reported measures of placental parasitaemia, cord blood parasitaemia, or cord blood haemoglobin.

Chemoprevention for all women

To evaluate the population effects of a policy of chemoprevention for all pregnant women, regardless of parity, this third analysis includes all trials which recruited women of any parity. Some of these presented results stratified by parity and were included in the analyses above, but a few additional trials did not provide their outcome data stratified by parity.

Maternal outcomes (see summary of findings Table 5).

For maternal mortality, only nine maternal deaths were recorded in trials recruiting women of all parities; 4/3019 with chemoprevention and 5/3007 without (four trials, 6026 participants, Analysis 1.1, low quality evidence).

For severe anaemia in the mother, there were very few events recorded in the two trials but the risk was lower with chemoprevention (RR 0.19, 95% CI 0.05 to 0.75; two trials, 1327 participants, Analysis 1.2, low quality evidence). For any anaemia, no population differences were demonstrated (RR 1.03, 95% CI 0.87 to 1.23; three trials, 3027 participants, Analysis 1.3, moderate quality evidence). Three trials reported mean haemoglobin, with only one very small trial from the early 1990s finding benefit with chemoprevention (three trials, 2223 participants, Analysis 1.4).

Clinical malaria (or history of fever) was reported in four of the trials across all parity groups. The older, and smaller trials, suggested a population benefit on clinical malaria but this was not seen in the two recent and much larger trials using two doses of SP (four trials, 3455 participants, Analysis 1.5, low quality evidence).

For parasitaemia at delivery, there was considerable heterogeneity between trials (I2 = 79%). Of the two most recent trials, both large, and both administering two doses of SP, one trial from Mozambique demonstrated a benefit with chemoprevention and one from Uganda did not (five trials, 3961 participants, Analysis 1.6, low quality evidence).

Infant outcomes (see summary of findings Table 6).

In trials recruiting women of all parities, no differences were demonstrated for spontaneous abortions (three trials, 5767 participants, Analysis 1.9, low quality evidence), stillbirths (five trials, 7130 participants, Analysis 1.10, moderate quality evidence), perinatal deaths (four trials, 5216 participants, Analysis 1.11, moderate quality evidence), or neonatal and infant deaths (five trials, 6313 participants, Analysis 1.12, moderate quality evidence). We also pooled across all trials for these outcomes (including those which only recruited women in their first or second pregnancies), and no differences were demonstrated.

Population benefits for the infants were not demonstrated for pre‐term birth (two trials, 1174 participants, Analysis 1.13, low quality evidence), low birthweight (four trials, 3644 participants, Analysis 1.14, low quality evidence), or mean birthweight (five trials, 6007 participants, Analysis 1.15, moderate quality evidence).

The effects of chemoprevention on placental parasitaemia were mixed (I2 = 94%), with large effects in two older trials administering monthly pyrimethamine or weekly chloroquine, and no effect demonstrated in the two more recent trials administering two doses of SP (four trials, 3200 participants, Analysis 1.18, low quality evidence).

One trial in Mozambique found a large effect in reducing the risk of cord blood anaemia (RR 0.49, 95% CI 0.30 to 0.80; one trial, 870 participants, Analysis 1.16), and increase in mean cord PCV (MD 1.01%, 95% CI 0.05 to 1.97; one trial, 990 participants, Analysis 1.17).

Adverse effects

We aggregated adverse effects across all parity groups. Reporting of adverse effects was generally poor. Only five trials specifically stated that no adverse effects attributable to the drugs were observed in the mothers, and the rest either did not report adverse effects or the information was unclear. Four trials reported adverse events following SP (Analysis 1.7), and one trial following mefloquine (Analysis 1.8). No differences were seen between the treatment and control groups.

Again, reporting of adverse events in the neonate was generally poor. Episodes of neonatal kernicterus were reported in two trials, and congenital anomalies in two trials, with no differences detected (Analysis 1.20).

Comparison 2. SP IPT chemoprevention for women in their first or second pregnancy

The above analysis examines the effects of drugs known to be effective in preventing malaria at the particular time the trials were carried out. As the WHO currently recommends intermittent dosing with SP, we performed an additional analysis to provide the effect estimates for SP compared to no drug or placebo. The analysis is exactly the same as comparison one, but we included only the six SP trials. These trials administered SP in two doses (Parise 1998i KEN; Njagi 2003i KEN; Njagi 2003ii KEN; Challis 2004 MOZ; Menendez 2008 MOZ; Ndyomugyenyi 2011 UGA), three doses (Shulman 1999 KEN), or monthly (Parise 1998ii KEN).

Maternal outcomes (see summary of findings Table 7).

For maternal death, no effect was demonstrated but the analysis is underpowered (Analysis 2.1).

For women of low parity, restricting the analysis to trials of SP did not substantially change the estimates of benefit on severe anaemia (RR 0.60, 95% CI 0.47 to 0.75; three trials, 2503 participants, Analysis 2.2, high quality evidence), mild anaemia (RR 0.88, 95% CI 0.88 to 0.96; three trials, 3219 participants, Analysis 2.3, moderate quality evidence), or mean haemoglobin (MD 0.41 g higher, 95% CI 0.27 to 0.54; three trials, 2995 participants, Analysis 2.4).

Similarly, the reduction in antenatal parasitaemia is consistent with the overall effect from trials of any chemoprevention (RR 0.38, 95% 0.24 to 0.59; four trials, 2832 participants, Analysis 2.5, high quality evidence), but there is insufficient data to draw conclusions on clinical malaria (RR 0.24, 95% CI 0.05 to 1.12; one trial, 174 participants, very low quality evidence (Analysis 2.6).

Infant outcomes (see summary of findings Table 8).

The trials and the meta‐analyses are underpowered to confidently detect or exclude effects on spontaneous abortion, stillbirth, perinatal deaths, or neonatal deaths, but restricting the analysis to trials of SP did not substantially change the estimates of effect (see Analysis 2.7; Analysis 2.8; Analysis 2.9; Analysis 2.10; low quality evidence). The trend is towards a reduction in pre‐term birth but the 95% CI is wide and includes the possibility of no effect (RR 0.85, 95% CI 0.66 to 1.10; two trials, 1493 participants, Analysis 2.11, low quality evidence).

Overall, chemoprevention with SP reduced the incidence of low birthweight but this effect seems to be reducing over time, with large effects in the older trials and no effect seen in the more recent trials using two doses of SP (four trials, 3043 participants, Analysis 2.12, moderate quality evidence). However, mean birthweight was higher with SP, and this effect was still present in the most recent trials (MD 105.5 g, 95% CI 68.02 to 142.9, four trials, 2693 participants, Analysis 2.13, moderate quality evidence).

Chemoprevention with SP reduced placental parasitaemia (RR 0.45, 95% CI 0.33 to 0.61; three trials, 1633 participants, Analysis 2.14, high quality evidence) but only one trial of SP reported cord parasitaemia (RR 0.47, 95% CI 0.22 to 1.01; one trial, 1335 participants, Analysis 2.15).

Adverse effects

No effects were detected with icterus (two trials, 2233 participants, Analysis 2.16) or congenital abnormalities (one trial, 1017 participants, Analysis 2.16).

Comparison 3. Chemoprevention for P. vivax

Only one trial reported on chemoprevention for P. vivax, conducted in Thailand with weekly prophylaxis with chloroquine. It was rated at low risk of bias on all criteria. It seemed to prevent completely all episodes of P. vivax malaria (RR 0.01, 95% CI 0.00 to 0.20; 942 participants, see Table 2), but had no effect on maternal anaemia, low birthweight, or mean birthweight. It was underpowered to assess effects on mortality.

