Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Review - Intervention

Mass drug administration for malaria

Authors' declarations of interest

Version published: 29 September 2021 Version history

https://doi.org/10.1002/14651858.CD008846.pub3
Collapse all Expand all

Background

Studies evaluating mass drug administration (MDA) in malarious areas have shown reductions in malaria immediately following the intervention. However, these effects vary by endemicity and are not sustained. Since the 2013 version of this Cochrane Review on this topic, additional studies have been published.

Objectives

Primary objectives

To assess the sustained effect of MDA with antimalarial drugs on:

‐ the reduction in malaria transmission in moderate‐ to high‐transmission settings;

‐ the interruption of transmission in very low‐ to low‐transmission settings.

Secondary objective

To summarize the risk of drug‐associated adverse effects following MDA.

Search methods

We searched several trial registries, citation databases, conference proceedings, and reference lists for relevant articles up to 11 February 2021. We also communicated with researchers to identify additional published and unpublished studies.

Selection criteria

Randomized controlled trials (RCTs) and non‐randomized studies comparing MDA to no MDA with balanced co‐interventions across study arms and at least two geographically distinct sites per study arm.

Data collection and analysis

Two review authors independently assessed trials for eligibility and extracted data. We calculated relative risk (RR) and rate ratios with corresponding 95% confidence intervals (CIs) to compare prevalence and incidence, respectively, in MDA compared to no‐MDA groups. We stratified analyses by malaria transmission and by malaria species. For cluster‐randomized controlled trials (cRCTs), we adjusted standard errors using the intracluster correlation coefficient. We assessed the certainty of the evidence using the GRADE approach. For non‐randomized controlled before‐and‐after (CBA) studies, we summarized the data using difference‐in‐differences (DiD) analyses.

Main results

Thirteen studies met our criteria for inclusion. Ten were cRCTs and three were CBAs.

Cluster‐randomized controlled trials

Moderate‐ to high‐endemicity areas (prevalence ≥ 10%)

We included data from two studies conducted in The Gambia and Zambia. 

At one to three months after MDA, the Plasmodium falciparum (hereafter, P falciparum) parasitaemia prevalence estimates may be higher compared to control but the CIs included no effect (RR 1.76, 95% CI 0.58 to 5.36; Zambia study; low‐certainty evidence); parasitaemia incidence was probably lower (RR 0.61, 95% CI 0.40 to 0.92; The Gambia study; moderate‐certainty evidence); and confirmed malaria illness incidence may be substantially lower,  but the CIs included no effect (rate ratio 0.41, 95% CI 0.04 to 4.42; Zambia study; low‐certainty evidence). 

At four to six months after MDA, MDA showed little or no effect on P falciparum parasitaemia prevalence (RR 1.18, 95% CI 0.89 to 1.56; The Gambia study; moderate‐certainty evidence) and, no persisting effect was demonstrated with parasitaemia incidence (rate ratio 0.91, 95% CI 0.55 to 1.50; The Gambia study).

Very low‐ to low‐endemicityareas (prevalence < 10%)

Seven studies from Cambodia, Laos, Myanmar (two studies), Vietnam, Zambia, and Zanzibar evaluated the effects of multiple rounds of MDA on P falciparum. Immediately following MDA (less than one month after MDA), parasitaemia prevalence was reduced (RR 0.12, 95% CI 0.03 to 0.52; one study; low‐certainty evidence). At one to three months after MDA, there was a reduction in both parasitaemia incidence (rate ratio 0.37, 95% CI 0.21 to 0.55; 1 study; moderate‐certainty evidence) and prevalence (RR 0.25, 95% CI 0.15 to 0.41; 7 studies; low‐certainty evidence). For confirmed malaria incidence, absolute rates were low, and it is uncertain whether MDA had an effect on this outcome (rate ratio 0.58, 95% CI 0.12 to 2.73; 2 studies; very low‐certainty evidence). 

For P falciparum prevalence, the relative differences declined over time, from RR 0.63 (95% CI 0.36 to 1.12; 4 studies) at four to six months after MDA, to RR 0.86 (95% CI 0.55 to 1.36; 5 studies) at 7 to 12 months after MDA. Longer‐term prevalence estimates showed overall low absolute risks, and relative effect estimates of the effect of MDA on prevalence varied from RR 0.82 (95% CI 0.20 to 3.34) at 13 to 18 months after MDA, to RR 1.25 (95% CI 0.25 to 6.31) at 31 to 36 months after MDA in one study.

Five studies from Cambodia, Laos, Myanmar (2 studies), and Vietnam evaluated the effect of MDA on Plasmodium vivax (hereafter, P vivax). One month following MDA, P vivax prevalence was lower (RR 0.18, 95% CI 0.08 to 0.40; 1 study; low‐certainty evidence). At one to three months after MDA, there was a reduction in P vivax prevalence (RR 0.15, 95% CI 0.10 to 0.24; 5 studies; low‐certainty evidence). The immediate reduction on P vivax prevalence was not sustained over time, from RR 0.78 (95% CI 0.63 to 0.95; 4 studies) at four to six months after MDA, to RR 1.12 (95% CI 0.94 to 1.32; 5 studies) at 7 to 12 months after MDA. One of the studies in Myanmar provided estimates of longer‐term effects, where overall absolute risks were low, ranging from RR 0.81 (95% CI 0.44 to 1.48) at 13 to 18 months after MDA, to RR 1.20 (95% CI 0.44 to 3.29) at 31 to 36 months after MDA.

Non‐randomized studies

Three CBA studies were conducted in moderate‐ to high‐transmission areas in Burkina Faso, Kenya, and Nigeria. There was a reduction in P falciparum parasitaemia prevalence in MDA groups compared to control groups during MDA (DiD range: ‐15.8 to ‐61.4 percentage points), but the effect varied at one to three months after MDA (DiD range: 14.9 to ‐41.1 percentage points). 

Authors' conclusions

In moderate‐ to high‐transmission settings, no studies reported important effects on P falciparum parasitaemia prevalence within six months after MDA. In very low‐ to low‐transmission settings, parasitaemia prevalence and incidence were reduced initially for up to three months for both P falciparum and P vivax; longer‐term data did not demonstrate an effect after four months, but absolute risks in both intervention and control groups were low. No studies provided evidence of interruption of malaria transmission. 

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.

Administration of antimalarial drugs to whole populations for reducing malaria

What is mass drug administration (MDA) for malaria?

MDA for malaria consists of giving a full treatment course of antimalarial medicine (even to persons with no symptoms of malaria and regardless of whether malaria is present) to every member of a defined population or every person living in a defined geographical area (except to those for whom the medicine could be harmful) at approximately the same time and often at repeated intervals.

How can MDA reduce malaria transmission in a population?

The life cycle of the malaria parasite consists of human liver, human blood, and mosquito stages. Malaria infection begins with the bite of an Anopheles species mosquito carrying the malaria parasite. During the bite, the infective mosquito injects the malaria parasite into the human host. After initially replicating in the liver, the parasites are released into the bloodstream. During the blood stage, parasites multiply in red blood cells, sometimes causing fever and other symptoms characteristic of malaria. Some of these parasites become a form which is infectious to mosquitoes. When the infected person is bitten again, the mosquito ingests blood containing the parasites, which then restarts the transmission cycle.

MDA with antimalarial drugs temporarily prevents new and clears existing malaria infections in the population. Depending on the characteristics of the antimalarial drug used, MDA targets parasites at different stages, which can lead to reduced disease burden and transmission in the population. Whether MDA can successfully reduce or interrupt malaria transmission may depend on how much malaria there is in the area; the use of other tools to control malaria, including preventing mosquito bites; the proportion of the population who receive at least one round of MDA; population movement; and when MDA rounds occur in relation to the peak malaria transmission season.

What was the aim of the review?

To guide policy‐making and future research for malaria control and elimination, the aim of this review was to update the evidence evaluating the effect of MDA compared to no MDA on malaria outcomes in moderate‐ to high‐transmission settings and very low‐ to low‐transmission settings. Our search of relevant published and unpublished literature identified 13 studies that met our inclusion criteria.

What are the main findings of the review?

Malaria burden was compared in people receiving MDA and those who did not receive MDA, at different time points. The findings differed by malaria transmission setting. In areas with malaria prevalence of 10% or higher (moderate‐ to high‐transmission areas), based on one trial, MDA did not reduce malaria in the population (low‐certainty evidence). In areas with malaria prevalence under 10% (very low‐ to low‐endemicity areas), evidence from seven trials indicates that MDA reduced malaria in the population immediately after MDA has stopped (low‐certainty evidence), but we are uncertain if the decline continues long‐term because the number of malaria cases in both intervention and control groups were low (very low‐certainty evidence).

In all settings of malaria transmission, the type of antimalarial drug used for MDA, co‐interventions such as mosquito control, coverage of MDA, and risk of re‐introduction may have an impact on the effect of MDA compared to no MDA. However, we were unable to explore these factors due to the limited number of studies.

How up to date is the review?

We included studies available up to 11 February 2021.

Authors' conclusions

Implications for practice

In moderate‐ to high‐transmission settings, only two studies contributed data to assess the effect of MDA on outcomes. Based on results from a single trial, MDA probably reduces parasitaemia incidence, but does not reduce parasitaemia prevalence at one to three months after MDA. However, it is worth noting that there was a large overall reduction in parasitaemia prevalence in both the intervention and control arms from baseline to post‐MDA. The second trial showed no effect of a single round of MDA at four to six months after MDA. Given the absence of data in moderate‐ to high‐transmission settings at time points after six months, we were unable to determine the longer‐term effects of MDA on malaria transmission.

In very low‐ to low‐transmission settings, MDA probably reduces P falciparum parasitaemia incidence at under one month, and P falciparum and P vivax prevalence at one to three months after MDA. The short‐ and long‐term effects of MDA on P falciparum and P vivax parasitaemia prevalence at time periods after four months is uncertain due to very low‐certainty evidence, but the immediate large reduction in parasitaemia prevalence is not sustained over time. Based on data provided in studies conducted in very low‐ to low‐transmission settings, we did not find evidence in any study of interruption of transmission as measured by a reduction to zero indigenous cases following MDA.

Other variables, such as type of antimalarial drug, MDA coverage, number of rounds, and co‐interventions, may affect the impact of MDA on malaria outcomes and should be considered when conducting MDA. Additionally, the degree of population mobility and potential for importation of parasites also plays an important role in the effect of MDA. These considerations should be weighed carefully in recommendations surrounding MDA.

Our findings in very low‐ to low‐transmission settings support the existing WHO Malaria Policy Advisory Committee's (MPAC) 2015 recommendations on the use of MDA in areas approaching elimination with high coverage of vector control and surveillance, good access to treatment, and limited risk of re‐introduction of infection (WHO 2015a). These recommendations are currently being updated through a revised guideline development process at WHO (WHO 2020b).

Implications for research

Given the addition of several cRCTs since the publication of the previous review on this topic (Poirot 2013), this updated review provides additional information about MDA in the context of a renewed interest in MDA as a strategy to accelerate progress towards malaria elimination. Although several studies, conducted more recently in very low‐ to low‐endemicity settings, attempted to collect data on outcomes at longer time points following MDA, the certainty of the evidence on the sustained effect of MDA was very low due to high risk of bias and large imprecision. Although of higher certainty evidence compared to trials conducted in very low‐ to low‐endemicity settings, none of the included studies in moderate‐ to high‐endemicity settings measured the effect of MDA after four to six months. Future studies should measure the longer‐term effect of MDA and ensure that outcomes from a sufficient number and representative sample of participants are collected to obtain more precise estimates of effect. In relation to study design, cRCTs should be designed with a sufficient number of clusters to help to ensure that measured and unmeasured confounders are balanced across randomized arms, studies designed for interrupted time series analysis should include sufficient pre‐ and post‐intervention data to adequately capture seasonal malaria trends, and co‐interventions should be balanced across study arms.

Summary of findings

Open in table viewer
Summary of findings 1. MDA compared to no MDA for Plasmodium falciparum malaria (moderate to high endemicity, short‐term follow‐up)

Patient or population: People of all ages living in an area with moderate to high endemicity of P falciparum malaria (≥ 10% prevalence)
Setting: Moderate to high endemicity defined as ≥10% prevalence of P falciparum
Intervention: MDA
Comparison: Control (no MDA or placebo)

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with no MDA

Risk with MDA

Follow‐up: 1 to 3 months

Parasitaemia prevalence

5 per 100

9 per 100
(3 to 27)

RR 1.76
(0.58 to 5.36)

786
(1 RCT)

⊕⊕⊝⊝
LOWa,b,c

Due to imprecision

At 1‐3 months post‐MDA, parasite prevalence may increase in MDA compared no MDA. However, the effects vary and it is possible that MDA makes little or no difference on parasitaemia prevalence.

Parasitaemia incidence

68 events per 100 person‐years

42 events per 100 person‐years
(27 to 63)

Rate ratio 0.61
(0.40 to 0.92)

739
(1 RCT)

⊕⊕⊕⊝
MODERATEa,b,d

Due to imprecision

At 1‐3 months post‐MDA, there is probably a reduction in parasitaemia incidence in MDA compared to no MDA.

Confirmed malaria illness incidence

28 per 1000 population

11 per 1000 population
(1 to 122)

Rate ratio 0.41
(0.04 to 4.42)

144,422
(1 RCT)

⊕⊕⊝⊝
LOWa,b,c

Due to imprecision

At 1‐3 months post‐MDA, there may be a reduction in confirmed malaria illness incidence in MDA compared to no MDA.

Follow‐up: 4 to 6 months

Parasitaemia prevalence

55 per 100

65 per 100
(49 to 86)

RR 1.18
(0.89 to 1.56)

1414
(1 RCT)

⊕⊕⊕⊝
MODERATEa,b,d

Due to imprecision

At 4‐6 months post‐MDA, there is probably little or no effect on parasitaemia prevalence in MDA compared to no MDA

Parasitaemia incidence

129 events per 100 person‐years

118 events per 100 person‐years
(71 to 194)

Rate ratio 0.91
(0.55 to 1.50)

1376
(1 RCT)

⊕⊝⊝⊝
VERY LOWa,b,d,e

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia incidence at 4‐6 months post‐MDA

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: confidence interval; cRCT: cluster‐randomized controlled trial; MDA: mass drug administration; RCT: randomized controlled trial; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect

aNot downgraded for inconsistency; the comparison presented is reported from a single study.
bNot downgraded for indirectness; outcome was evaluated in all ages or assessed in children (considered the most appropriate population to measure malaria transmission in moderate‐ to high‐endemicity areas).
cDowngraded 2 levels for imprecision due to very wide CIs.
dDowngraded 1 level for imprecision due to wide CIs.
eDowngraded 2 levels for risk of bias since malaria cases in outcome were defined as fever plus parasitaemia > 5000, which excludes all afebrile and low density infections and results in an underestimate of the outcome.

Open in table viewer
Summary of findings 2. MDA compared to no MDA for Plasmodium falciparum malaria (very low to low endemicity, short‐term follow‐up)

Patient or population: People of all ages living in an area with very low to low endemicity of P falciparum malaria (< 10% prevalence)
Setting: Very low to low endemicity defined as < 10% prevalence of P falciparum
Intervention: MDA
Comparison: Control (no MDA or placebo)

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with Control

Risk with MDA

Follow‐up: < 1 month

Parasitaemia prevalence
follow up: range < 1 month

12 per 100

1 per 100
(0 to 6)

RR 0.12
(0.03 to 0.52)

1232
(1 RCT)

⊕⊕⊝⊝
LOWa‐d

Due to risk of bias and imprecision

At < 1 month post‐MDA, there may a reduction in parasitaemia prevalence in MDA compared to no MDA.

Follow‐up: 1 to 3 months

Parasitaemia prevalence

3 per 100

1 per 100

(0 to 1)

RR: 0.25

(0.15 to 0.41)

17,454
(7 RCTs)5

⊕⊕⊝⊝
LOWe‐i

Due to risk of bias

At 1‐3 months post‐MDA, there may a reduction in parasitaemia prevalence in MDA compared to no MDA.

Parasitaemia incidence

15 events per 100 person‐years

5 events per 100 person‐years
(3 to 10)

Rate ratio 0.37
(0.21 to 0.66)

736
(1 RCT)

⊕⊕⊕⊝
MODERATEb,d,,j

Due to imprecision

At 1‐3 months post‐MDA, there is probably a reduction in parasitaemia incidence in MDA compared to no MDA.

Confirmed malaria illness incidence

6 per 1000 population

4 per 1000 population
(1 to 17)

Rate ratio: 0.58

(0.12 to 2.73)

130,651
(2 RCTs)

⊕⊝⊝⊝
VERY LOWa,j‐l

Due to risk of bias and imprecision

We do not know if MDA has an effect on confirmed malaria illness incidence at 1‐3 months post‐MDA compared to no MDA.

Follow‐up: 4 to 6 months

Parasitaemia prevalence

5 per 100

3 per 100

(2 to 6)

RR: 0.63

(0.36 to 1.12)

5670

(4 RCTs)

⊕⊝⊝⊝
VERY LOWc,d,f,l

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia prevalence at 4‐6 months post‐MDA compared to no MDA.

 

Confirmed malaria illness incidence

4 per 1000 population

4 per 1000 population
(0 to 53)

Rate ratio 0.93
(0.07 to 12.43)

23,251
(1 RCT)

⊕⊝⊝⊝
VERY LOWa,b,k,l

Due to risk of bias and imprecision

We do not know if MDA has an effect on confirmed malaria illness incidence at 4‐6 months post‐MDA compared to no MDA.

Follow‐up: 7 to 12 months

Parasitaemia prevalence

5 per 100

4 per 100

(3 to 6)

RR: 0.86

(0.55 to 1.36)

7760

(5 RCTs)

⊕⊝⊝⊝
VERY LOWc,d,l,m

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia prevalence at 7‐12 months post‐MDA compared to no MDA.

Confirmed malaria illness incidence

11 per 1000 population

5 per 1000 population (2 to 12)

Rate ratio 0.47

(0.21 to 1.03)

26,576
(3 RCTs)

⊕⊝⊝⊝
VERY LOWd,f,j,l

Due to risk of bias and imprecision

We do not know if MDA has an effect on confirmed malaria illness incidence at 7‐12 months post‐MDA compared to no MDA.

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: confidence interval; cRCT: cluster‐randomized controlled trial; MDA: mass drug administration; RCT: randomized controlled trial; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect

aDowngraded 1 level for risk of bias due to several criteria scored as high or unclear risk of bias.
bNot downgraded for inconsistency; the comparison presented is reported from a single study.
cNot downgraded for indirectness; all ages are at similar risk of malaria transmission, given the local epidemiology of malaria in this study setting and the outcomes were assessed by a highly sensitive diagnostic method (ultrasensitive PCR).
dDowngraded 1 level for imprecision due to wide CIs.
eEight included studies reported parasitaemia prevalence during the 1‐3 month post‐MDA follow‐up period; however, one study did not contribute data in the meta‐analysis due to no events at baseline before MDA or at any follow‐up time points.
fDowngraded 2 levels for risk of bias due to several criteria scored as high or unclear risk of bias, including baseline imbalance and high risk of contamination in several studies.
gNot downgraded for inconsistency based on I2 statistic; however, reasons for heterogeneity explored in post‐hoc sub‐group analysis by continent (sub‐Saharan Africa and Southeast Asia; Analysis 4.1).
hNot downgraded for indirectness; all ages are at similar risk of malaria transmission, given the local epidemiology of malaria in this study setting so there is no concern with assessing this outcome in different age groups across studies.
iNot downgraded for imprecision due to appreciable benefit of pooled effect as reported by seven studies.
jNot downgraded for indirectness since outcome was assessed in all ages and by routine detection methods.
kDowngraded 2 levels for imprecision due to very wide CIs.
lNot downgraded for inconsistency based on I2 statistic.
mDowngraded 2 levels for risk of bias due to several criteria scored as high or unclear risk of bias, including baseline imbalance, high risk of contamination, and a large unexplained increase in sampled population in the MDA group at this time point.
 

Open in table viewer
Summary of findings 3. MDA compared to no MDA for Plasmodium falciparum malaria (very low to low endemicity, long‐term follow‐up)

Patient or population: People of all ages living in an area with very low to low endemicity of P falciparum malaria ( < 10% prevalence)
Setting: Very low to low endemicity defined as < 10% prevalence of P falciparum
Intervention: MDA
Comparison: Control (no MDA or placebo)

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with Control

Risk with MDA

Follow‐up: 13 to 18 months

Parasitaemia prevalence

4 per 100

4 per 100
(1 to 14)

RR 0.82
(0.20 to 3.34)

1537
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐d

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia prevalence at 13‐18 months post‐MDA compared to no MDA.

Confirmed malaria illness incidence

17 per 1000 population

13 per 1000 population
(3 to 51)

Rate ratio 0.77
(0.20 to 3.03)

23,251
(1 RCT)

⊕⊝⊝⊝
VERY LOWb,d‐f

Due to risk of bias and imprecision

We do not know if MDA has an effect on confirmed malaria illness incidence at 13‐18 months post‐MDA compared to no MDA.

Follow‐up: 19 to 24 months

Parasitaemia prevalence
follow‐up: range 19 to 24 months

3 per 100

1 per 100
(0 to 6)

RR 0.34
(0.06 to 1.97)

1393
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐d

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia prevalence at 19‐24 months post‐MDA compared to no MDA.

Follow‐up: 25 months and above

Parasitaemia prevalence
follow‐up: range 25 to 30 months

3 per 100

3 per 100
(1 to 12)

RR 0.89
(0.22 to 3.62)

1521
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐d

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia prevalence at 25‐30 months post‐MDA compared to no MDA.

Parasitaemia prevalence
follow‐up: range 31 to 36 months

3 per 100

4 per 100
(1 to 19)

RR 1.25
(0.25 to 6.31)

1679
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐d

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia prevalence at 31‐36 months post‐MDA compared to no MDA.

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; cRCT: cluster‐randomized controlled trial; MDA: mass drug administration; RCT: randomized controlled trial; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect

aDowngraded 2 levels for risk of bias due to several criteria scored as high or unclear risk of bias, including baseline imbalance, high risk of contamination, and a large unexplained increase in sampled population in the MDA group at this time point.
bNot downgraded for inconsistency; the comparison presented is reported from a single study.
cNot downgraded for indirectness; all ages are at similar risk of malaria transmission, given the local epidemiology of malaria in this study setting and the outcomes were assessed by a highly sensitive diagnostic method (ultrasensitive PCR).
dDowngraded 2 levels for imprecision due to very wide CIs.
eDowngraded 1 level for risk of bias due to several criteria scored as high or unclear risk of bias.
fNot downgraded for indirectness since outcome was assessed in all ages and by routine detection methods.

Open in table viewer
Summary of findings 4. MDA compared to no MDA for P vivax malaria (very low to low endemicity, short‐term follow‐up)

Patient or population: People of all ages living in an area with very low to low endemicity of P vivax malaria (< 10% prevalence)
Setting: Very low to low endemicity defined as < 10% prevalence of P vivax
Intervention: MDA
Comparison: Control (no MDA or placebo)

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with Control

Risk with MDA

Follow‐up: < 1 month

Parasitaemia prevalence
follow up: range < 1 month

27 per 100

5 per 100
(2 to 11)

RR 0.18
(0.08 to 0.40)

1232
(1 RCT)

⊕⊕⊝⊝
LOWa‐d

Due to risk of bias and imprecision

At < 1 month post‐MDA, there may a reduction in parasitaemia prevalence in MDA compared to no MDA.

Follow‐up: 1 to 3 months

Parasitaemia prevalence

12 per 100

2 per 100

(1 to 3)

RR: 0.15

(0.10 to 0.24)

6896

(5 RCTs)

⊕⊕⊝⊝
LOWc,e‐g

Due to risk of bias

At 1‐3 months post‐MDA, there may a reduction in parasitaemia prevalence in MDA compared to no MDA.

Follow‐up: 4 to 6 months

Parasitaemia prevalence

11 per 100

9 per 100

(7 to 10)

RR: 0.78

(0.63 to 0.95)

 

5670

(4 RCTs)

⊕⊝⊝⊝
VERY LOWc,d,f,h

Due to risk of bias and imprecision

We do not know if MDA reduces parasitaemia prevalence at 4‐6 months post‐MDA compared to no MDA.

Follow‐up: 7 to 12 months

Parasitaemia prevalence

9 per 100

11 per 100

(9 to 13)

RR: 1.12

(0.94 to 1.34)

7760

(5 RCTs)

⊕⊝⊝⊝
VERY LOWc,d,h,i

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia prevalence at 7‐12 months post‐MDA compared to no MDA.

Confirmed malaria illness incidence

41 per 1000 population

57 per 1000 population (40 to 80)

Rate ratio: 1.38

(0.97 to 1.95)

3325

(2 RCTs)

⊕⊝⊝⊝
VERY LOWd,f,h,j

Due to risk of bias and imprecision

We do not know if MDA has an effect on confirmed malaria illness incidence at 7‐12 months post‐MDA compared to no MDA.

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; cRCT: cluster‐randomized controlled trial; MDA: mass drug administration; RCT: randomized controlled trial; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect

aDowngraded 1 level for risk of bias due to several criteria scored as high or unclear risk of bias.
bNot downgraded for inconsistency; the comparison presented is reported from a single study.
cNot downgraded for indirectness; all ages are at similar risk of malaria transmission, given the local epidemiology of malaria in this study setting and the outcomes were assessed by a highly sensitive diagnostic method (ultrasensitive PCR).
dDowngraded 1 level for imprecision due to wide CIs.
eNot downgraded for imprecision due to appreciable benefit of pooled effect as reported by five studies.
fDowngraded 2 levels for risk of bias due to several criteria scored as high or unclear risk of bias, including baseline imbalance and high risk of contamination.
gNot downgraded for inconsistency despite the large value of the I2 statistic since the direction of effect was consistent with large imprecision.
hNot downgraded for inconsistency based on I2 statistic.
iDowngraded 2 levels for risk of bias due to several criteria scored as high or unclear risk of bias, including a large unexplained increase in sampled population in the MDA group at this time point.
jNot downgraded for indirectness since outcome was assessed in all ages and by routine detection methods.

Open in table viewer
Summary of findings 5. MDA compared to no MDA for P vivax malaria (very low to low endemicity, long‐term follow‐up)

Patient or population: People of all ages living in an area with very low to low endemicity of P vivax malaria (< 10% prevalence)
Setting: Very low to low endemicity < 10% prevalence of P vivax
Intervention: MDA
Comparison: Control (no MDA or placebo)

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with Control

Risk with MDA

Follow‐up: 13 to 18 months

Parasitaemia prevalence

17 per 100

14 per 100
(8 to 25)

RR 0.81
(0.44 to 1.48)

1537
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐d

Due to risk of bias and imprecision

We do not know if MDA reduces parasitaemia prevalence at 13‐18 months post‐MDA compared to no MDA.

Follow‐up: 19 to 24 months

Parasitaemia prevalence
follow‐up: range 19 to 24 months

11 per 100

9 per 100
(4 to 20)

RR 0.84
(0.38 to 1.83)

1393
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐d

Due to risk of bias and imprecision

We do not know if MDA reduces parasitaemia prevalence at 19‐24 months post‐MDA compared to no MDA.