Open in table viewer
Table 2. Chloroquine versus placebo (effect on P. vivax malaria)

Outcomes

Trials

Participants

Effect estimate

Comment

Death (mother)

1

951

Risk ratio 0.34 (0.01, 8.28)

‐

Severe anaemia

1

‐

‐

Not reported

Anaemia

1

951

Risk ratio 1.00 (0.92, 1.08)

Defined as PCV < 30%

Clinical malaria

1

‐

‐

Not reported

P. vivax parasitaemia

1

942

Risk ratio 0.01 (0.00, 0.20)

History of antenatal parasitaemia. Nine women censored (they had P.

falciparum infection prior to their first P. vivax episode)

Adverse effects with chloroquine

1

951

Risk ratio 2.03 (0.18, 22.31)

The 5 most commonly reported adverse events were headache, anorexia,

sleep disorder, dizziness and weakness. CQ group: drug suspended in two

cases (1 ‐ constipation,1‐ nausea)

One woman in the placebo group was complaining of visual problems

Spontaneous abortion

1

951

Risk ratio 0.71 (0.36, 1.39)

‐

Stillbirth

1

865

Risk ratio 0.24 (0.03, 2.17)

‐

Perinatal deaths

1

‐

‐

Not reported

Neonatal and infant mortality

1

‐

‐

Not reported

Preterm birth (All)

1

733

Risk ratio 0.93 (0.46, 1.85)

‐

Preterm birth (Para 0)

1

141

Risk ratio 2.41 (0.63, 9.24)

‐

Preterm birth (Para 2+)

1

592

Risk ratio 0.62 (0.26, 1.46)

‐

Low birthweight (All)

1

733

Risk ratio 1.02 (0.71, 1.46)

‐

Low birthweight (Para 0)

1

141

Risk ratio 1.20 (0.65, 2.21)

‐

Low birthweight (Para 2+)

1

592

Risk ratio 0.94 (0.60, 1.47)

‐

Mean birthweight (All)

1

733

Mean difference ‐8.20 (‐73.41, 57.02)

‐

Mean birthweight (Para 0)

1

141

Mean difference ‐36.00 (‐188.73, 116.73)

Mean (SD) 2741 ± 481 versus 2777 ± 435 in the CQ versus placebo group

Mean birthweight (Para 2+)

1

592

Mean difference ‐2.00 (‐74.12, 70.12)

Mean (SD) 2954 ± 423 versus 2956 ± 471 in the CQ versus placebo group

Placental malaria

1

‐

‐

Not reported

Cord blood haemoglobin

1

‐

‐

Not reported

Cord blood parasitaemia

1

‐

‐

Not reported

Adverse effects (baby)

1

864

Risk ratio 1.22 (0.33, 4.50)

Congenital anomalies: Amniotic banding, brachydactyly; anophthalmia,

Down's syndrome,; amniotic banding,

absent digit toes; two cleft lip, one cleft palate in the placebo group.

Discussion

Summary of main results

We included 17 trials, enrolling 14,481 pregnant women, in this Cochrane Review.

For women in their first or second pregnancy, malaria chemoprevention reduces the risk of moderate to severe anaemia by around 40% (high quality evidence), and the risk of any anaemia by around 17% (high quality evidence). Malaria chemoprevention reduces the risk of antenatal parasitaemia by around 61% (high quality evidence), and two trials reported a reduction in febrile illness (low quality evidence). There were only 16 maternal deaths and these trials were underpowered to detect an effect on maternal mortality (very low quality evidence).

For infants of women in their first and second pregnancies, malaria chemoprevention probably increases mean birthweight by around 93 g (moderate quality evidence), reduces low birthweight by around 27% (moderate quality evidence), and reduces placental parasitaemia by around 46% (high quality evidence). Fewer trials evaluated spontaneous abortions, still births, perinatal deaths, or neonatal deaths, and these analyses were underpowered to detect clinically important differences.

In multigravid women, chemoprevention has similar effects on antenatal parasitaemia (high quality evidence) but there are too few trials to evaluate effects on other outcomes.

In trials giving chemoprevention to all pregnant women irrespective of parity, the average effects of chemoprevention measured in all women indicated it may prevent severe anaemia (low quality evidence), but consistent benefits have not been shown for other outcomes.

In an analysis confined only to intermittent preventive therapy with SP, the estimates of effect and the quality of the evidence were similar.

A summary of a single trial in Thailand of prophylaxis against vivax showed chloroquine prevented vivax infection (RR 0.01, 95% CI 0.00 to 0.20; 942 participants).

Overall completeness and applicability of evidence

Trials were almost exclusively from Africa and published between 1964 and 2011. These trials, from a variety of settings and using varied chemoprevention regimens, found fairly consistent clinically important benefits for low parity women and their infants.

However, it is possible that with the introduction of ACTs, declining malaria transmission in some areas of Africa, and increasing quality of antenatal services, that the attributable fraction of malaria towards maternal anaemia and low birthweight has been reduced and the large effects seen in these trials may be attenuated by less malaria and better individualized care of women during pregnancy.

Quality of the evidence

The evidence for effects on maternal, foetal and neonatal mortality is generally considered of low or very low quality because the trials and the meta‐analysis remain significantly underpowered to confidently prove or exclude clinically important effects.

For women of low parity, we considered the evidence of clinically important effects on anaemia and antenatal parasitaemia to be of high quality, meaning we can have confidence in these results. For the infants of women of low parity, we considered the effects on birthweight to only be of moderate quality because of the high risk of bias of most of the older trials. This means we can have only moderate confidence in the magnitude of these effects.

Trials did not describe the routine health services available to detect and treat malaria infection in both intervention and control arms, but many trials were done some years ago in areas with very basic curative health services available. However, in the future with declining levels of malaria the individual management of illness and malaria at clinic may become an important option to control malaria in pregnancy.

Potential biases in the review process

It seems unlikely that we have missed any trials. As trials did not systematically document adverse effects, it is likely that these have been underestimated in this review.

Agreements and disagreements with other studies or reviews

The findings of this Cochrane Review are consistent with previous editions (Garner 2006; ter Kuile 2007). The findings are also consistent with the findings of a review comparing observational and randomized evidence (McClure 2013). McClure 2013 points out that the fairly modest effects seen in RCTs, where delivery of care is often strengthened and adherence assured, were attenuated in the observational studies where, the authors surmise, delivery of the intervention and adherence to it may be attenuated. However, this contrasts with a study estimating the effects of IPT with SP on low birthweight and neonatal mortality from survey data: the trial estimates are remarkably similar to the results observed with IPT with SP from the trial data reported in this and previous analysis (Eisele 2012; ter Kuile 2007).

Drugs for preventing malaria in pregnancy: conceptual framework.
Figures and Tables -
Figure 1

Drugs for preventing malaria in pregnancy: conceptual framework.

Navigate to figure in ReviewOpen in new tab

Study flow diagram.
Figures and Tables -
Figure 2

Study flow diagram.

Navigate to figure in ReviewOpen in new tab

Risk of bias summary: review authors' judgements about each risk of bias item for each included trial.
Figures and Tables -
Figure 3

Risk of bias summary: review authors' judgements about each risk of bias item for each included trial.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 1 Death (mother).
Figures and Tables -
Analysis 1.1

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 1 Death (mother).

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 2 Severe anaemia (mother).
Figures and Tables -
Analysis 1.2

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 2 Severe anaemia (mother).

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 3 Anaemia (mother).
Figures and Tables -
Analysis 1.3

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 3 Anaemia (mother).

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 4 Mean haemoglobin (g/dL).
Figures and Tables -
Analysis 1.4

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 4 Mean haemoglobin (g/dL).

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 5 Clinical malaria (mother).
Figures and Tables -
Analysis 1.5

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 5 Clinical malaria (mother).

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 6 Parasitaemia (mother).
Figures and Tables -
Analysis 1.6

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 6 Parasitaemia (mother).