Follow‐up: 25 months and above

Parasitaemia prevalence
follow‐up: range 25 to 30 months

11 per 100

9 per 100
(4 to 21)

RR 0.89
(0.41 to 1.94)

1521
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐d

Due to risk of bias and imprecision

We do not know if MDA reduces parasitaemia prevalence at 25‐30 months post‐MDA compared to no MDA.

Parasitaemia prevalence
follow‐up: range 31 to 36 months

6 per 100

7 per 100
(3 to 20)

RR 1.20
(0.44 to 3.29)

1679
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐c,e

Due to risk of bias and imprecision

We do not know if MDA reduces parasitaemia prevalence at 31‐36 months post‐MDA compared to no MDA.

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: confidence interval; cRCT: cluster‐randomized controlled trial; MDA: mass drug administration; RCT: randomized controlled trial; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect

aDowngraded 2 levels for risk of bias due to several criteria scored as high or unclear risk of bias, including a large unexplained increase in sampled population in the MDA group at this time point.
bNot downgraded for inconsistency; the comparison presented is reported from a single study.
cNot downgraded for indirectness; all ages are at similar risk of malaria transmission, given the local epidemiology of malaria in this study setting and the outcomes were assessed by a highly sensitive diagnostic method (ultrasensitive PCR).
dDowngraded 1 level for imprecision due to wide CIs.
eDowngraded 2 levels for imprecision due to very wide CIs.

Background

Description of the condition

Malaria remains a leading cause of morbidity and mortality, and was responsible for an estimated 229 million clinical cases and 409,000 deaths in 2019, mostly in children under five years of age living in sub‐Saharan Africa. Malaria is a parasitic disease that is transmitted to humans by the bite of a female Anopheles species mosquito. Among the four Plasmodium species of human malaria (P falciparum, P vivax, P ovale, and P malariae), P falciparum and P vivax cause the largest number of malaria cases, while P falciparum is responsible for a majority of severe and fatal infections. Without effective treatment, malaria may manifest with severe complications and lead to death (WHO 2020a).

Since 2000, the global burden of malaria has been reduced considerably due to the expansion of effective malaria control strategies such as vector control and case management (Bhatt 2015; WHO 2020a). Currently recommended vector control tools include the use of insecticide‐treated bed nets (ITNs), indoor residual spraying (IRS) with insecticides, and larval source management. Case management of malaria, which relies on the prompt diagnosis (parasitological confirmation) and administration of effective antimalarial treatment, has been enhanced by the use of rapid diagnostic tests (RDTs) and artemisinin‐based combination therapy (ACTs). Intermittent preventive treatment (IPT) with antimalarial drugs in certain high risk populations, such as pregnant women, children, and infants, is another key malaria control strategy. However, challenges in achieving and maintaining high levels of coverage with these proven interventions, as well as the emergence of insecticide resistance in mosquitos and antimalarial drug resistance in parasites, threaten the success of these interventions. Therefore, additional strategies must be considered to complement existing tools to accelerate progress towards malaria elimination. Recent World Malaria Reports from the World Health Organization (WHO) have noted a stagnation in reductions of malaria cases (WHO 2020a). This finding underscores the need for continued evaluation and improved targeting of malaria control and elimination strategies to identify those activities that will have a positive impact on malaria transmission.

In the past decade, there has been renewed interest in mass drug administration (MDA), which was a component of many malaria elimination programmes during the eradication era in the mid‐twentieth century. MDA has been used widely for controlling and eliminating the five most highly prevalent neglected tropical diseases (lymphatic filariasis, onchocerciasis, schistosomiasis, soil‐transmitted helminths, and trachoma) (Hotez 2009). However, the role of MDA for malaria has been less clear due to the fear of rapid emergence of antimalarial drug resistance and the varying success of the intervention in sustaining reductions in malaria burden and interrupting malaria transmission. A Cochrane Review found MDA to be effective in reducing malaria parasitaemia prevalence within one month following the intervention (Poirot 2013), but results on sustaining the impact four to six months after the intervention were mixed or the evidence was of low quality. Since the publication of the previous review, additional studies have been conducted to assess both the immediate‐ and longer‐term effects of MDA in low‐transmission and moderate‐ to high‐transmission settings. The additional evidence provides the opportunity to address gaps in knowledge surrounding MDA as a strategy to accelerate progress towards malaria elimination across transmission settings.

Description of the intervention

MDA for malaria consists of administering a full therapeutic course of antimalarial medicine (irrespective of the presence of symptoms or infection) to every member of a defined population or every person living in a defined geographical area (except to those for whom the medicine is contraindicated) at approximately the same time and often at repeated intervals. MDA for malaria is intended to reduce or interrupt transmission, curtail morbidity and mortality, or prevent reoccurrence and resulting malaria transmission (WHO 2015a).

Many early studies of MDA for malaria (dating back to the 1930s) observed a marked decrease in parasite prevalence following MDA, but often MDA did not result in interruption of transmission. Subsequent studies that evaluated MDA in conjunction with other malaria control tools, including the Garki Project conducted in 1969 in Northern Nigeria (Molineaux 1980), often demonstrated substantial decreases in, but not interruption of, malaria transmission (Greenwood 2004; von Seidlein 2003). In Nicaragua, MDA was administered to approximately 8 million people in 1981 and resulted in marked immediate decline in malaria incidence, but rates eventually reverted back to those observed prior to MDA (Garfield 1983). Notably, MDA in conjunction with ITNs on Aneityum Island, Vanuatu in 1991 was an exception, and interruption of transmission was achieved following eight rounds of MDA with chloroquine, sulfadoxine‐pyrimethamine, and primaquine (Kaneko 2000 VUT). On another island setting (Anjouan Island, Comoros), two rounds of MDA, with artemisinin‐piperaquine or artemisinin‐piperaquine with single low dose primaquine, were associated with a rapid 99% decrease in malaria cases (from 7362 before to 47 cases after MDA rounds) (Deng 2018 COM). More recently, the role of MDA to curb malaria morbidity and mortality in complex emergency settings was highlighted during the Ebola outbreak in West Africa (Aregawi 2016 SLE; Kuehne 2016).

How the intervention might work

MDA targets the human parasite reservoir through the administration of a curative antimalarial dose to the entire population, irrespective of symptoms or parasitaemia. MDA provides a prophylactic effect, whereby new infections are temporarily prevented, and a treatment effect, in which parasitaemia is cleared. Unlike case management, in which symptomatic cases are treated, MDA also targets asymptomatic malaria infections, that are believed to  contribute to ongoing malaria transmission (Lindblade 2013). Theoretically, in combination with high levels of vector control, clearing the human parasite reservoir could lead to an interruption of transmission in areas of low endemicity or a temporary decline in malaria burden in higher transmission settings. However, MDA effectiveness may be limited by factors such as antimalarial properties and operational considerations including population coverage, risk of re‐introduction, and timing.

The Plasmodium parasite life cycle includes stages in humans (liver and blood) and mosquitoes. In particular, parasitic infections during the exo‐erythrocytic and erythrocytic cycles in humans have important implications for clinical illness, diagnosis, treatment, and prevention. Once Plasmodium sporozoites are injected by mosquitoes in a human host, parasites invade and multiply in liver cells, leading to hepatocytic rupture and the release of parasites into the blood stream. P vivax and P ovale can persist in a dormant liver stage as hypnozoites, leading to relapsing erythrocytic stage infections months or years after initial infection. During the blood stage, asexual reproduction of merozoites in red blood cells results in a cycle of parasite maturation and erythrocytic destruction, triggering clinical symptoms such as fever, headache, and chills. Some parasites undergo sexual differentiation (gametocytes) in humans, which subsequently leads to malaria transmission once the gametocytes are ingested during a mosquito blood meal.

The life cycle of the Plasmodium parasite offers numerous vulnerable stages that could be exploited in a MDA strategy to reduce disease burden and transmission. Most antimalarial drugs are effective schizonticides against asexual blood stage parasites and can eliminate blood stream infections and prevent illness, and may reduce the overall density of asexual parasites, leading to a decline in gametocyte density. Artemisinins and 8‐aminoquinolines (e.g. primaquine and tafenoquine) have gametocytocidal properties and act directly on circulating gametocytes. Primaquine has long been the only available drug with unique activity against mature gametocytes of P falciparum and the hypnozoite stage of P vivax and P ovale species, reducing the possibility of transmission and relapse (WHO 2015b). Tafenoquine, another 8‐aminoquinoline, recently received regulatory approval for the radical cure of P vivax infections in persons older than 16 years (GlaxoSmithKline 2018). The addition of ivermectin, or other endectocides, could potentially reduce malaria transmission by shortening the lifespan of mosquitos, thus preventing parasite development in the mosquito. Although ivermectin is not an antimalarial drug, it is often used to control other parasitic infections in malaria‐endemic settings and the effect of MDA with ivermectin on malaria transmission has been summarized in a separate Cochrane Review (Chaccour 2010; Tesh 1990; de Souza 2021).

Operationally, high levels of MDA coverage may not be achieved since a proportion of the population may be excluded from MDA due to contraindications to antimalarials or health conditions, absence during treatment rounds, or refusal. Although conducting multiple rounds of MDA can improve intervention coverage, the same people are often repeatedly excluded from MDA rounds, which reduces the effective coverage, or the proportion of the population receiving at least one round of MDA. High rates of population movement or migration also pose a challenge by increasing the risk of re‐introduction. Additionally, mathematical modelling has provided insights into the optimal timing in which rounds should take place in relation to the malaria transmission season. Implementation of MDA prior to the rainy season has been shown to have the greatest effect, due to widespread parasite clearance just prior to the high transmission season (Brady 2017; Okell 2011).

Why it is important to do this review

Despite unprecedented success in malaria control, progress has stalled in recent years (Alonso 2017). In addition to increased financial and political commitment, technical and operational strategies are needed to accelerate progress towards malaria elimination. Given the variable success of MDA, more recent studies have been conducted with consideration of transmission setting, seasonality of malaria, duration of intervention, MDA coverage, antimalarials with longer prophylactic action, and measurement of longer‐term effects. Recent mathematical models have suggested that malaria reduction following MDA is more likely to be sustained in lower transmission settings compared to higher, in conjunction with vector control interventions, and when the intervention is continued over two years compared to one (Brady 2017; Okell 2011). Additionally, there has been recent debate within the malaria community on the role of MDA in malaria elimination (Eisele 2019; Mendis 2019).

Currently, the WHO recommends the use of MDA for P falciparum in areas approaching elimination, for the elimination of P falciparum in areas of multidrug resistance in the Greater Mekong subregion, epidemics, and complex emergency settings (WHO 2015a). At the time the recommendations were released (2015), insufficient evidence was available to provide guidance on the use of MDA for P vivax, or for the use of MDA for P falciparum in areas of moderate to high transmission. However, use of MDA could be considered in areas approaching interruption of transmission of P falciparum with good coverage of vector control, implementation of surveillance, access to treatment, and minimal risk of re‐introduction of infection. Since the publication of the previous Cochrane Review (Poirot 2013), and the release of the WHO guidelines, additional studies have been conducted in both very low‐ to low‐transmission settings and moderate‐ to high‐transmission settings, including both P falciparum and P vivax. Due to renewed interest in MDA as a strategy to accelerate progress towards malaria elimination, an update to the previous review to address whether MDA provides sustained impact in very low‐ to low‐transmission settings and moderate‐ to high‐transmission settings will provide evidence to evaluate the appropriateness of existing recommendations, guide future policies, highlight current knowledge gaps, and direct future research.

Objectives

Primary objectives

To assess the sustained effect of mass drug administration (MDA) with antimalarial drugs on:

  • the reduction in malaria transmission in moderate‐ to high‐transmission settings;

  • the interruption of transmission in very low‐ to low‐transmission settings.

Secondary objective

To summarize the risk of drug‐associated adverse effects following MDA.

Methods

Criteria for considering studies for this review

Types of studies

We considered the following study designs.

  • Randomized controlled trials (RCTs) with: (a) the unit of randomization being a cluster, and (b) at least two clusters per arm.

  • Randomized cross‐over trials with: (a) the unit of randomization being a cluster, (b) at least two clusters per arm, and (c) a suitable washout period during which the intervention is no longer applied.

  • Quasi‐experimental designs, including stepped wedge where sites are randomly allocated.

  • Non‐randomized controlled before‐and‐after studies (CBAs) with: (a) a contemporaneous control group, and (b) at least two sites per arm.

  • Interrupted time series (ITS) studies with: (a) a clearly defined point in time when the intervention occurred, and (b) at least three data points collected over one year both before and after MDA.

We excluded studies if we observed any of the following.

  • The follow‐up periods for the intervention and control periods were not identical.

  • The intervention was applied at the individual level.

  • There was no control arm for RCTs or CBAs, or insufficient pre‐ or post‐intervention data points for ITS.

Types of participants

Children and adults living in areas of any malaria endemicities were included and analysed by transmission setting according to the objectives. Due to the nature of the intervention, we included only studies that were carried out on entire populations at the same time. We excluded studies in which chemoprevention was delivered in the form of individually‐timed, intermittent, preventive treatment in sub‐populations (i.e. pregnant women, children or infants) or seasonal malaria chemoprevention. We did not exclude studies in special groups (i.e. refugees and soldiers), but none of these met eligibility criteria.

Types of interventions

Intervention

For the purpose of this review, we defined MDA as the direct administration of a full therapeutic course of antimalarial medicine (irrespective of the presence of symptoms or infection) to every member of a defined population or every person living in a defined geographical area (except for those for whom the medicine was contraindicated) at approximately the same time and often at repeated intervals. We included only studies that provided doses of antimalarials intended for curative purposes.

Control

We included studies in which all other malaria and non‐malaria co‐interventions were balanced in all arms. These included the use of ITNs, IRS, source reduction activities, environmental management, mass drug campaigns for other neglected tropical diseases, and mass nutritional supplementation activities such as vitamin A distribution.

Types of outcome measures

Studies must have reported at least one primary outcome for inclusion.

Primary outcomes

  • Parasitaemia prevalence, as measured by microscopy, malaria RDT, or molecular method, such as polymerase chain reaction (PCR).

  • Parasitaemia incidence (e.g. incidence of infection through active surveillance as measured in a cohort).

  • Confirmed malaria illness incidence, defined as febrile illness with diagnostically confirmed parasitaemia (WHO 2015c) (e.g. incidence of confirmed clinical infection as measured in passive or routine surveillance data collected at health facilities).

Secondary outcomes

  • All‐cause and malaria‐specific mortality.

  • Gametocytaemia prevalence, as measured by microscopy or molecular method.

  • Known adverse effects related to MDA drugs using WHO definitions (Edwards 2000).

Search methods for identification of studies

Electronic searches

Search strategy for identification of studies

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

Databases

We searched the following databases on 11 February 2021, using the search terms and strategy described in Appendix 1.

  • Cochrane Infectious Diseases Group Specialized Register (searched 11 February 2021).

  • Cochrane Central Register of Controlled Trials (CENTRAL), published in the Cochrane Library (Issue 2 of 12, February 2021).

  • MEDLINE (PubMed; 1966 to 11 February 2021).

  • Embase (Ovid; 1947 to 11 February 2021).

  • LILACS (Latin American and Caribbean Health Science Information database, BIREME; 1982 to 11 February 2021).

We also searched the WHO International Clinical Trials Registry Platform (ICTRP; www.who.int/ictrp/en/) and ClinicalTrials.gov (clinicaltrials.gov/ct2/home) for trials in progress, using 'malaria' and 'mass drug administration' as search terms.

Searching other resources

Reference lists 

We checked the reference lists of all studies and articles identified by the above methods and of previously published reviews, as well as references listed in review articles (Greenwood 2004; Newby 2015; Poirot 2013; Shanks 2012; von Seidlein 2003).

Conference proceedings

We searched the following recent proceedings for relevant abstracts.

  • Sixth Multilateral Initiative on Malaria Pan‐African Malaria Conference (Durban, South Africa; October 2013).

  • American Society of Tropical Medicine and Hygiene (ASTMH) 62nd Annual Meeting (Washington, DC, USA; November 2013).

  • ASTMH 63rd Annual Meeting (New Orleans, LA, USA; November 2014).

  • ASTMH 64th Annual Meeting (Philadelphia, PA, USA; October 2015).

  • ASTMH 65th Annual Meeting (Atlanta, GA, USA; November 2016).

  • ASTMH 66th Annual Meeting (Baltimore, MD, USA; November 2017).

  • ASTMH 67th Annual Meeting (New Orleans, Louisiana, USA; October 2018).

Researchers and organizations

In addition to the electronic searches described above, we reached out to the Malaria Elimination Initiative and other relevant groups to identify both published and unpublished studies that might be available from other sources. We also contacted the Malaria Eradication Research Agenda Consortium and the Bill and Melinda Gates Foundation.

Data collection and analysis

Selection of studies

Two review authors (MPS and MD) independently assessed the titles and abstracts of trials identified by the literature searches. We obtained the full‐text versions of any potentially relevant articles identified by at least one of the review authors. The same two review authors assessed the full‐text articles of potentially relevant studies for inclusion, using an eligibility form based on predetermined inclusion criteria. We resolved any disagreements by discussion and consensus, with arbitration by a third review author (KAL) when necessary. We ensured that multiple publications of the same trial were included only once. We listed studies excluded after full‐text assessment, together with their reasons for exclusion, in the Characteristics of excluded studies table. We illustrated the study selection process using a Preferred Reporting Items of Systematic reviews and Meta‐Analyses (PRISMA) diagram.

Data extraction and management

Two review authors (MPS and MD) independently extracted information from the trials using pre‐piloted electronic data extraction forms. In case of differences in extracted data, the two review authors discussed these differences to reach consensus. If unresolved, further discussion involved a third review author (KAL). In case of missing data, we contacted the original study author(s) for clarification.

We extracted the following data.

  • Trial design: type of trial; method of participant selection; adjustment for clustering (for cRCTs); sample size; method of blinding of participants and personnel.

  • Participants: trial settings (country, transmission season, endemicity, antimalarial drug resistance context, parasite and vector species of interest) and population characteristics; recruitment rates; withdrawal and loss to follow‐up.

  • Intervention: MDA regimen, coverage, and timing (number of rounds/campaigns, years of implementation, and timing relative to transmission season).

  • Outcomes: definition of outcome; diagnostic or surveillance method; number of events; number of participants or person time; time point at which outcome was assessed in relation to MDA; statistical power; unit of analysis; incomplete outcomes or missing data.

For dichotomous data (parasitaemia prevalence and gametocytaemia prevalence), we extracted the number of participants experiencing each outcome and the number of participants in each treatment group. For count data (parasitaemia incidence, confirmed malaria illness incidence, and mortality), we extracted the number of events in the treatment and control groups, and the total person time at risk in each group or the rate ratio, and a measure of variance (for example, standard error). We did not extract any continuous data.

For cRCTs, we recorded the number of clusters randomized, number of clusters analysed, and the intracluster correlation coefficient (ICC) value.

For non‐randomized studies, there were no studies that reported adjusted measures of intervention effects, so we were unable to obtain an effect estimate that controlled for confounding.

We included pre‐intervention data up to one year prior to the intervention. We included all post‐MDA data, and outcomes were reported according to designated time points of: less than 1 month, 1 to 3 months, 4 to 6 months, 7 to 12 months, 13 to 18 months, 19 to 24 months, 25 to 30 months, and 31 to 36 months after MDA. For studies with multiple rounds of MDA, we defined 'during MDA' as the intervention time period ‐ i.e. the time between the start of the first round and end of the last round of MDA ‐ and the 'post‐MDA' follow‐up period as the time following the last round of MDA.

Assessment of risk of bias in included studies

Two review authors (MPS and MD) independently assessed the risk of bias for each included cluster‐RCT using the Cochrane risk of bias tool and the five additional criteria listed in Section 16.3.2 of the Cochrane Handbook for Systematic Reviews of Interventions that relate specifically to cluster‐RCTs (Higgins 2011a; Higgins 2011b). We assessed non‐randomized controlled studies for risk of bias using the Cochrane Effective Practice and Organisation of Care (EPOC) risk of bias tool. We resolved any discrepancies through discussion or by consulting a third review author (KAL). We classified judgements of risk of bias as either at low, high, or unclear risk of bias, and we used summary graphs (risk of bias summary and risk of bias graph) to display results.

Measures of treatment effect

We compared intervention and control groups by calculating risk ratios (RR) or rate ratios for incidence data. We presented all results with their corresponding 95% confidence intervals (CIs). At all time periods, a RR less than 1.0 indicates that parasitaemia prevalence was lower in the MDA compared to control arm, while a RR greater than 1.0 indicates that parasitaemia prevalence was higher in the MDA compared to control arm. Similarly, a rate ratio less than 1.0 can be interpreted as a lower rate of malaria infection (incidence) in the MDA compared to control arm, and a rate ratio greater than 1.0 reflects a higher malaria incidence measured in the MDA compared to control arm.

For non‐randomized studies, since adjusted effect measures were not provided, we did not present a measure of treatment effect. Instead, we described the effect of MDA compared to no MDA using a difference‐in‐differences (DiD) analysis in which we calculated the difference in dichotomous outcomes between pre‐ and during‐MDA time periods, and pre‐ and post‐MDA time periods within intervention and control arms and then took the difference of those values between intervention and control groups.

Unit of analysis issues

Since a variety of analytical methods were used to adjust for clustering in cRCTs, we extracted raw data and estimated effective sample sizes adjusted for clustering using either: (1) the study‐provided ICC (as indicated in the Characteristics of included studies tables), or, (2) for studies that did not report an ICC, an estimated ICC calculated as the average of other study‐provided ICCs in the same malaria transmission setting (0.02766 for areas of very low to low transmission, and 0.1225 for areas of moderate to high transmission). We only presented results from cRCTs that were adjusted for clustering. No cRCTs included multiple intervention arms so adjustment for multiple comparison in meta‐analysis was not required (Richardson 2016).

Dealing with missing data

We attempted to contact study investigators to obtain missing data. We applied no imputation measures for working with missing data.

Assessment of heterogeneity

We inspected forest plots for overlapping CIs and assessed statistical heterogeneity in each meta‐analysis using the I² statistic and Chi² test. We regarded heterogeneity as moderate if the I² statistic values were between 30% and 60%; substantial if they were between 50% and 90%; and considerable if they were between 75% and 100%. We regarded a Chi² test statistic with a P value of less than or equal to 0.10 as indicative of statistically significant heterogeneity. We explored clinical and methodological heterogeneity through consideration of the trial populations, methods, and interventions, and by visualization of trial results.

If there was considerable heterogeneity (i.e. an I² statistic value of 75% to 100%) and inconsistency in the direction of the effect, then we did not perform a meta‐analysis.

Assessment of reporting biases

Since there were fewer than 10 trials included in each meta‐analysis, we did not investigate reporting biases (such as publication bias) using funnel plots.

Data synthesis

We analysed data using Review Manager 5 (RevMan 5) (Review Manager 5). We used a fixed‐effect meta‐analysis to combine data if heterogeneity was absent. We pooled estimates where considerable heterogeneity was present if the direction of effect was consistent.

Malaria endemicity was classified as very low to low (prevalence of P falciparum or P vivax < 10%), or moderate to high (≥ 10%) (WHO 2017). Study‐specific endemicity was defined by baseline malaria prevalence by microscopy or RDT or annual parasite incidence in the control group, and preferentially using data from (1) children, or (2) all ages. If only molecular data were available, we used a tool developed by Okell 2012 to estimate microscopy prevalence from PCR data in the control group at baseline.

For the main objective in moderate‐ to high‐transmission settings, we defined a reduction in malaria transmission as a 50% reduction in median malaria parasite prevalence or incidence, or both, at 12 months post‐intervention. For the main objective in very low‐ to low‐transmission settings, we considered interruption of transmission as a reduction in number of indigenous malaria infections to zero at six months post‐intervention. To evaluate both objectives, we stratified analyses by study design (i.e. cRCTs and non‐randomized controlled studies) and post‐intervention time periods (i.e. less than 1 month after MDA, 1 to 3 months, 4 to 6 months, 7 to 12 months, 13 to 18 months, 19 to 24 months, 25 to 30 months, and 31 to 36 months after MDA). For studies with data from multiple time points within the same post‐intervention time period (Landier 2017 MMRa; Pongvongsa 2018 LAO; Tripura 2018 KHM; von Seidlein 2019 VNM), we used the latest time point for analysis for that category. We did not stratify studies by number of MDA rounds due to few studies after previous stratification. We only conducted a meta‐analysis if we identified a sufficient number of studies (> 1) with both an outcome indicator estimate and a measure of precision.

Subgroup analysis and investigation of heterogeneity

We stratified outcomes by malaria transmission setting (very low‐ to low‐transmission and moderate‐ to high‐transmission settings) and Plasmodium species (P falciparum or P vivax). Given the few number of studies after stratification, we did not carry out subgroup analyses to explore causes of heterogeneity. However, for a single outcome where studies conducted in sub‐Saharan Africa and Southeast Asia were combined (P falciparum prevalence at one to three months after MDA in very low‐ to low‐endemicity settings), we carried out a supplementary post‐hoc subgroup analysis by continent (Africa and Asia), to consider whether the geographical differences in malaria epidemiology may explain heterogeneity in effect of MDA.

Sensitivity analysis

Due to an insufficient number of studies, we did not perform sensitivity analyses on the primary outcomes to assess the effect of excluding trials at high risk of bias (for baseline imbalance and incomplete outcome data) on the overall results. We did not undertake sensitivity analyses to investigate the impact of varying the ICC value on meta‐analysis results since the ICC values were obtained directly from studies or applied from comparable studies.

Summary of findings and assessment of the certainty of the evidence

We assessed the certainty of evidence using the GRADE approach (Guyatt 2011). We rated each important outcome as described by Balshem 2011.

  • High: we are very confident that the true effect lies close to that of the estimate of the effect.

  • Moderate: we are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect.

  • Low: our confidence in the effect estimate is limited. The true effect may be substantially different from the estimate of the effect.

  • Very low: we have very little confidence in the effect estimate. The true effect is likely to be substantially different from the estimate of effect.

RCTs started as high quality evidence but could be downgraded if there were valid reasons within the following five categories: risk of bias, imprecision, inconsistency, indirectness, and publication bias. We summarized our findings in a summary of findings table.

Results

Description of studies

Detailed descriptions of included studies, excluded studies, studies awaiting classification, and ongoing studies are provided in the Characteristics of included studies, Characteristics of excluded studies, Characteristics of studies awaiting classification, and Characteristics of ongoing studies tables, respectively.

Results of the search

The last published version of this review included 32 studies: two cRCTs, eight non‐randomized trials, and 22 uncontrolled before‐and‐after studies (Poirot 2013).

Following the revised inclusion criteria, which restricted the review to more rigorous study designs with a control group and balanced co‐interventions across study arms (described in detail in Differences between protocol and review), the updated literature search (to 11 February 2021) identified 462 records. After de‐duplication and removal of studies excluded by the previous review's literature search in 2012, we screened 251 titles. Of those, we assessed 39 full‐text articles for study eligibility (Figure 1).