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 7 Adverse effects with SP.
Figures and Tables -
Analysis 1.7

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 7 Adverse effects with SP.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 8 Adverse effects with mefloquine.
Figures and Tables -
Analysis 1.8

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 8 Adverse effects with mefloquine.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 9 Spontaneous abortion.
Figures and Tables -
Analysis 1.9

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 9 Spontaneous abortion.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 10 Stillbirth.
Figures and Tables -
Analysis 1.10

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 10 Stillbirth.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 11 Perinatal deaths.
Figures and Tables -
Analysis 1.11

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 11 Perinatal deaths.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 12 Neonatal and infant mortality.
Figures and Tables -
Analysis 1.12

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 12 Neonatal and infant mortality.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 13 Preterm birth.
Figures and Tables -
Analysis 1.13

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 13 Preterm birth.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 14 Low birthweight.
Figures and Tables -
Analysis 1.14

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 14 Low birthweight.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 15 Mean birthweight (baby).
Figures and Tables -
Analysis 1.15

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 15 Mean birthweight (baby).

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 16 Cord blood anaemia.
Figures and Tables -
Analysis 1.16

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 16 Cord blood anaemia.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 17 Cord blood haemoglobin.
Figures and Tables -
Analysis 1.17

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 17 Cord blood haemoglobin.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 18 Placental parasitemia (fetus).
Figures and Tables -
Analysis 1.18

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 18 Placental parasitemia (fetus).

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 19 Cord blood parasitaemia.
Figures and Tables -
Analysis 1.19

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 19 Cord blood parasitaemia.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 20 Adverse effects (baby).
Figures and Tables -
Analysis 1.20

Comparison 1 Preventive antimalarials versus placebo/no intervention, Outcome 20 Adverse effects (baby).

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 1 Death (mother).
Figures and Tables -
Analysis 2.1

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 1 Death (mother).

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 2 Severe anaemia (mother).
Figures and Tables -
Analysis 2.2

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 2 Severe anaemia (mother).

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 3 Anaemia (mother).
Figures and Tables -
Analysis 2.3

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 3 Anaemia (mother).

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 4 Mean haemoglobin (g/dL).
Figures and Tables -
Analysis 2.4

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 4 Mean haemoglobin (g/dL).

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 5 Parasitaemia (mother).
Figures and Tables -
Analysis 2.5

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 5 Parasitaemia (mother).

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 6 Clinical malaria (mother).
Figures and Tables -
Analysis 2.6

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 6 Clinical malaria (mother).

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 7 Spontaneous abortion.
Figures and Tables -
Analysis 2.7

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 7 Spontaneous abortion.

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 8 Stillbirth.
Figures and Tables -
Analysis 2.8

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 8 Stillbirth.

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 9 Perinatal deaths.
Figures and Tables -
Analysis 2.9

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 9 Perinatal deaths.

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 10 Neonatal and infant mortality.
Figures and Tables -
Analysis 2.10

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 10 Neonatal and infant mortality.

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 11 Preterm birth.
Figures and Tables -
Analysis 2.11

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 11 Preterm birth.

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 12 Low birthweight.
Figures and Tables -
Analysis 2.12

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 12 Low birthweight.

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 13 Mean birthweight (baby).
Figures and Tables -
Analysis 2.13

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 13 Mean birthweight (baby).

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 14 Placental parasitemia (fetus).
Figures and Tables -
Analysis 2.14

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 14 Placental parasitemia (fetus).

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 15 Cord blood parasitaemia.
Figures and Tables -
Analysis 2.15

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 15 Cord blood parasitaemia.

Navigate to figure in ReviewOpen in new tab

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 16 Adverse effects (baby).
Figures and Tables -
Analysis 2.16

Comparison 2 IPT with SP versus placebo/no intervention, Outcome 16 Adverse effects (baby).

Navigate to figure in ReviewOpen in new tab
Summary of findings for the main comparison. Summary of findings table 1

Malaria chemoprevention for pregnant women (parity 0‐1) living in endemic areas: maternal outcomes

Patient or population: Pregnant women (parity 0‐1)
Settings: Malaria‐endemic areas
Intervention: Malaria chemoprevention (any regimen)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

Chemoprevention

Mortality
All‐cause death

7 per 1000

8 per 1000
(3 to 20)

RR 1.15
(0.44 to 3.06)

2097
(3 trials)

⊕⊝⊝⊝
very low1,2

Severe anaemia
During the third trimester

145 per 1000

87 per 1000
(68 to 108)

RR 0.60
(0.47 to 0.75)

2503
(3 trials)

⊕⊕⊕⊕
high3,4,5,6

Anaemia

649 per 1000

539 per 1000
(480 to 604)

RR 0.83
(0.74 to 0.93)

3662
(5 trials)

⊕⊕⊕⊕
high3,6,7,8

Uncomplicated clinical malaria

173 per 1000

64 per 1000
(31 to 128)

RR 0.37
(0.18 to 0.74)

307
(2 trials)

⊕⊕⊝⊝
low4,9,10

Antenatal parasitaemia

286 per 1000

111 per 1000
(74 to 165)

RR 0.39
(0.26 to 0.58)

3663
(8 trials)

⊕⊕⊕⊕
high3,6,7,11

Severe adverse effects12

‐

‐

‐

‐

‐

*The basis for the assumed risk (eg, the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 Downgraded by 1 for risk of bias: Only one of these trials adequately described allocation concealment to be considered at low risk of selection bias.
2 Downgraded by 2 for imprecision: These trials were not adequately powered to detect a difference in mortality. Only 15 deaths occurred in these three trials.To confidently detect a 25% reduction in maternal mortality in a setting of 350 deaths/100,000 would require a sample size of over 100,000.
3 No serious risk of bias: Exclusion of the trials at high risk of bias did not change the statistical significance or clinical importance of the result.
4 No serious inconsistency: This finding was consistent across all trials and statistical heterogeneity was low.
5 No serious indirectness: These trials were conducted in Kenya and Mozambique between 1996 and 2005, all three trials administered IPT with SP. The definition of severe anaemia was variable; Hb < 8 g/dL, Hb < 7 g/dL, or PCV < 21%.
6 No serious imprecision: This result is statistically significant and the meta‐analysis is adequately powered to detect this effect.
7 No serious inconsistency: Although statistical heterogeneity was high, all trials favoured chemoprevention but there was variability in the size of the effect.
8 No serious indirectness: These trials were conducted in Nigeria, Kenya and Uganda between 1978 and 1999. Three trials administered IPT as SP, one gave weekly chloroquine, and one gave daily proguanil. The definition of anaemia was variable: Hb < 12 g/dL, Hb < 11 g/dL, Hb < 10 g/dL, PCV < 33% and PCV < 30%.
9 Downgraded by 1 for risk of bias. Both trials had high or unclear risk of selection bias and an attrition rate above 20%.
10 Downgraded by 1 for indirectness: Both these trials, from Cameroon 1993 and Mozambique 2002, measured fever history only as proxy for malaria illness.
11 Not downgraded for inconsistency. Despite substantive quantitative heterogeneity (I2 69% across six trials), all show at least a reduction of 23%, often more
11 No serious indirectness: These trials were conducted in The Gambia, Nigeria, Kenya and Mozambique between 1978 and 2005. Five trials gave IPT as SP, one gave pyrimethamine‐dapsone, one pyrimethamine, and one proguanil.
12 Reporting of adverse events was generally poor. No severe adverse events were reported.

Figures and Tables -
Summary of findings for the main comparison. Summary of findings table 1
Navigate to table in Review
Summary of findings 2. Summary of findings table 2

Malaria chemoprevention for pregnant women (parity 0‐1) living in endemic areas: infant outcomes

Patient or population: Pregnant women (parity 0‐1)
Settings: Malaria‐endemic areas
Intervention: Malaria chemoprevention (any regimen)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

Chemoprevention

Spontaneous abortion

33 per 1000

21 per 1000
(13 to 33)

RR 0.65
(0.41 to 1.02)

2876
(5 trials)

⊕⊕⊝⊝
low1,2,3,4

Stillbirth

33 per 1000

32 per 1000
(21 to 49)

RR 0.97
(0.64 to 1.49)

2703
(3 trials)

⊕⊕⊝⊝
low2,4,5,6,

Perinatal mortality

104 per 1000

76 per 1000
(56 to 104)

RR 0.73
(0.54 to 1.00)

1620
(2 trials)

⊕⊕⊝⊝
low2,4,5,7,

Neonatal mortality

37 per 1000

23 per 1000
(14 to 39)

RR 0.62
(0.37 to 1.05)

2156
(2 trials)

⊕⊕⊝⊝
low2,4,5,7,

Preterm birth

164 per 1000

140 per 1000
(108 to 181)

RR 0.85
(0.66 to 1.10)

1493
(2 trials)

⊕⊕⊝⊝
low1,2,4

Low birthweight

152 per 1000

110 per 1000
(92.7 to 132.2)

RR 0.73
(0.61 to 0.87)

3619
(8 trials)

⊕⊕⊕⊝
moderate9,10

Mean birthweight

The mean birthweight in the control groups ranged from

2723 g to 3079 g

The mean birthweight in the intervention groups was
92.72 g higher
(62.05 higher to 123.39 higher)

‐

3936
(9 trials)

⊕⊕⊕⊝
moderate5,10

Placental parasitaemia

307 per 1000

160 per 1000
(132 to 211)

RR 0.54
(0.43 to 0.69)

2830
(7 trials)

⊕⊕⊕⊕
high3,11,12

Cord blood haemoglobin

The mean haemoglobin in the control group was
15.8 g/dL

The mean haemoglobin in the intervention groups was
1.8 g/dL lower
(3.46 lower to 0.14 lower)

‐

64
(1 trial)

⊕⊝⊝⊝
very low1,13,14

*The basis for the assumed risk (eg the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 Downgraded by 1 for serious risk of bias: None of the trials described adequate measures to prevent selection bias.
2 No serious inconsistency: The effect is consistent across trials and statistical heterogeneity is low.
3 No serious indirectness: These trials were conducted in The Gambia, Cameroon, Kenya and Mozambique between 1990 and 2002. One gave chemoprevention as weekly chloroquine and four trials gave IPT with SP.
4 Downgraded by 1 for serious imprecision: The 95% CI is wide and sample remains underpowered to detect or rule out an effect.
5 Downgraded by 1 for serious risk of bias: Only one trial adequately described methods to prevent selection bias.
6 No serious indirectness: Trials were conducted in Cameroon and Kenya between 1993 and 1997. One trial gave weekly chloroquine and the others gave IPT as SP.
7 No serious indirectness: The trials were conducted in The Gambia and Kenya between 1984 and 1997. One trial used IPT with SP and one gave pyrimethamine‐dapsone which is no longer in use.
8 No serious indirectness: Both trials were conducted in Kenya and used IPT with SP.
9 Downgraded by 1 for serious risk of bias: Only two of these trials were at low risk of selection bias.
10 No serious indirectness: These trials were conducted in The Gambia, Cameroon, Kenya, Uganda and Mozambique between 1986 and 2005. The majority of trials used IPT with SP.
11 No serious inconsistency: Although statistical heterogeneity was high, all trials favoured chemoprevention but there was variability in the size of the effect.
12 No serious indirectness: These trials were conducted in The Gambia, Cameroon, Kenya, Uganda and Mozambique between 1990 and 2002. The majority of trials used IPT with SP.
13 Downgraded by 1 for serious indirectness: This single trial used a regimen that is no longer in use (proguanil).
14 Downgraded by 1 for serious imprecision: Only a single small trial has evaluated this comparison.

Figures and Tables -
Summary of findings 2. Summary of findings table 2
Navigate to table in Review
Summary of findings 3. Summary of findings table 3

Malaria chemoprevention for pregnant women (parity 2+) living in endemic areas: maternal outcomes

Patient or population: Pregnant women (parity 2+)
Settings: Malaria‐endemic areas
Intervention: Malaria chemoprevention (any regimen)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

Chemoprevention

Mortality
All‐cause death

5 per 1000

7 per 1000
(2 to 26

RR 1.47
(0.42 to 5.21)

2239
(1 trial)

⊕⊝⊝⊝
very low1,2,3

Severe anaemia
During the third trimester

68 per 1000

65 per 1000
(28 to 153)

RR 0.96
(0.41 to 2.25)

2682
(2 trials)

⊕⊕⊝⊝
low1,4,5

Anaemia

The mean PCV in the control group was
30.4 %

The mean PCV in the intervention group was
0.3 % higher
(0.7 lower to 1.3 higher)

‐

244
(1 trial)

⊕⊝⊝⊝
very low6,7,8

Uncomplicated clinical malaria

‐

‐

‐

‐
(0 trials)

‐

Antenatal parasitaemia

108 per 1000

41 per 1000
(30 to 54)

RR 0.38
(0.28 to 0.50)

3022
(4 trials)

⊕⊕⊕⊕
high9,10

Severe adverse events11

‐

‐

‐

‐

‐

*The basis for the assumed risk (eg the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 No serious risk of bias: These trials are at low risk of bias.
2 Downgraded by 1 for serious indirectness: This single trial was conducted in The Gambia between 2002 and 2004 and administered IPT as monthly SP. The findings may not be easily generalised to elsewhere.
3 Downgraded by 2 for very serious imprecision: Only ten deaths occurred in this trial. Much larger trials would be needed to detect or exclude effects on maternal mortality.
4 No serious indirectness: These two trials were conducted in The Gambia in 2002‐2004 and Mozambique between 2003 and 2005.
5 Downgraded by 2 for very serious imprecision: The 95% CI are very wide and include the possibility of both clinically important benefits and harms.
6 Downgraded by 1 for serious risk of bias: This single trial is at unclear risk of selection bias.
7 Downgraded by 1 for serious indirectness: This trial administered chemoprevention as pyrimethamine‐dapsone which is no longer in use.
8 Downgraded by 1 for serious imprecision: A much larger sample size is required to confidently detect or exclude an effect.
9 No serious risk of bias: Two of the four trials were at low risk of selection bias and exclusion of the other two trials did not change the size of the effect.
10 No serious indirectness: These three trials were conducted in The Gambia, Nigeria and Mozambique between 1986 and 2005. The biggest and most recent trial administered IPT with SP (two doses)

11 Reporting of adverse events was generally poor. No severe adverse events were reported.

Figures and Tables -
Summary of findings 3. Summary of findings table 3
Navigate to table in Review
Summary of findings 4. Summary of findings table 4

Malaria chemoprevention for pregnant women (parity 2+) living in endemic areas: infant outcomes

Patient or population: Pregnant women (parity 2+)
Settings: Malaria‐endemic areas
Intervention: Malaria chemoprevention (any regimen)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

Chemoprevention

Spontaneous abortion

‐

‐

‐

‐

(0 trials)

‐

Stillbirth

‐

‐

‐

‐

(0 trials)

‐

Perinatal deaths

‐

‐

‐

‐

(0 trials)

‐

Neonatal mortality

26 per 1000

38 per 1000
(23 to 62)

RR 1.46

(0.90 to 2.38)

2017

(1 trial)

⊕⊝⊝⊝
very low1,2,3

Preterm birth

‐

‐

‐

‐

(0 trials)

‐

Low birthweight

60 per 1000

63 per 1000
(46 to 85)

RR 0.86
(0.63 to 1.17)

2743
(3 trials)

⊕⊕⊝⊝
low3,4,5

Mean birthweight

‐

‐

‐

‐

(0 trials)

‐

Placental parasitaemia

‐

‐

‐

‐

(0 trials)

‐

Cord blood haemoglobin

‐

‐

‐

‐

(0 trials)

‐

*The basis for the assumed risk (eg the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 No serious risk of bias: This single trial was at low risk of selection bias.
2 Downgraded by 1 for serious indirectness: This single trial was conducted in The Gambia between 2002 and 2004 and administered IPT as monthly SP. The findings may not be easily generalised to elsewhere.
3 Downgraded by 2 for serious imprecision: The 95% CI is very wide and includes clinically important effects and no effect. A much larger sample size is required to confidently detect or exclude an effect.
4 No serious risk of bias: These trials are at low risk of selection bias.
5 No serious indirectness: These trials were conducted in The Gambia, Mozambique, and Uganda between 2002 and 2008.