Figure 1
Study flow diagram

Study flow diagram

Included studies

A total of 13 studies met the criteria for inclusion, comprising five studies included in the previous review (Escudie 1962 BFA; Molineaux 1980 NGA; Roberts 1964 KEN; Shekalaghe 2011 TZA; von Seidlein 2003 GMB), and eight new studies (Eisele 2020 ZMBa; Eisele 2020 ZMBb; Landier 2017 MMRa; McLean 2021 MMR; Morris 2018 TZA; Pongvongsa 2018 LAO; Tripura 2018 KHM; von Seidlein 2019 VNM). Since clusters in the Eisele trial were randomized by areas of low and high malaria transmission and the outcomes were stratified by endemicity (specified a priori by design), we considered this trial as two studies (Eisele 2020 ZMBa; Eisele 2020 ZMBb). Four studies were part of a multi‐county trial conducted in Southeast Asia (Landier 2017 MMRa; Pongvongsa 2018 LAO; Tripura 2018 KHM; von Seidlein 2019 VNM). We analysed these as separate studies due to differences in study design (differences in antimalarial drug used for MDA, timing of MDA in relation to the transmission season, and follow‐up time periods) and heterogeneity of effects.

Cluster‐randomized trials

Ten cRCTs were included in the qualitative syntheses: two from moderate‐ to high‐endemicity areas (high: Eisele 2020 ZMBb; von Seidlein 2003 GMB), and eight from very low‐ to low‐endemicity areas (very low: Morris 2018 TZA; Shekalaghe 2011 TZA; Tripura 2018 KHM; von Seidlein 2019 VNM; low: Eisele 2020 ZMBa; Landier 2017 MMRa; McLean 2021 MMR; Pongvongsa 2018 LAO).

Five studies were conducted in sub‐Saharan Africa and five studies were conducted in Southeast Asia. Trial locations in sub‐Saharan Africa were Southern Province, Zambia (Eisele 2020 ZMBa; Eisele 2020 ZMBb), Unguja Island, Tanzania (Morris 2018 TZA), Lower Moshi, Tanzania (Shekalaghe 2011 TZA), and Farafenni, The Gambia (von Seidlein 2003 GMB). Study locations in Southeast Asia included Kayin (Karen) state, Myanmar (Landier 2017 MMRa), Southeast Myanmar (McLean 2021 MMR), Savannakhet province, Laos (Pongvongsa 2018 LAO), Battambang province, Cambodia (Tripura 2018 KHM), and Binh Phuoc and Ninh Thuan provinces, Vietnam (von Seidlein 2019 VNM).

One cRCT conducted in an area of very low malaria transmission of Tanzania reported zero events at both baseline and follow‐up time points for several outcomes and was excluded from quantitative syntheses (Shekalaghe 2011 TZA). The nine remaining cRCTs provided data for comparison of MDA versus no MDA in this review: three studies compared MDA to placebo or no MDA (McLean 2021 MMR; Morris 2018 TZA; von Seidlein 2003 GMB), four studies compared MDA to delayed MDA (Landier 2017 MMRa; Pongvongsa 2018 LAO; Tripura 2018 KHM; von Seidlein 2019 VNM), while two studies (Eisele 2020 ZMBa; Eisele 2020 ZMBb) compared MDA and focal MDA (household‐level MDA where a member tested positive by RDT) separately to no MDA and we included only the MDA arm for comparison.

Interventions

Characteristics of the intervention have been summarized in Table 1.

Open in table viewer
Table 1. Description of studies

Study ID (Design)

Year(s) of study

Malaria endemicitya

Plasmodium species

Antimalarial drug resistance

MDA group

Control group

Co‐intervention(s)b

Outcomes reported (months of follow‐up post‐MDAc)

Drug

Rounds, interval, and duration implemented

Population targeted (coverage)

Eisele 2020 ZMBa (cRCT)

2014‐2017

Low

P falciparum

Widespread resistance to CQ and SP, but no evidence of resistance to artemisinin

DHAp

4 rounds administered at start of rainy season, during rainy season, during dry season, and at start of rainy season over 15 months

37,694

(79% in round 1; 63% in round 2; 76% in round 3; 66% in round 4)

No drug and no placebo

IRS, ITNs, and enhanced standard of care

  • Parasitaemia prevalence (2)

  • Parasitaemia incidence (2)

  • Confirmed malaria case incidence (2)

  • Adverse effects

Eisele 2020 ZMBb (cRCT)

2014‐2017

High

P falciparum

Widespread resistance to CQ and SP, but no evidence of resistance to artemisinin

DHAp

4 rounds administered at start of rainy season, during rainy season, during dry season, and at start of rainy season over 15 months

45,442

(79% in round 1; 63% in round 2; 76% in round 3; 66% in round 4)

No drug and no placebo

IRS, ITNs, and enhanced standard of care

  • Parasitaemia prevalence (2)

  • Parasitaemia incidence (2)

  • Confirmed malaria case incidence (2)

  • Adverse effects

Escudie 1962 BFA (CBA)

1960‐1961

High

P falciparum,

P ovale,

P malariae

ND

AQ‐PQ or CQ‐PQ

(Low frequency MDA)

7 rounds administered 28 days apart over 7 months

1890

(75% to 91% per round)

No drug and no placebo

None (IRS arms excluded)

  • Parasitaemia prevalence (3)

  • Gametocytaemia prevalence (3)

(High frequency MDA)

15 rounds administered 14 days apart over 7 months

2560

(84% to 97% per round)

  • Parasitaemia prevalence (3)

  • Gametocytaemia prevalence (3)

Landier 2017 MMRa (cRCT)

2013‐2015

Low

P falciparum,

P vivax

Artemisinin resistance firmly established

DHAp with PQ

3 rounds administered 1 month apart over 3 months

1434 (66% in round 1, 56% in round 2, and 65% in round 3)

Delayed MDA

ITNs, uninterrupted access to case management

  • Parasitaemia prevalence (7)

  • Confirmed malaria illness incidence (7)

  • Adverse effects

McLean 2021 MMR (cRCT)

2014‐2017

Very low

P falciparum,

P vivax

Artemisinin resistance: Kelch 13 mutation in 57% of samples at baseline

DHAp with PQ

3 rounds administered 1 month apart over 3 months

4622

(86% in round 1, 86% in round 2, 88% in round 3)

No drug and no placebo

ITNs, routine malaria control by village health workers

  • Parasitaemia prevalence (31)

  • Adverse effects

Molineaux 1980 NGA (CBA)

1970‐1975

High

P falciparum,

P malariae,

P ovale

ND

SP

(Low frequency MDA)

9 rounds administered 10 weeks apart over 18 months

14,129

(73% to 92% per round)

No drug and no placebo

IRS

  • Parasitaemia prevalence (0)

  • Gametocytaemia prevalence (0)

(High frequency MDA)

23 rounds administered 2 weeks apart during the wet seasons and 10 weeks apart during the dry seasons over 18 months

1810

(72% to 91% per round)

Morris 2018 TZA (cRCT)

2016‐2017

Very low

P falciparum,

P malariae,

P ovale, and

P vivax

No evidence of resistance to first line treatment AS‐AQ

DHAp with PQ

2 rounds administered 4 weeks apart over 6 weeks

10,944

(91% in round 1, 88% in round 2)

No drug and no placebo

IRS and ITNs

  • Parasitaemia prevalence (0)

  • Confirmed malaria illness incidence (14)

  • Adverse effects

Pongvongsa 2018 LAO (cRCT)

2016‐2017

Low

P falciparum,

P vivax

ND

DHAp with PQ

3 rounds administered 1 month apart over 3 months

1006 (81% in round 1, 80% in round 2, and 82% in round 3)

Delayed MDA

ITNs, uninterrupted access to case management

  • Parasitaemia prevalence (10)

  • Adverse effects

Roberts 1964 KEN (CBA)

1953‐1954

Moderate

P falciparum

ND

Pyrimethamine

2 rounds administered 1 year apart over 13 months

101,000

(95% in round 1, 93% in round 2)

No drug and no placebo

None

  • Parasitaemia prevalence (7)

Shekalaghe 2011 TZA (cRCT)

2008

Very low

P falciparum

ND

SP+AS with PQ

1 round over 16 days

1110

(95%)

Placebo

ITNs, single treatment campaign for trachoma with azithromycin

  • Parasitaemia prevalence (4)

  • Confirmed malaria illness incidence (4)

  • Gametocytaemia prevalence (4)

  • Adverse effects

Tripura 2018 KHM (cRCT)

2014‐2016

Very low

P falciparum,

P vivax

Reduced susceptibility to artemisinins and ACT partner drug resistance

DHAp

3 rounds administered 1 month apart over 3 months

858 (74% in round 1, 60% in round 2, and 71% in round 3)

Delayed MDA

ITNs, uninterrupted access to case management

  • Parasitaemia prevalence (10)

  • Confirmed malaria case incidence (9)

  • Adverse effects

von Seidlein 2003 GMB (cRCT)

1999

High

P falciparum

ND

SP+AS

1 round over 1 month

12,331

(89%)

Placebo

None

  • Parasitaemia prevalence (5)

  • Parasitaemia incidence (5)

  • Gametocytaemia prevalence (5)

  • Malaria‐specific mortality

  • Adverse effects

von Seidlein 2019 VNM (cRCT)

2013‐2015

Very low

P falciparum,

P vivax

No evidence of resistance to DHAp at the start of study, but treatment failure to DHAp has increased following study

DHAp with PQ

3 rounds administered 1 month apart over 3 months

1439 (83% in round 1, 98% in round 2, and 99% in round 3)

Delayed MDA

ITNs, uninterrupted access to case management

  • Parasitaemia prevalence (10)

  • Adverse effects

ACT = artemisinin‐based combination therapy, AQ = amodiaquine, AS = artesunate, CBA = controlled before‐and‐after study, CQ = chloroquine, cRCT = cluster‐randomized controlled trial, DHAp = dihydroartemisinin piperaquine, ITNs = insecticide‐treated bed nets, IRS = indoor residual spraying, MDA = mass drug administration, PQ = primaquine, SP = sulfadoxine‐ (or sulfalene‐) pyrimethamine, NA = not applicable, ND = not described.

aMalaria endemicity classified as very low (> 0% to < 1%), low (1% to < 10%), moderate (10% to < 35%) or high (≥ 35%) (WHO 2017).
bCo‐interventions were balanced across intervention and control groups, as per inclusion criteria.
cPost‐MDA refers to the length of time, in months, after the last round of MDA that the outcome was evaluated.

Moderate‐ to high‐transmission areas

One study administered four rounds of MDA with dihydroartemisinin piperaquine (DHAp) to all persons older than three months, except for pregnant women in the first trimester (Eisele 2020 ZMBb). The four MDA rounds occurred at the start of the rainy season, during the rainy season, during the dry season, and again at the start of the rainy season. MDA coverage was 88.1% in round one and 72.0% in round two. Both study arms received LLINs and IRS with Actellic (an insecticide) at baseline before MDA. The second trial administered a single round of MDA with sulfadoxine‐pyrimethamine plus artesunate at the start of the rainy season to all persons six months of age or older, except for pregnant women (von Seidlein 2003 GMB). MDA coverage among the target population was 89% and no co‐interventions were provided.

Very low‐ to low‐transmission areas

Seven studies administered multiple rounds of MDA with dihydroartemisinin piperaquine (Eisele 2020 ZMBa; Landier 2017 MMRa; McLean 2021 MMR; Morris 2018 TZA; Pongvongsa 2018 LAO; Tripura 2018 KHM; von Seidlein 2019 VNM). Five of these studies also added a single dose of primaquine at each round (Landier 2017 MMRa; McLean 2021 MMR; Morris 2018 TZA; Pongvongsa 2018 LAO; von Seidlein 2019 VNM). The interval between rounds varied across studies. In Eisele 2020 ZMBa, MDA was administered at the start of the rainy season, during the rainy season, during the dry season, and at the start of the rainy season. In McLean 2021 MMR, three rounds of MDA were administered one month apart during the dry season. The two rounds of MDA in Morris 2018 TZA were administered two months apart. The three rounds of MDA in Landier 2017 MMRa and Pongvongsa 2018 LAO, conducted one month apart, were administered at the start and during rainy season. Three rounds of MDA in Tripura 2018 KHM were administered during rainy season, while three rounds of MDA took place at the end of the transmission season in von Seidlein 2019 VNM.

Most studies administering MDA with dihydroartemisinin piperaquine excluded pregnant women in their first trimester, with the exception of Tripura 2018 KHM, which excluded all pregnant women from MDA. Studies administering a single dose of primaquine excluded all pregnant women from primaquine. After these exclusions, MDA was administered to all persons three months of age and older (Eisele 2020 ZMBa), six months of age and older (Landier 2017 MMRa; Morris 2018 TZA; Pongvongsa 2018 LAO; Tripura 2018 KHM; von Seidlein 2019 VNM), or 12 months of age and older (McLean 2021 MMR). MDA coverage ranged from 66% to 95% in the first round and was generally lower (56% to 99%) in subsequent rounds for all studies except von Seidlein 2019 VNM, in which coverage increased in subsequent rounds. All studies provided ITNs to all study arms. In addition, Eisele 2020 ZMBa and Morris 2018 TZA included a single round of IRS at baseline; Landier 2017 MMRa, Pongvongsa 2018 LAO, Tripura 2018 KHM, and von Seidlein 2019 VNM provided uninterrupted access to case management in study villages; and Shekalaghe 2011 TZA included a single treatment campaign for trachoma with azithromycin.

One study administered a single round of sulfadoxine‐pyrimethamine plus artesunate with a single dose of primaquine three to five weeks prior to the start of the rainy season (Shekalaghe 2011 TZA). All persons older than 12 months received MDA. However, pregnant women or individuals who were anaemic did not receive primaquine.

Outcomes

Outcome data are summarized in Table 2.

Open in table viewer
Table 2. Description of outcomes

Study ID (design)

Parasitaemia prevalence

Parasitaemia incidence

Confirmed malaria illness incidence

All‐cause or malaria‐specific mortality

Gametocytaemia prevalence

Adverse effects

Eisele 2020 ZMBa (cRCT)

Yes

Yes

Yes

No

No

Yes

Eisele 2020 ZMBb (cRCT)

Yes

Yes

Yes

No

No

Yes

Escudie 1962 BFA (CBA)

Yes

No

No

No

Yes

No

Landier 2017 MMRa (cRCT)

Yes

No

Yes

No

No

Yes

McLean 2021 MMR (cRCT)

Yes

No

No

No

No

Yes

Molineaux 1980 NGA (CBA)

Yes

No

No

No

Yes

No

Morris 2018 TZA (cRCT)

Yes

No

Yes

No

No

Yes

Pongvongsa 2018 LAO (cRCT)

Yes

No

No

No

No

Yes

Roberts 1964 KEN (CBA)

Yes

No

No

No

No

No

Shekalaghe 2011 TZA (cRCT)

Yes

No

Yes

No

Yes

Yes

Tripura 2018 KHM (cRCT)

Yes

No

Yes

No

No

Yes

von Seidlein 2003 GMB (cRCT)

Yes

Yes

Yes

Yes

Yes

Yes

von Seidlein 2019 VNM (cRCT)

Yes

No

No

No

No

Yes

CBA = controlled before‐and‐after study, cRCT = cluster‐randomized controlled trial

Moderate‐ to high‐transmission areas

All outcomes reported were for P falciparum. In Eisele 2020 ZMBb, parasitaemia prevalence was measured by RDT prior to MDA, during MDA (following the first two rounds), and one to three months after MDA. In von Seidlein 2003 GMB, parasitaemia prevalence was measured by microscopy prior to MDA and four to six months after MDA. Both studies also reported parasitaemia incidence, but measured the outcome differently. Eisele 2020 ZMBb measured parasitaemia incidence by RDT through active surveillance in a cohort of persons three months of age and older, for two months following the fourth round of MDA. In von Seidlein 2003 GMB, parasitaemia incidence was captured through weekly surveillance in children under 11 years for cases with temperatures of 37.5 °C and higher, and P falciparum parasitaemia above 5000 parasites per µL by microscopy for five months following MDA. Eisele 2020 ZMBb reported confirmed malaria illness incidence prior to MDA and one to three months after MDA. Only von Seidlein 2003 GMB reported the secondary outcomes of gametocytaemia prevalence (pre‐MDA and four to six months after MDA) and malaria‐specific mortality. Both studies monitored adverse effects in the MDA study arm.

Very low‐ to low‐transmission areas

All studies reported P falciparum parasitaemia prevalence, but the measurement of outcomes differed across studies with respect to follow‐up time points, population sampled, and detection method. Eisele 2020 ZMBa, Landier 2017 MMRa, McLean 2021 MMR, Pongvongsa 2018 LAO, Tripura 2018 KHM, and von Seidlein 2019 VNM reported prevalence at baseline before MDA. The timing of the first prevalence measure reported by Morris 2018 TZA started and extended past the first round of MDA and, therefore, was classified as during MDA. Eisele 2020 ZMBa also reported parasitaemia prevalence during MDA following the first two MDA rounds. All studies reported the outcome at one to three months after MDA. Four studies reported prevalence at four to six months after MDA (Landier 2017 MMRa; Pongvongsa 2018 LAO; Tripura 2018 KHM; von Seidlein 2019 VNM). Five studies provided prevalence at 7 to 12 months after MDA (Landier 2017 MMRa; McLean 2021 MMR; Pongvongsa 2018 LAO; Tripura 2018 KHM; von Seidlein 2019 VNM). Additionally, McLean 2021 MMR reported outcomes at 13 to 18 months after MDA and at all additional follow‐up intervals through to 31 to 36 months after MDA.

Eisele 2020 ZMBa measured parasitaemia prevalence by RDT in children aged 5 years and older. Five studies evaluated the outcome by ultrasensitive PCR (uPCR) in either adults aged 18 to 55 years (McLean 2021 MMR), or persons six months of age and older (Landier 2017 MMRa; Pongvongsa 2018 LAO; Tripura 2018 KHM; von Seidlein 2019 VNM). Morris 2018 TZA measured prevalence by 18s‐quantitative PCR (qPCR) in persons of all ages. P falciparum parasitaemia incidence was only reported in Eisele 2020 ZMBa and measured by RDT through active surveillance in a cohort of persons three months of age and older, for two months following MDA. Confirmed malaria illness incidence was reported prior to MDA and at: one to three months after MDA in Eisele 2020 ZMBa, Morris 2018 TZA, and Tripura 2018 KHM; four to six months after MDA in Morris 2018 TZA; 7 to 12 months after MDA in Landier 2017 MMRa, Morris 2018 TZA, and Tripura 2018 KHM; and 13 to 18 months after MDA only in Morris 2018 TZA. All studies monitored adverse effects.

Shekalaghe 2011 TZA reported parasitaemia prevalence by microscopy and quantitative nucleic acid sequence‐based amplification  (QT‐NASBA), confirmed malaria illness incidence, and gametocytaemia prevalence by microscopy in all ages prior to and after MDA up to four months, but these data were omitted from quantitative synthesis at all time points due to either zero events pre‐MDA (outcomes by microscopy) or unclear timing of post‐MDA events (outcomes by QT‐NASBA).

P vivax parasitaemia prevalence was reported by Landier 2017 MMRa, McLean 2021 MMR, Pongvongsa 2018 LAO, Tripura 2018 KHM, and von Seidlein 2019 VNM, and included the same pre‐MDA and post‐MDA follow‐up time points, detection method, and population sampled as P falciparum parasitaemia. Confirmed malaria illness incidence for P vivax was reported prior to MDA in Tripura 2018 KHM, and at 7 to 12 months after MDA in Landier 2017 MMRa; and Tripura 2018 KHM.

Non‐randomized controlled studies

We included three CBA studies in qualitative syntheses. All studies were conducted in areas of moderate to high endemicity (moderate: Roberts 1964 KEN; high: Escudie 1962 BFA; Molineaux 1980 NGA). All studies were conducted in sub‐Saharan Africa in the countries of Burkina Faso (Escudie 1962 BFA), Kenya (Roberts 1964 KEN), and Nigeria (Molineaux 1980 NGA).

Escudie 1962 BFA included six arms (chloroquine plus primaquine or amodiaquine plus primaquine every four weeks; chloroquine plus primaquine or amodiaquine plus primaquine every two weeks; chloroquine plus primaquine or amodiaquine plus primaquine every four weeks with IRS; chloroquine plus primaquine or amodiaquine plus primaquine every two weeks with IRS; IRS only; non‐IRS control). Of these six arms, three arms with two comparisons contributed to this review: chloroquine plus primaquine or amodiaquine plus primaquine every four weeks ('low frequency'); chloroquine plus primaquine or amodiaquine plus primaquine every two weeks ('high frequency'); and non‐IRS control. Given similar mechanisms of antimalarial action for chloroquine and amodiaquine, and the combined reporting of outcomes at post‐MDA time points, these drugs were not analysed as separate groups within the six arms described. Of the four arms included in Molineaux 1980 NGA (low frequency MDA plus IRS, high frequency MDA plus IRS, IRS only, and no intervention), we included three arms with two comparisons in this review: low frequency MDA plus IRS, high frequency MDA plus IRS, and IRS only). Roberts 1964 KEN compared MDA to no MDA and all data provided comparisons for this review.

Interventions

Characteristics of the intervention have been summarized in Table 1.

All studies administered multiple rounds of MDA using different antimalarials and at different frequencies of MDA rounds. Escudie 1962 BFA administered MDA with a single dose of chloroquine plus primaquine or amodiaquine plus primaquine every four weeks for seven rounds, or every two weeks for 15 rounds to the entire population (all ages). MDA coverage ranged from 75% to 91% per round in the four‐week low frequency MDA arm and 84% to 97% per round in the two‐week high frequency MDA arm. Molineaux 1980 NGA administered MDA with sulfalene‐pyrimethamine either every 10 weeks ('low frequency MDA') or every 2 weeks during the wet season and every 10 weeks during the dry season ('high frequency MDA') to all ages except infants prior to their first malaria episode. MDA coverage was similar across arms, ranging from 73% to 92% per round in the low frequency MDA arm and 72% to 91% in the high frequency MDA arm. In Roberts 1964 KEN, two rounds of MDA with pyrimethamine were administered once a year (at the start of the rainy season) to all ages with coverage of 95% in round one and 93% in round two.

Molineaux 1980 NGA administered IRS using propoxur for three to four rounds per year in all arms, but there were no co‐interventions in the other two studies.

Outcomes

Outcome data are summarized in Table 2.

All outcomes reported were for P falciparum. All studies reported parasitaemia prevalence by microscopy at pre‐MDA and during‐MDA time points, but outcomes were assessed for different lengths of post‐MDA follow‐up and in different age groups. Molineaux 1980 NGA provided no post‐MDA data, Escudie 1962 BFA reported parasitaemia prevalence at one to three months after MDA, and Roberts 1964 KEN provided parasitaemia prevalence at time points up to 7 to 12 months after MDA. Two studies assessed parasitaemia prevalence in all age groups (Molineaux 1980 NGA; Roberts 1964 KEN), and Escudie 1962 BFA assessed parasitaemia prevalence only in children aged two to nine years. Gametocytaemia prevalence was reported by Escudie 1962 BFA in children aged two to nine years prior to MDA, during MDA, and one to three months after MDA, and by Molineaux 1980 NGA in all ages, prior to MDA and during MDA (no post‐MDA time points). No other outcomes, including adverse effects, were reported by these studies.

Excluded studies

We excluded 25 articles for the following reasons: 13 were duplicates (cross‐referenced articles of another study); 10 failed to meet the study design inclusion criteria; and two were the incorrect intervention or inadequate treatment dose. We have provided detailed reasons in the Characteristics of excluded studies.

Two additional excluded studies were unpublished trials that are awaiting classification (Characteristics of studies awaiting classification). We made several attempts to contact Song TGO, but did not receive a response. We were unable to obtain sufficient information from authors of El‐Sayed SDN to assess the trial for eligibility. However, we plan to screen the study should the authors publish the results. 

Risk of bias in included studies

Our summary assessment for risks of bias are shown in Figure 2 and Figure 3, and details are provided for each study in the Characteristics of included studies tables.

Figure 2
Risk of bias summary: review authors' judgements about each risk of bias item for each included study

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

Figure 3
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies

Allocation

We assessed nine studies as low risk of bias for random sequence generation due to the use of a computerized randomization algorithm (Eisele 2020 ZMBa; Eisele 2020 ZMBb; Landier 2017 MMRa; Morris 2018 TZA; Pongvongsa 2018 LAO; Shekalaghe 2011 TZA; Tripura 2018 KHM; von Seidlein 2003 GMB; von Seidlein 2019 VNM). We assessed the remaining four studies as high risk of bias, due to lack of randomization (Escudie 1962 BFA; Molineaux 1980 NGA; Roberts 1964 KEN), or the use of a non‐computerized randomization method (McLean 2021 MMR). We judged all 10 cRCTs to be at low risk for bias related to allocation concealment since allocation was performed by an institution, and all non‐randomized studies to be at high risk of bias for allocation concealment.

Blinding

Parasitaemia prevalence

We assessed two placebo‐controlled studies as low risk for performance bias since participants and study staff were blinded (Shekalaghe 2011 TZA; von Seidlein 2003 GMB). We judged all other studies, which included non‐MDA control groups, as unclear risk for performance bias since participants were not blinded and it was unclear whether or how this could affect this outcome.

Outcome assessment by laboratory staff was blinded to study arm in nine studies, and we judged these as low risk (Landier 2017 MMRa; McLean 2021 MMR; Molineaux 1980 NGA; Morris 2018 TZA; Pongvongsa 2018 LAO; Shekalaghe 2011 TZA; Tripura 2018 KHM; von Seidlein 2003 GMB; von Seidlein 2019 VNM). Three studies did not describe blinding of laboratory staff, and we judged these as unclear risk of detection bias (Eisele 2020 ZMBa; Eisele 2020 ZMBb; Roberts 1964 KEN). Outcome assessment in one study was clearly unblinded, and we considered this study as high risk of detection bias (Escudie 1962 BFA).

Parasitaemia incidence

We judged all three studies reporting parasitaemia incidence as low risk of both performance and detection bias since this outcome was evaluated through active surveillance, so lack of blinding of participants and study staff was presumed to minimally affect the outcome (Eisele 2020 ZMBa; Eisele 2020 ZMBb), or the trial was placebo‐controlled (von Seidlein 2003 GMB).

Confirmed malaria illness incidence

We judged one placebo‐controlled trial as having low risk of performance and detection bias due to blinding of participants, study staff, and outcome assessment (Shekalaghe 2011 TZA). We judged five studies as having unclear risk since it was unclear how lack of blinding would affect care‐seeking behaviours by participants, or outcome assessment (Eisele 2020 ZMBa; Eisele 2020 ZMBb; Landier 2017 MMRa; Morris 2018 TZA; Tripura 2018 KHM).

Incomplete outcome data

Parasitaemia prevalence

We judged one trial that reported large variations in the size of the population evaluated for parasitaemia prevalence outcomes at different time points as having high risk of bias since, based on the described methodology, the population evaluated should have remained consistent (McLean 2021 MMR). The remaining eight studies had a low risk of attrition bias.

Parasitaemia incidence

We judged one study as high risk of bias due to a large proportion of missed malaria infections (particularly, afebrile or low parasite density malaria infections) due to outcome definitions (von Seidlein 2003 GMB). We judged two studies as unclear risk of bias since loss to follow‐up was not provided by baseline endemicity strata (Eisele 2020 ZMBa, Eisele 2020 ZMBb).

Confirmed malaria illness incidence

We judged all six studies reporting this outcome as having low risk of bias (Eisele 2020 ZMBa; Eisele 2020 ZMBb; Landier 2017 MMRa; Morris 2018 TZA; Tripura 2018 KHM; Shekalaghe 2011 TZA).