Figures and Tables -
Summary of findings 4. Summary of findings table 4
Navigate to table in Review
Summary of findings 5. Summary of findings table 5

Malaria chemoprevention for all pregnant women (all parities) living in endemic areas: maternal outcomes

Patient or population: Pregnant women (all parities)
Settings: Malaria‐endemic areas
Intervention: Malaria chemoprevention (any regimen)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

Chemoprevention

Mortality
All‐cause death

1 per 1000

1 per 1000
(0 to 3)

RR 0.84
(0.25 to 2.74)

6026
(4 trials)

⊕⊕⊝⊝
low1,2,3

Severe anaemia
During the third trimester

26 per 1000

5 per 1000
(1 to 19)

RR 0.19
(0.05 to 0.75)

1327
(2 trials)

⊕⊕⊝⊝
low2,4,5,6

Anaemia

206 per 1000

212 per 1000
(179 to 253)

RR 1.03
(0.87 to 1.23)

3027
(3 trials)

⊕⊕⊕⊝
moderate1,2,7,8

Uncomplicated clinical malaria

114 per 1000

42 per 1000
(13 to 140)

RR 0.37
(0.11 to 1.23)

3455
(4 trials)

⊕⊕⊝⊝
low1,9,10

Antenatal parasitaemia

152 per 1000

106 per 1000
(67 to 172)

RR 0.70
(0.44 to 1.13)

3455
(4 trials)

⊕⊕⊝⊝
low1,8,11

Severe adverse effects12

‐

‐

‐

‐

(0 trials)

‐

*The basis for the assumed risk (eg the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 No serious risk of bias: The two most recent trials adequately described allocation concealment to be considered at low risk of selection bias.
2 No serious inconsistency: This finding was consistent across all trials and statistical heterogeneity was low.
3 Downgraded by 2 for very serious imprecision: These trials were not adequately powered to detect a difference in mortality. Only nine deaths occurred in these four trials. To confidently detect a 25% reduction in maternal mortality in a setting of 350 deaths/100,000 would require a sample size of over 100,000.
4 No serious risk of bias: One of these two trials adequately described allocation concealment to be at low risk of bias.
5 Downgraded by 1 for serious indirectness: Only a single trial from Mozambique provides data on the currently used regimen of IPT as two doses of SP. The definition of severe anaemia was PCV <21%.
6 Downgraded by 1 for serious imprecision: The number of events is very low and the trials underpowered to be confident in these results.
7 No serious indirectness: These trials were conducted in Thailand, Mozambique and Uganda between 1988 and 2008. The two recent trials administered IPT as two doses of SP. The definition of anaemia was variable; Hb < 11 g/dL, PCV < 33% and PCV <30%.
8 Downgraded by 1 for serious imprecision: Although the finding is of no effect. The 95% CI includes what may be clinically important differences.
9 Downgraded by 1 for serious inconsistency: The two old trials from 1957 and 1988 suggest clinically important benefits with chemoprophylaxis ‐ however, the two recent trials providing two doses of SP find no evidence of an effect.
10 Downgraded by 1 for serious indirectness: The finding of no effect in the two recent trials may be due to the declining efficacy of two doses of SP.
11 Downgraded for by 1 for serious inconsistency. There is substantive heterogeneity between trials (I2 = 79%), and this finding of no effect is in contrast to findings of benefit in both women of low parity and multigravidae. The finding of no effect in two of the recent trials may reflect declining efficacy in the regimens used.
12 Reporting of adverse events was generally poor. No severe adverse events were reported.

Figures and Tables -
Summary of findings 5. Summary of findings table 5
Navigate to table in Review
Summary of findings 6. Summary of findings table 6

Malaria chemoprevention for pregnant women (all parities) living in endemic areas: infant outcomes

Patient or population: Pregnant women (all parities)
Settings: Malaria‐endemic areas
Intervention: Malaria chemoprevention (any regimen)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

Chemoprevention

Spontaneous abortion

12 per 1000

11 per 1000
(7 to 16)

RR 0.89
(0.58 to 1.36)

5767
(3 trials)

⊕⊕⊝⊝
low1,2,3,4

Stillbirth

22 per 1000

22 per 1000
(17 to 30)

RR 1.02
(0.76 to 1.36)

7130
(5 trials)

⊕⊕⊕⊝
moderate1,2,5

Perinatal mortality

33 per 1000

41 per 1000
(31 to 54)

RR 1.24
(0.94 to 1.63)

5216
(4 trials)

⊕⊕⊕⊝
moderate1,2,5

Neonatal mortality

62 per 1000

56 per 1000
(44 to 72)

RR 0.91
(0.71 to 1.16)

6313
(5 trials)

⊕⊕⊕⊝
moderate1,2,5

Preterm birth

85 per 1000

81 per 1000
(55 to 117)

RR 0.95
(0.65 to 1.38)

1174
(2 trials)

⊕⊕⊝⊝
low2,5,6,10

Low birthweight

119 per 1000

126 per 1000
(106 to 151)

RR 1.06
(0.89 to 1.27)

3644
(4 trials)

⊕⊕⊝⊝
low1,2,5,10

Mean birthweight

The mean birthweight in the control groups ranged from

2797 g to 3161 g

The mean birthweight in the intervention groups was
0.54 g lower
(24.6 g lower to 23.6 g higher)

‐

6007
(5 trials)

⊕⊕⊕⊝
moderate1,7,8,10

Placental parasitaemia

181 per 1000

80 per 1000
(27 to 233)

RR 0.44
(0.15 to 1.29)

3200
(4 trials)

⊕⊕⊝⊝
low1,9,10

*The basis for the assumed risk (eg the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 No serious risk of bias: The two most recent trials adequately described allocation concealment to be considered at low risk of selection bias.
2 No serious inconsistency: The finding of no difference is consistent across trials and statistical heterogeneity is low
3 No serious indirectness: These trials were conducted in the Burkina Faso, Mozambique and Uganda between 1988 and 2008. One gave chemoprevention as weekly chloroquine and two trials gave IPT with SP.
4 Downgraded by 2 for very serious imprecision: The 95% CI is wide and sample remains underpowered to detect or rule out an effect.
5 Downgraded by 1 for serious imprecision: The 95% CI is wide and sample remains underpowered to detect or rule out an effect.
6 No serious risk of bias: The most recent trial adequately described allocation concealment to be considered at low risk of selection bias.
7 No serious inconsistency: Although substantial statistical heterogeneity is present (I2 = 72%), this relates to the oldest trial which found a benefit with chemoprevention. The subsequent four trials have consistently found no clinically important difference.
8 No serious imprecision: The 95% CI probably excludes clinically important benefits.
9 Downgraded by 1 for serious inconsistency: The two old trials from 1957 and 1988 suggest clinically important benefits with chemoprophylaxis ‐ however, the two recent trials providing two doses of SP find no evidence of an effect.
10 Downgraded by 1 for serious indirectness: The finding of no effect in the recent trials may be due to the declining efficacy of two doses of SP which is no longer recommended.