Selective reporting

All studies reported the outcomes that were pre‐specified and were judged as having low risk of reporting bias.

Other potential sources of bias

No cRCTs were judged to have a high or unclear risk of bias from recruitment, loss of clusters, incorrect analyses, or comparability with individually randomized trials. All studies were judged to be at low risk for other biases.

Baseline imbalance

We judged five studies to have high risk of bias due to unbalanced baseline malaria prevalence in intervention and control arms (Landier 2017 MMRa; Morris 2018 TZA; Pongvongsa 2018 LAO; Roberts 1964 KEN; Tripura 2018 KHM). Two studies did not report baseline demographic characteristics (other risk factors for malaria), and we judged these to have unclear risk of bias (Escudie 1962 BFA; Molineaux 1980 NGA). The remaining studies were balanced for malaria and demographic characteristics across arms at baseline.

Contamination protection

The study arms for six studies were either paired by geographic proximity or located in close geographic proximity with a high potential for population movement. We judged these studies as having a high risk of contamination bias (Landier 2017 MMRa; McLean 2021 MMR; Pongvongsa 2018 LAO; Roberts 1964 KEN; Tripura 2018 KHM; von Seidlein 2019 VNM). Two studies did not describe methods for contamination protection, and we judged these as unclear risk (Escudie 1962 BFA; Morris 2018 TZA). The remaining four studies included buffer zones to minimize contamination of outcome assessment, and we judged these as low risk (Eisele 2020 ZMBa; Eisele 2020 ZMBb; Shekalaghe 2011 TZA; von Seidlein 2003 GMB).

Effects of interventions

See: Summary of findings 1 MDA compared to no MDA for Plasmodium falciparum malaria (moderate to high endemicity, short‐term follow‐up); Summary of findings 2 MDA compared to no MDA for Plasmodium falciparum malaria (very low to low endemicity, short‐term follow‐up); Summary of findings 3 MDA compared to no MDA for Plasmodium falciparum malaria (very low to low endemicity, long‐term follow‐up); Summary of findings 4 MDA compared to no MDA for P vivax malaria (very low to low endemicity, short‐term follow‐up); Summary of findings 5 MDA compared to no MDA for P vivax malaria (very low to low endemicity, long‐term follow‐up)

Cluster‐randomized trials

Moderate‐ to high‐transmission areas
Effects reported on P falciparum outcomes

At one to three months after MDA

Eisele 2020 ZMBb found a non‐significant increase in P falciparum parasitaemia prevalence (RR 1.76, 95% CI 0.58 to 5.36; 1 study; Analysis 1.1), a significant reduction in parasitaemia incidence by 39% (rate ratio 0.61, 95% CI 0.40 to 0.92; 1 study; Analysis 1.2), and a non‐significant reduction in confirmed malaria illness incidence by 59% (RR 0.41, 95% CI 0.04 to 4.51; 1 study; Analysis 1.3) in MDA compared to non‐MDA clusters at one to three months after MDA.

At four to six months after MDA

As reported by von Seidlein 2003 GMB, at four to six months after MDA, MDA did not reduce P falciparum parasitaemia prevalence (RR 1.18, 95% CI 0.89 to 1.56; 1 study; Analysis 1.1), and also did not reduce the secondary outcomes of P falciparum gametocytaemia prevalence (RR 1.13, 95% CI 0.20 to 6.54; 1 study; Analysis 1.4), and malaria‐specific mortality (RR 1.42, 95% CI 0.12 to 17.15; 1 study; Analysis 1.5).

Very low‐ to low‐transmission areas
Effects reported on P falciparum outcomes

At less than one month after MDA

McLean 2021 MMR reported that MDA reduced P falciparum parasitaemia prevalence by 88% immediately (less than one month) after MDA (RR 0.12, 95% CI 0.03 to 0.52; 1 study; Analysis 2.1).

At one to three months after MDA

Based on results from seven studies, MDA significantly reduced P falciparum parasitaemia prevalence by 75% one to three months after MDA (RR 0.25, 95% 0.15 to 0.41; 7 studies; Analysis 2.1). The largest reduction was observed by Landier 2017 MMRa in Myanmar (RR 0.06, 95% CI 0.01 to 0.45; 1 study; Analysis 2.1), while the opposite effect was found by Morris 2018 TZA in Zanzibar (RR 1.34, 95% CI 0.30 to 5.92; Analysis 2.1). Although the I2 value(31%) did not indicate inconsistency in the results, we explored differences in malaria epidemiology by continent as a cause of heterogeneity in this effect. When the effects were sub‐grouped by continent (Africa and Asia) in post‐hoc analysis (Analysis 4.1), there was no effect of MDA in two studies in sub‐Saharan Africa (pooled RR 0.97, 95% CI 0.32 to 2.98; Eisele 2020 ZMBa; Morris 2018 TZA), but a larger reduction in the five studies in Southeast Asia (RR 0.19, 95% CI 0.11 to 0.33; Landier 2017 MMRa; McLean 2021 MMR; Pongvongsa 2018 LAO; Tripura 2018 KHM; von Seidlein 2019 VNM).

Eisele 2020 ZMBa measured the effect of MDA compared to no MDA on P falciparum parasitaemia incidence in very low‐ to low‐endemicity areas, and found a statistically significant 63% reduction one to three months after MDA (rate ratio 0.37, 95% CI 0.21 to 0.66; 1 study; Analysis 2.2).

MDA also reduced confirmed malaria illness incidence (rate ratio 0.58, 95% CI 0.12 to 2.73, 2 studies; Analysis 2.3). However, the effect was imprecise and not statistically significant.

At four to six months after MDA

Based on findings from four studies in Southeast Asia, at four to six months after MDA, there was a non‐significant reduction in P falciparum parasitaemia prevalence by 37% (RR 0.63, 95% CI 0.36 to 1.12; 4 studies; Analysis 2.1) in MDA compared to non‐MDA clusters. Effects across studies included few malaria cases and were therefore very imprecise. The strongest reduction was observed by Pongvongsa 2018 LAO in Laos (RR 0.07, 95% CI 0 to 1.31; Analysis 2.1), while the weakest effect was found by Landier 2017 MMRa in Myanmar (RR 0.89, 95% CI 0.27 to 2.95; Analysis 2.1).

There was no effect of MDA on confirmed malaria illness incidence four to six months after MDA as reported by a single study, Morris 2018 TZA (rate ratio 0.93, 95% CI 0.07 to 12.43; Analysis 2.3).

At 7 to 12 months after MDA

At 7 to 12 months after MDA, there was a small but non‐significant reduction in P falciparum parasitaemia prevalence (RR 0.86, 95% CI 0.55 to 1.36; 5 studies; Analysis 2.1) in MDA compared to non‐MDA clusters. Due to few malaria cases, the effect estimates across studies were very imprecise. Additionally, the direction of effect was inconsistent across studies, with two studies showing a non‐significant increase in P falciparum parasitaemia prevalence in the MDA compared to no‐MDA arm (Landier 2017 MMRa; von Seidlein 2019 VNM), while three studies found that MDA reduced parasitaemia prevalence (McLean 2021 MMR; Pongvongsa 2018 LAO; Tripura 2018 KHM).

MDA reduced confirmed malaria illness incidence 7 to 12 months after MDA (rate ratio 0.47, 95% CI 0.21 to 1.03; 3 studies; Analysis 2.3). However, the effect was not statistically significant and there was substantial heterogeneity (I2=72%, Analysis 2.3).

At longer time periods (> 12 months) after MDA

At 13 to 18 months after MDA, MDA did not significantly reduce P falciparum parasitaemia prevalence in McLean 2021 MMR (RR 0.82, 95% CI 0.20 to 3.34; 1 study; Analysis 2.1) or confirmed malaria illness incidence in Morris 2018 TZA (rate ratio 0.77, 95% CI 0.2 to 3.03; 1 study; Analysis 2.3). McLean 2021 MMR reported no effect on parasitaemia prevalence at 19 to 24 months, 25 to 30 months and 31 to 36 months after MDA (Analysis 2.1).

Effects reported on P vivax outcomes

At less than one month after MDA

McLean 2021 MMR showed that MDA had an immediate and large reduction on P vivax parasitaemia prevalence at less than one month after MDA (RR 0.18, 95% CI 0.08‐0.40; 1 study; Analysis 3.1).

At one to three months after MDA

Although there was considerable heterogeneity across five studies (I2 = 84%) in the effect of MDA on P vivax parasitaemia prevalence one to three months after MDA, we meta‐analysed the results since the direction of effect was consistent and all reported effect estimates were very imprecise due to a small number of malaria events. At one to three months after MDA, MDA significantly reduced P vivax parasitaemia prevalence by 85% (RR 0.15, 95% CI 0.10 to 0.24; 5 studies; Analysis 3.1).

At four to six months after MDA

MDA significantly reduced P vivax parasitaemia prevalence by 22% at four to six months after MDA (RR 0.78, 95% CI 0.63 to 0.95; 4 studies; Analysis 3.1).

At 7 to 12 months after MDA

At 7 to 12 months after MDA, MDA did not reduce P vivax parasitaemia prevalence (RR 1.12, 95% CI 0.94 to 1.34; 5 studies; Analysis 3.1). There was a non‐significant increase in confirmed malaria illness incidence (pooled rate ratio 1.38, 95% CI 0.97 to 1.95; 2 studies; Analysis 3.2) in MDA compared to non‐MDA clusters.

At longer time periods (> 12 months) after MDA

McLean 2021 MMR reported that MDA did not reduce parasitaemia prevalence at 13 to 18 months, 19 to 24 months, 25 to 30 months, and 31 to 36 months after MDA (Analysis 3.1).

Adverse effects

Adverse effects (AEs) of MDA were reported by all ten cRCTs included in qualitative synthesis (Eisele 2020 ZMBa; Eisele 2020 ZMBb; Landier 2017 MMRa; McLean 2021 MMR; Morris 2018 TZA; Pongvongsa 2018 LAO; Shekalaghe 2011 TZA; Tripura 2018 KHM; von Seidlein 2003 GMB; von Seidlein 2019 VNM). We have provided details by study in the Characteristics of included studies tables. McLean 2021 MMR reported AEs by relatedness to MDA drug, but did not provide diagnoses of AEs. Eisele 2020 ZMBa and Eisele 2020 ZMBb reported AEs by MDA and fMDA arms combined and in aggregate across low‐ and high‐transmission strata clusters.

Two studies reported one serious AE with MDA using dihydroartemisinin piperaquine (Eisele 2020 ZMBa or Eisele 2020 ZMBb), or dihydroartemisinin piperaquine plus primaquine (McLean 2021 MMR). One study reported that 0.5% of all AEs reported after MDA with dihydroartemisinin piperaquine plus primaquine were perceived as serious by participants (Morris 2018 TZA). Two studies administering dihydroartemisinin piperaquine plus primaquine reported multiple serious AEs that were not drug‐related (Landier 2017 MMRa; von Seidlein 2019 VNM). Landier 2017 MMRa also reported three cases of black urine. One study reported a possibly drug‐related severe skin reaction following administration of MDA using sulfadoxine‐pyrimethamine plus artesunate with primaquine (Shekalaghe 2011 TZA).

Four studies indicated stomach pains or diarrhoea and vomiting as commonly reported AEs to MDA with dihydroartemisinin piperaquine (Eisele 2020 ZMBa; Eisele 2020 ZMBb), dihydroartemisinin piperaquine plus primaquine (Morris 2018 TZA, Pongvongsa 2018 LAO, von Seidlein 2019 VNM), and sulfadoxine‐pyrimethamine plus artesunate (von Seidlein 2003 GMB). Dizziness and fever were mentioned as minor AEs to MDA with dihydroartemisinin piperaquine (Morris 2018 TZA, Tripura 2018 KHM) and sulfadoxine‐pyrimethamine plus artesunate (von Seidlein 2003 GMB). Dizziness was also a common AE in two studies with MDA using dihydroartemisinin piperaquine plus primaquine (Landier 2017 MMRa, Pongvongsa 2018 LAO). Three studies also indicated complaints of nausea, headache, and fatigue in relation to MDA with dihydroartemisinin piperaquine plus primaquine (Morris 2018 TZA, von Seidlein 2019 VNM), or dihydroartemisinin piperaquine alone (Tripura 2018 KHM). Pruritis or itching were reported by studies administering dihydroartemisinin piperaquine plus primaquine (Landier 2017 MMRa; Pongvongsa 2018 LAO) and a study administering MDA with sulfadoxine‐pyrimethamine plus amodiaquine (von Seidlein 2003 GMB). Common cold or dry cough was reported as a common minor AE in two studies using MDA with dihydroartemisinin piperaquine (Eisele 2020 ZMBa; Eisele 2020 ZMBb), or MDA with dihydroartemisinin piperaquine plus primaquine (Pongvongsa 2018 LAO).

Among the four studies that reported the distribution of AEs (as either percentage of participants or percentage of doses with an AE), the frequency of at least one AE report was highest in MDA with sulfadoxine‐pyrimethamine plus amodiaquine (33% of participants, von Seidlein 2003 GMB), and lower in MDA with dihydroartemisinin piperaquine (0.24% (Eisele 2020 ZMBa or Eisele 2020 ZMBb) to 11.6% (Morris 2018 TZA) of participants) or dihydroartemisinin piperaquine plus primaquine (3.6% of participants in McLean 2021 MMR).

Non‐randomized studies

P falciparum parasitaemia prevalence

In the three non‐randomized trials, we summarized changes in parasitaemia prevalence from baseline to during‐MDA or post‐MDA, in MDA compared to no‐MDA groups, using a DiD analysis (Table 3).