Figures and Tables -
Summary of findings 6. Summary of findings table 6
Navigate to table in Review
Summary of findings 7. Summary of findings table 7

Intermittent preventive treatment with SP for pregnant women (parity 0‐1) living in malaria endemic areas: maternal outcomes

Patient or population: Pregnant women (parity 0‐1)
Settings: Malaria‐endemic areas
Intervention: Intermittent preventive treatment with SP (2 doses, 3 doses, or monthly dosing)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

IPT (SP)

Mortality
All‐cause death

7 per 1000

8 per 1000
(3 to 20)

RR 1.15
(0.44 to 3.06)

2097
(2 trials)

⊕⊝⊝⊝
very low1,2

Severe anaemia
During the third trimester

145 per 1000

87 per 1000
(68 to 108)

RR 0.60
(0.47 to 0.75)

2503
(3 trials)

⊕⊕⊕⊕
high3,4,5,6

Anaemia

617 per 1000

543 per 1000
(480 to 604)

RR 0.88
(0.81 to 0.96)

3291
(4 trials)

⊕⊕⊕⊝
moderate1,6,7,8

Uncomplicated clinical malaria

9 per 100

2 per 100
(0 to 10)

RR 0.24
(0.05 to 1.12)

174
(1 trial)

⊕⊝⊝⊝
very low9,10,11

Antenatal parasitaemia

286 per 1000

108 per 1000
(69 to 169)

RR 0.38
(0.24 to 0.59)

2832
(4 trials)

⊕⊕⊕⊕
high3,6,7,12

Severe adverse effects13

‐

‐

‐

‐

(0 trials)

‐

*The basis for the assumed risk (eg the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 Downgraded by 1 for risk of bias: Only one of these trials adequately described allocation concealment to be considered at low risk of selection bias.
2 Downgraded by 2 for imprecision: These trials were not adequately powered to detect a difference in mortality. Only 15 deaths occurred in these two trials. To confidently detect a 50% reduction in maternal mortality in a setting of 350 deaths/100,000 would require a sample size of over 100,000.
3 No serious risk of bias: Exclusion of the trials at high risk of bias did not change the statistical significance or clinical importance of the result.
4 No serious inconsistency: This finding was consistent across all trials and statistical heterogeneity was low.
5 No serious indirectness: These trials were conducted in Kenya and Mozambique between 1996 and 2005, all three trials administered IPT with SP. The definition of severe anaemia was variable; Hb < 8 g/dL, Hb < 7g/dL, or PCV < 21%.
6 No serious imprecision: This result is statistically significant and the meta‐analysis is adequately powered to detect this effect.
7 No serious inconsistency: Although statistical heterogeneity was high, all trials favoured IPT with SP but there was variability in the size of the effect.
8 No serious indirectness: These trials were conducted Kenya between 1996 and 1999. The definition of anaemia was variable; Hb < 11 g/dL, Hb < 10 g/dL.
9 Downgraded by 1 for risk of bias: This trial is at unclear risk of selection bias.
10 Downgraded by 1 for indirectness: This trial from Mozambique 2002, measured fever history only as proxy for malaria illness.
11 Downgraded by 1 for serious imprecision: The 95% CI is wide and includes clinically important benefits and no effect.
12 No serious indirectness: These trials were conducted in the Kenya and Mozambique between 1996 and 2005.
13Reporting of adverse events was generally poor. No severe adverse events were reported.

Figures and Tables -
Summary of findings 7. Summary of findings table 7
Navigate to table in Review
Summary of findings 8. Summary of findings table 8

Intermittent preventive treatment with SP for pregnant women (parity 0‐1) living in malaria endemic areas: infant outcomes

Patient or population: Pregnant women (parity 0‐1)
Settings: Malaria‐endemic areas
Intervention: Intermittent preventive treatment with SP (2 doses, 3 doses, or monthly dosing)
Control: Placebo or no intervention

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(trials)

Quality of the evidence
(GRADE)

Assumed risk

Corresponding risk

Control

IPT (SP)

Spontaneous abortion

34 per 1000

21 per 1000
(13 to 33)

RR 0.61
(0.38 to 0.99)

2567
(3 trials)

⊕⊕⊝⊝
low1,2,3,4

Stillbirth

33 per 1000

32 per 1000
(21 to 49)

RR 0.97
(0.64 to 1.47)

2703
(3 trials)

⊕⊕⊝⊝
low2,4,5,6

Perinatal mortality

80 per 1000

62 per 1000
(42 to 94)

RR 0.78
(0.52 to 1.17)

1237
(1 trial)

⊕⊕⊝⊝
low7

Neonatal mortality

37 per 1000

23 per 1000
(14 to 39)

RR 0.62
(0.37 to 1.05)

2156
(2 trials)

⊕⊕⊝⊝
low2,4,5,6

Preterm birth

164 per 1000

140 per 1000
(108 to 181)

RR 0.85
(0.66 to 1.10)

1493
(2 trials)

⊕⊕⊝⊝
low1,2,4

Low birthweight

128 per 1000

104 per 1000
(86 to 127)

RR 0.81
(0.67 to 0.99)

3043
(4 trials)

⊕⊕⊕⊝
moderate8,9

Mean birthweight

The mean birthweight in the control groups ranged from

2908 g to 3079 g

The mean birthweight in the intervention groups was
84.18 g higher
(40.1 to 128.3 higher)

‐

2127
(3 trials)

⊕⊕⊕⊝
moderate5,9

Placental parasitaemia

225 per 1000

101 per 1000
(74 to 137)

RR 0.45
(0.33 to 0.61)

1633
(3 trials)

⊕⊕⊕⊝
moderate5,10

Cord blood haemoglobin

‐

‐

‐

‐

(0 trials)

‐

*The basis for the assumed risk (eg the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1 Downgraded by 1 for serious risk of bias: None of the trials described adequate measures to prevent selection bias.
2 No serious inconsistency: The effect is consistent across trials and statistical heterogeneity is low
3 No serious indirectness: These trials were conducted in the Kenya and Mozambique between 1996 and 2002.
4 Downgraded by 1 for serious imprecision: The 95% CI is wide and sample remains underpowered to detect or rule out an effect.
5 Downgraded by 1 for serious risk of bias: Only one trial adequately described methods to prevent selection bias.
6 No serious indirectness: Trials were conducted in Kenya between 1996 and 1997.
7 Downgraded by 2 for serious imprecision: The 95% CI is wide and sample remains underpowered to detect or rule out an effect.
8 Downgraded by 1 for serious risk of bias: Only two of these trials were at low risk of selection bias.
9 No serious indirectness: These trials were conducted in the Kenya, Uganda and Mozambique between 1996 and 2008.
10 No serious inconsistency: Although statistical heterogeneity was high, all trials favoured chemoprevention but there was variability in the size of the effect.

Figures and Tables -
Summary of findings 8. Summary of findings table 8
Navigate to table in Review
Table 1. Optimal information size calculations: Chemoprevention versus placebo

Outcome

Assumed risk

Source

Clinically important relative reduction

Sample size required1,2

Maternal mortality

350/100,000

Analysis 1.1

25%

125228

Severe anaemia

150/1000

Analysis 1.2

25%

2540

Anaemia

650/1000

Analysis 1.3

25%

284

Malaria

170/1000

Analysis 1.5

25%

2194

Parasitaemia

290/1000

Analysis 1.6

25%

1124

Spontaneous abortions

32/1000

Analysis 1.9

25%

13348

Still births

33/1000

Analysis 1.10

25%

12932

Neonatal deaths

37/1000

Analysis 1.12

25%

11492

Preterm birth

160/1000

Analysis 1.13

25%

2356

Low birthweight

150/1000

Analysis 1.14

25%

2540

Placental parasitaemia

300/1000

Analysis 1.18

25%

1074

1 All calculations are based on: 2‐sided tests, with a ratio of 1:1, power of 0.8, and confidence level of 0.05.
2 All calculations were performed using: http://www.sealedenvelope.com/power/binary‐superiority

Figures and Tables -
Table 1. Optimal information size calculations: Chemoprevention versus placebo
Navigate to table in Review
Table 2. Chloroquine versus placebo (effect on P. vivax malaria)

Outcomes

Trials

Participants

Effect estimate

Comment

Death (mother)

1

951

Risk ratio 0.34 (0.01, 8.28)

‐

Severe anaemia

1

‐

‐

Not reported

Anaemia

1

951

Risk ratio 1.00 (0.92, 1.08)

Defined as PCV < 30%

Clinical malaria

1

‐

‐

Not reported

P. vivax parasitaemia

1

942

Risk ratio 0.01 (0.00, 0.20)

History of antenatal parasitaemia. Nine women censored (they had P.

falciparum infection prior to their first P. vivax episode)

Adverse effects with chloroquine

1

951

Risk ratio 2.03 (0.18, 22.31)

The 5 most commonly reported adverse events were headache, anorexia,

sleep disorder, dizziness and weakness. CQ group: drug suspended in two

cases (1 ‐ constipation,1‐ nausea)

One woman in the placebo group was complaining of visual problems

Spontaneous abortion

1

951

Risk ratio 0.71 (0.36, 1.39)

‐

Stillbirth

1

865

Risk ratio 0.24 (0.03, 2.17)

‐

Perinatal deaths

1

‐

‐

Not reported

Neonatal and infant mortality

1

‐

‐

Not reported

Preterm birth (All)

1

733

Risk ratio 0.93 (0.46, 1.85)

‐

Preterm birth (Para 0)

1

141

Risk ratio 2.41 (0.63, 9.24)

‐

Preterm birth (Para 2+)

1

592

Risk ratio 0.62 (0.26, 1.46)

‐

Low birthweight (All)

1

733

Risk ratio 1.02 (0.71, 1.46)

‐

Low birthweight (Para 0)

1

141

Risk ratio 1.20 (0.65, 2.21)

‐

Low birthweight (Para 2+)

1

592

Risk ratio 0.94 (0.60, 1.47)

‐

Mean birthweight (All)

1

733

Mean difference ‐8.20 (‐73.41, 57.02)

‐

Mean birthweight (Para 0)

1

141

Mean difference ‐36.00 (‐188.73, 116.73)

Mean (SD) 2741 ± 481 versus 2777 ± 435 in the CQ versus placebo group

Mean birthweight (Para 2+)

1

592

Mean difference ‐2.00 (‐74.12, 70.12)

Mean (SD) 2954 ± 423 versus 2956 ± 471 in the CQ versus placebo group

Placental malaria

1

‐

‐

Not reported

Cord blood haemoglobin

1

‐

‐

Not reported

Cord blood parasitaemia

1

‐

‐

Not reported

Adverse effects (baby)

1

864

Risk ratio 1.22 (0.33, 4.50)

Congenital anomalies: Amniotic banding, brachydactyly; anophthalmia,

Down's syndrome,; amniotic banding,

absent digit toes; two cleft lip, one cleft palate in the placebo group.
Figures and Tables -
Table 2. Chloroquine versus placebo (effect on P. vivax malaria)
Navigate to table in Review
Comparison 1. Preventive antimalarials versus placebo/no intervention

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Death (mother) Show forest plot

9

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

1.1 Para 0‐1

4

2097

Risk Ratio (M‐H, Fixed, 95% CI)

1.15 [0.44, 3.06]

1.2 Multigravidae

1

2239

Risk Ratio (M‐H, Fixed, 95% CI)

1.47 [0.42, 5.21]

1.3 All women

4

6026

Risk Ratio (M‐H, Fixed, 95% CI)

0.84 [0.25, 2.74]

2 Severe anaemia (mother) Show forest plot

6

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

2.1 Para 0‐1

4

2503

Risk Ratio (M‐H, Random, 95% CI)

0.60 [0.47, 0.75]

2.2 Multigravidae

2

2682

Risk Ratio (M‐H, Random, 95% CI)

0.96 [0.41, 2.25]

2.3 All women

2

1327

Risk Ratio (M‐H, Random, 95% CI)

0.19 [0.05, 0.75]

3 Anaemia (mother) Show forest plot

10

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

3.1 Para 0‐1

7

3662

Risk Ratio (M‐H, Random, 95% CI)

0.83 [0.74, 0.93]

3.2 All women

3

3027

Risk Ratio (M‐H, Random, 95% CI)

1.03 [0.87, 1.23]

4 Mean haemoglobin (g/dL) Show forest plot

10

Mean Difference (IV, Fixed, 95% CI)

Subtotals only

4.1 Baseline Hb

5

3004

Mean Difference (IV, Fixed, 95% CI)

0.04 [‐0.10, 0.17]

4.2 Para 0‐1

7

3363

Mean Difference (IV, Fixed, 95% CI)

0.41 [0.29, 0.54]

4.3 Multigravidae

2

676

Mean Difference (IV, Fixed, 95% CI)

0.01 [‐0.23, 0.24]

4.4 All women

3

2223

Mean Difference (IV, Fixed, 95% CI)

0.13 [0.00, 0.25]

5 Clinical malaria (mother) Show forest plot

6

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

5.1 Para 0‐1

2

307

Risk Ratio (M‐H, Random, 95% CI)

0.37 [0.18, 0.74]

5.2 All women

4

3455

Risk Ratio (M‐H, Random, 95% CI)

0.37 [0.11, 1.23]

6 Parasitaemia (mother) Show forest plot

13

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

6.1 Para 0‐1

10

3663

Risk Ratio (M‐H, Random, 95% CI)

0.39 [0.26, 0.58]

6.2 Multigravidae

4

3022

Risk Ratio (M‐H, Random, 95% CI)

0.38 [0.28, 0.50]

6.3 All women

5

3961

Risk Ratio (M‐H, Random, 95% CI)

0.70 [0.44, 1.13]

7 Adverse effects with SP Show forest plot

5

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

7.1 Skin reactions

2

1472

Risk Ratio (M‐H, Fixed, 95% CI)

0.85 [0.27, 2.65]

7.2 Nausea and vomiting

2

1472

Risk Ratio (M‐H, Fixed, 95% CI)

1.69 [0.22, 12.81]

7.3 Any other adverse effects

3

2599

Risk Ratio (M‐H, Fixed, 95% CI)

0.72 [0.38, 1.36]

8 Adverse effects with mefloquine Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

8.1 Dizziness

1

107

Risk Ratio (M‐H, Fixed, 95% CI)

1.6 [0.90, 2.83]

8.2 Vertigo

1

339

Risk Ratio (M‐H, Fixed, 95% CI)

1.02 [0.81, 1.28]

8.3 Vomiting

1

339

Risk Ratio (M‐H, Fixed, 95% CI)

0.77 [0.53, 1.10]

8.4 Itching

1

339

Risk Ratio (M‐H, Fixed, 95% CI)

1.00 [0.73, 1.38]

8.5 Visual abnormalities

1

339

Risk Ratio (M‐H, Fixed, 95% CI)

1.00 [0.72, 1.39]

9 Spontaneous abortion Show forest plot

10

8643

Risk Ratio (M‐H, Fixed, 95% CI)

0.77 [0.56, 1.05]

9.1 Para 0‐1

7

2876

Risk Ratio (M‐H, Fixed, 95% CI)

0.65 [0.41, 1.02]

9.2 All women

3

5767

Risk Ratio (M‐H, Fixed, 95% CI)

0.89 [0.58, 1.36]

10 Stillbirth Show forest plot

9

9833

Risk Ratio (M‐H, Fixed, 95% CI)

1.01 [0.79, 1.28]

10.1 Para 0‐1

4

2703

Risk Ratio (M‐H, Fixed, 95% CI)

0.97 [0.63, 1.49]

10.2 All women

5

7130

Risk Ratio (M‐H, Fixed, 95% CI)

1.02 [0.76, 1.36]

11 Perinatal deaths Show forest plot

6

6836

Risk Ratio (M‐H, Fixed, 95% CI)

0.99 [0.81, 1.22]

11.1 Para 0‐1

2

1620

Risk Ratio (M‐H, Fixed, 95% CI)

0.73 [0.54, 1.00]

11.2 All women

4

5216

Risk Ratio (M‐H, Fixed, 95% CI)

1.24 [0.94, 1.63]

12 Neonatal and infant mortality Show forest plot

9

10486

Risk Ratio (M‐H, Fixed, 95% CI)

0.93 [0.76, 1.14]

12.1 Para 0‐1 (neonatal death: day 0‐28)

3

2156

Risk Ratio (M‐H, Fixed, 95% CI)

0.62 [0.37, 1.05]

12.2 Para 1+ (deaths up to six weeks)

1

2017

Risk Ratio (M‐H, Fixed, 95% CI)

1.46 [0.90, 2.38]

12.3 All women (neonatal and infant death: day 0‐1 year)

5

6313

Risk Ratio (M‐H, Fixed, 95% CI)

0.91 [0.71, 1.16]

13 Preterm birth Show forest plot

5

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

13.1 Para 0‐1

3

1493

Risk Ratio (M‐H, Fixed, 95% CI)

0.85 [0.66, 1.10]

13.2 All women

2

1174

Risk Ratio (M‐H, Fixed, 95% CI)

0.95 [0.65, 1.38]

14 Low birthweight Show forest plot

13

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

14.1 Para 0‐1

10

3619

Risk Ratio (M‐H, Fixed, 95% CI)

0.73 [0.61, 0.87]

14.2 Multigravidae

3

2743

Risk Ratio (M‐H, Fixed, 95% CI)

0.86 [0.65, 1.15]

14.3 All women

4

3644

Risk Ratio (M‐H, Fixed, 95% CI)

1.06 [0.89, 1.27]

15 Mean birthweight (baby) Show forest plot

15

Mean Difference (IV, Fixed, 95% CI)

Subtotals only

15.1 Para 0‐1

11

3936

Mean Difference (IV, Fixed, 95% CI)

92.72 [62.05, 123.39]

15.2 All women

5

6007

Mean Difference (IV, Fixed, 95% CI)

‐0.54 [‐24.66, 23.58]

16 Cord blood anaemia Show forest plot

2

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

16.1 Para 0‐1

1

64

Risk Ratio (M‐H, Fixed, 95% CI)

2.94 [0.78, 11.05]

16.2 All women

1

870

Risk Ratio (M‐H, Fixed, 95% CI)

0.49 [0.30, 0.80]

17 Cord blood haemoglobin Show forest plot

2

Mean Difference (IV, Fixed, 95% CI)

Subtotals only

17.1 Para 0‐1

1

64

Mean Difference (IV, Fixed, 95% CI)

‐1.80 [‐3.46, ‐0.14]

17.2 All women

1

990

Mean Difference (IV, Fixed, 95% CI)

1.01 [0.05, 1.97]

18 Placental parasitemia (fetus) Show forest plot

13

Risk Ratio (M‐H, Random, 95% CI)

Subtotals only

18.1 Para 0‐1

9

2830

Risk Ratio (M‐H, Random, 95% CI)

0.54 [0.43, 0.69]

18.2 All women

4

3200

Risk Ratio (M‐H, Random, 95% CI)

0.44 [0.15, 1.29]

19 Cord blood parasitaemia Show forest plot

3

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

19.1 Para 0‐1

2

1335

Risk Ratio (M‐H, Fixed, 95% CI)

0.47 [0.22, 1.01]

19.2 All women

1

2629

Risk Ratio (M‐H, Fixed, 95% CI)

0.73 [0.47, 1.14]

20 Adverse effects (baby) Show forest plot

5

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

20.1 Neonatal icterus

3

2233

Risk Ratio (M‐H, Fixed, 95% CI)

0.84 [0.63, 1.13]

20.2 Congenital anomalies

2

1328

Risk Ratio (M‐H, Fixed, 95% CI)

3.53 [0.58, 21.33]
Figures and Tables -
Comparison 1. Preventive antimalarials versus placebo/no intervention
Navigate to table in Review
Comparison 2. IPT with SP versus placebo/no intervention

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Death (mother) Show forest plot

3

1926

Risk Ratio (M‐H, Fixed, 95% CI)

0.99 [0.30, 3.22]

2 Severe anaemia (mother) Show forest plot

4

2503

Risk Ratio (M‐H, Random, 95% CI)

0.60 [0.47, 0.75]

3 Anaemia (mother) Show forest plot

5

3219

Risk Ratio (M‐H, Random, 95% CI)

0.88 [0.81, 0.96]

4 Mean haemoglobin (g/dL) Show forest plot

5

2995

Mean Difference (IV, Fixed, 95% CI)

0.41 [0.27, 0.54]

5 Parasitaemia (mother) Show forest plot

7

3456

Risk Ratio (M‐H, Random, 95% CI)

0.38 [0.24, 0.59]

6 Clinical malaria (mother) Show forest plot

1

174

Risk Ratio (M‐H, Random, 95% CI)

0.24 [0.05, 1.12]

7 Spontaneous abortion Show forest plot

5

2572

Risk Ratio (M‐H, Fixed, 95% CI)

0.61 [0.38, 0.99]

8 Stillbirth Show forest plot

3

2572

Risk Ratio (M‐H, Fixed, 95% CI)

0.96 [0.62, 1.50]

9 Perinatal deaths Show forest plot

1

1237

Risk Ratio (M‐H, Fixed, 95% CI)

0.78 [0.52, 1.17]

10 Neonatal and infant mortality Show forest plot

3

2156

Risk Ratio (M‐H, Fixed, 95% CI)

0.62 [0.37, 1.05]

11 Preterm birth Show forest plot

3

1493

Risk Ratio (M‐H, Fixed, 95% CI)

0.85 [0.66, 1.10]

12 Low birthweight Show forest plot

7

3043

Risk Ratio (M‐H, Fixed, 95% CI)

0.81 [0.67, 0.99]

13 Mean birthweight (baby) Show forest plot

6

2693

Mean Difference (IV, Fixed, 95% CI)

105.50 [68.02, 142.98]

14 Placental parasitemia (fetus) Show forest plot

6

2257

Risk Ratio (M‐H, Random, 95% CI)

0.45 [0.33, 0.61]

15 Cord blood parasitaemia Show forest plot

2

1335

Risk Ratio (M‐H, Fixed, 95% CI)

0.47 [0.22, 1.01]

16 Adverse effects (baby) Show forest plot

4

3250

Risk Ratio (M‐H, Fixed, 95% CI)

0.85 [0.64, 1.15]

16.1 Neonatal icterus

3

2233

Risk Ratio (M‐H, Fixed, 95% CI)

0.84 [0.63, 1.13]

16.2 Congenital anomalies

1

1017

Risk Ratio (M‐H, Fixed, 95% CI)

2.94 [0.12, 71.90]
Figures and Tables -
Comparison 2. IPT with SP versus placebo/no intervention
Navigate to table in Review