Open in table viewer
Table 3. Difference‐in‐differences analysis of P falciparum parasitaemia prevalence in non‐randomized studies

Study

Intervention, % (n)

Control % (n)

Difference‐in‐differences, percentage pointsa

Pre‐MDA

During MDA

Post‐MDA

Pre‐MDA

During MDA

Post‐MDA

During MDA

Post‐MDA

1 to 3 months

4 to 6 months

7 to 12 months

1 to 3 months

4 to 6 months

7 to 12 months

1 to 3 months

4 to 6 months

7 to 12 months

Escudie 1962 BFAb

Low frequency MDA with AQ‐PQ or CQ‐PQ

67.6 (190)

21.6 (75)

38.3 (105)

ND

ND

59.4 (129)

74.8 (517)

386 (72.3)

ND

ND

‐61.4

‐42.1

ND

ND

High frequency MDA with AQ‐PQ or CQ‐PQ

33.6 (131)

12.6 (59)

61.4 (286)

ND

ND

59.4 (129)

74.8 (517)

386 (72.3)

ND

ND

‐36.3

14.9

ND

ND

Molineaux 1980 NGAc

Low frequency MDA with SP

41.8 (525)

1.9 (40)

ND

ND

ND

49.1 (493)

32.5 (380)

ND

ND

ND

‐23.2

ND

ND

ND

High frequency MDA with SP

44.9 (754)

7.3 (109)

ND

ND

ND

49.1 (493)

32.5 (380)

ND

ND

ND

‐20.9

ND

ND

ND

Roberts 1964 KEN

MDA with pyrimethamine

8.3 (25)

9 (188)

2.9 (26)

(27) (4.5)

(15) (5)

18 (154)

34.4 (723)

40.7 (366)

37 (222)

26 (78)

‐15.8

‐28.1

‐22.8

‐11.3

AQ = amodiaquine, CQ = chloroquine, MDA = mass drug administration, ND = no data, PQ = primaquine, SP = sulfalene‐pyrimethamine

aCalculated as difference in proportion at the time period of during MDA or post‐MDA minus the proportion at pre‐MDA in the intervention and control separately and the difference in these two proportion differences between the intervention and control groups.
bMDA with AQ‐PQ or CQ‐PQ either every 4 weeks ('low frequency MDA') or every 2 weeks ('high frequency MDA').
cMDA with sulfalene‐pyrimethamine either every 10 weeks (low frequency MDA') or every 2 weeks during the wet season and every 10 weeks during the dry season ('high frequency MDA') to all ages except infants prior to their first malaria episode.

MDA appeared to reduce parasitaemia prevalence in all studies during MDA compared to pre‐MDA, with a wide range of DiD percentage point reductions from ‐15.8 (Roberts 1964 KEN), to ‐61.4 (Escudie 1962 BFA). The smallest reduction was observed in Roberts 1964 KEN, in which two rounds of MDA using pyrimethamine were administered one year apart and, therefore, the during‐MDA time period included multiple surveys between the two rounds. Similar percentage point reductions from baseline were observed between both the low‐frequency MDA plus IRS (‐23.2 percentage points) and high frequency MDA plus IRS (‐20.9 percentage points) compared to the control (IRS only) group in Molineaux 1980 NGA during the period that MDA was administered with sulfalene‐pyrimethamine. In Escudie 1962 BFA, there was a substantially greater DiD reduction in parasitaemia prevalence in the low‐frequency MDA group (amodiaquine plus primaquine or chloroquine plus primaquine every four weeks, ‐61.4 percentage points) compared to the high‐frequency MDA group (amodiaquine plus primaquine or chloroquine plus primaquine every two weeks; ‐36.3 percentage points) at during‐MDA compared to pre‐MDA time periods.

At one to three months after MDA, parasitaemia prevalence was reduced in the low‐frequency MDA arm of Escudie 1962 BFA by ‐42.1 percentage points, but there was an increase in parasitaemia prevalence in the high‐frequency MDA arm compared to control, as reflected by a +14.9 percentage point DiD. There was an initial reduction in parasitaemia prevalence (‐28.1 percentage points) found in Roberts 1964 KEN at one to three months after MDA, which decreased over time to ‐22.8 percentage points at four to six months after MDA, and to ‐11.3 percentage points at 7 to 12 months after MDA. Given the introduction of new interventions in Molineaux 1980 NGA immediately following MDA, there were no post‐MDA data available.

P falciparum gametocytaemia prevalence

In the two non‐randomized trials reporting gametocytaemia prevalence, we summarized changes in the effect of MDA from baseline to during‐MDA or post‐MDA using DiD analysis (Table 4).

Open in table viewer
Table 4. Difference‐in‐differences analysis of P falciparum gametocytaemia prevalence in non‐randomized studies

Study

Intervention, % (n)

Control, % (n)

Difference‐in‐differences

percentage pointa

Pre‐MDA

During MDA

Post‐MDA

1 to 3 months

Pre‐MDA

During MDA

Post‐MDA

1 to 3 months

During MDA

Post‐MDA

1 to 3 months

Escudie 1962 BFAb

Low frequency MDA with AQ‐PQ or CQ‐PQ

20.3 (57)

0.9 (3)

38.3 (35)

19.4 (42)

14 (97)

19.1 (102)

‐14.1

18.3

High frequency MDA with AQ‐PQ or CQ‐PQ

8.2 (32)

1.9 (9)

61.4 (107)

19.4 (42)

14 (97)

19.1 (102)

‐1.0

53.4

Molineaux 1980 NGAc

Low frequency MDA with SP

10.1 (127)

0.6 (12)

ND

12.4 (124)

7.9 (92)

ND

‐5.0

ND

High frequency MDA with SP

12.4 (208)

3.2 (48)

ND

12.4 (124)

7.9 (92)

ND

‐4.7

ND

AQ = amodiaquine, CQ = chloroquine, MDA = mass drug administration, ND = no data, SP = sulfalene‐pyrimethamine

aCalculated as difference in proportion at the time period of during MDA or post‐MDA minus the proportion at pre‐MDA in the intervention and control separately and the difference in these two proportion differences between the intervention and control groups.
bMDA with AQ‐PQ or CQ‐PQ either every 4 weeks ('low frequency MDA') or every 2 weeks ('high frequency MDA').
cMDA with sulfalene‐pyrimethamine either every 10 weeks ('low frequency MDA') or every 2 weeks during the wet season and every 10 weeks during the dry season ('high frequency MDA') to all ages except infants prior to their first malaria episode.

From pre‐MDA to during‐MDA, with the exception of the low‐frequency MDA arm in Escudie 1962 BFA, there was a small DiD reduction on gametocytaemia prevalence, ranging from ‐1 percentage point (Escudie 1962 BFA high‐frequency MDA) to ‐5.0 percentage points in both arms of Molineaux 1980 NGA. A large reduction from pre‐MDA (‐14.1 percentage points) was observed in the low‐frequency MDA arm of Escudie 1962 BFA during MDA. However, compared to the pre‐MDA period, gametocytaemia prevalence at one to three months after MDA was increased in both the low‐frequency MDA arm (18.3 percentage points) and high‐frequency MDA arm (53.4 percentage points) in Escudie 1962 BFA compared to no MDA.

Discussion

Summary of main results

We included 13 studies in this review: 10 cRCTs, of which two were from areas of moderate to high transmission, and eight were from areas of very low to low endemicity; and three CBAs, all from settings of moderate to high transmission.

Areas of moderate‐ to high‐endemicity (cRCTs)

The two included studies were both conducted in sub‐Saharan Africa. We included MDA versus no‐MDA comparisons from von Seidlein 2003 GMB, which examined the effect of one round of MDA with sulfadoxine‐pyrimethamine plus artesunate, and from Eisele 2020 ZMBb, which examined the effect of four rounds of MDA with dihydroartemisinin piperaquine. No co‐interventions were implemented in the von Seidlein 2003 GMB study, but IRS, ITNs, and enhanced community case management practices were included in all arms of Eisele 2020 ZMBb.

Based on these data, in comparison to no‐MDA:

  • at one to three months after MDA, P falciparum parasitaemia prevalence may be higher (low‐certainty evidence), parasitaemia incidence is probably lower (moderate‐certainty evidence), and confirmed malaria illness incidence may be lower (low‐certainty evidence) in MDA compared to no‐MDA, as reported by a single study (Eisele 2020 ZMBb);

  • at four to six months after MDA, MDA probably leads to little or no effect on P falciparum parasitaemia prevalence (moderate‐certainty evidence) and we do not know the effect on parasitaemia incidence (very low‐certainty evidence) as reported by a single study (von Seidlein 2003 GMB).

Longer‐term effects of MDA on outcomes were not reported by any included studies in moderate‐ to high‐transmission settings.

Areas of very low‐ to low‐endemicity (cRCTs)

Of the eight studies included in qualitative synthesis, three were carried out in sub‐Saharan Africa and five were conducted in Southeast Asia. One study that administered one round of MDA with sulfadoxine‐pyrimethamine plus artesunate plus primaquine was excluded from quantitative synthesis due to insufficient data (Shekalaghe 2011 TZA). Of the remaining seven trials, we included one comparison (multiple rounds of MDA with dihydroartemisinin piperaquine plus primaquine) each from Landier 2017 MMRa, McLean 2021 MMR, Morris 2018 TZA, Pongvongsa 2018 LAO, and von Seidlein 2019 VNM; and one comparison (multiple rounds of MDA with dihydroartemisinin piperaquine) each from Eisele 2020 ZMBa and Tripura 2018 KHM. All studies implemented ITNs, while Morris 2018 TZA and Eisele 2020 ZMBa also included IRS in both intervention and control arms.

Based on these data, in comparison to no‐MDA:

  • at less than one month after MDA, P falciparum parasitaemia prevalence may be lower in MDA compared to no‐MDA (low‐certainty evidence);

  • at one to three months after MDA, MDA probably reduces P falciparum parasitaemia incidence (moderate‐certainty evidence) and may reduce P falciparum parasitaemia prevalence (low‐certainty evidence), but we do not know if MDA has an effect on confirmed malaria illness incidence (very low‐certainty evidence);

  • at 4 to 6 months and 7 to 12 months after MDA, we do not know the effect of MDA on P falciparum parasitaemia prevalence and confirmed malaria illness incidence (very low‐certainty evidence);

  • as reported by a single study (McLean 2021 MMR), we do not know if MDA has an effect on P falciparum parasitaemia prevalence at longer‐term follow‐up periods (13 to 36 months, very low‐certainty evidence).

Despite the very low‐certainty evidence available for outcomes after four months post‐MDA, there was a general trend of a substantial reduction in P falciparum parasitaemia prevalence immediately following MDA and the strength of the reduction declined over time.

Five studies provided data for comparison of P vivax outcomes:

  • at less than one month after MDA, P vivax parasitaemia prevalence may be lower in MDA compared to no‐MDA (low‐certainty evidence);

  • at one to three months after MDA, P vivax parasitaemia prevalence may be lower in MDA compared to no‐MDA (low‐certainty evidence);

  • at four to six months after MDA, we do not know if MDA reduces P vivax parasitaemia prevalence (very low‐certainty evidence);

  • at 7 to 12 months after MDA, we do not know the effect of MDA on P vivax parasitaemia prevalence and confirmed malaria illness incidence (very low‐certainty evidence);

  • as reported by a single study (McLean 2021 MMR), we do not know if MDA has an effect on P vivax parasitaemia prevalence at longer‐term follow‐up periods (13 to 36 months, very low‐certainty evidence).

Very low‐certainty evidence was available for multiple outcomes. However, similar to P falciparum outcomes in this setting, there was a general trend of a large and immediate reduction in P vivax parasitaemia prevalence following MDA that waned over time.

In relation to the Objectives of this review:

  • in moderate‐ to high‐transmission settings, no studies provided data at longer‐term time periods to observe a 50% reduction in median malaria parasite prevalence or incidence, or both, at 12 months after MDA. However, the available studies showed a reduction in parasitaemia incidence, but not parasitaemia prevalence, at time periods prior to six months after MDA;

  • in very low‐ to low‐transmission settings, there was a strong reduction in P falciparum and P vivax parasitaemia prevalence immediately following MDA, that decreased over time. However, there was no evidence of interruption of transmission, defined as a reduction in number of indigenous malaria infections to zero at six months after MDA.

Overall completeness and applicability of evidence

This review highlights a renewed interest in MDA as a strategy to accelerate progress towards malaria elimination, given the more recent studies conducted since the previous review on this topic (Poirot 2013). In areas of moderate to high transmission, no studies reported outcomes under six months after MDA, but available evidence from earlier post‐MDA time periods showed that MDA decreased parasitaemia incidence, but not parasitaemia prevalence. In very low‐ to low‐transmission areas, several studies found substantial short‐term reductions in parasitaemia prevalence following MDA that were not sustained longer‐term. None of these studies found a reduction to zero indigenous malaria cases at any post‐MDA time period.

In both malaria transmission settings, we found that MDA compared to no MDA led to significant reductions in P falciparum parasitaemia incidence at one to three months after MDA, and in very low‐ to low‐transmission settings, correspondingly large and statistically significant reductions in parasitaemia prevalence at less than one month after MDA (one study) and at one to three months after MDA (seven studies). However, in moderate‐ to high‐transmission settings, this trend was not mirrored by a reduction in parasitaemia prevalence, based on a single study reporting at one to three months and four to six months after MDA. We believe one explanation for this counterintuitive finding of a reduction in incidence but not prevalence is due to differences in the study design and populations used to measure these outcomes. Studies measured incidence in a fixed cohort and prevalence in a random sample of the population in cross‐sectional surveys. Since the inclusion criteria for cohort studies typically requires establishment of a period of residency in the study area, cohort participants may also be more likely to receive at least one round of MDA compared to a cross‐section of the population that can include individuals who recently moved to the study area after the MDA round (i.e. did not receive the intervention) but prior to the survey. This distinction also raises the concern about risk of re‐introduction of malaria. It is possible that we found no effect of MDA on several outcomes due to the high risk of re‐introduction in the study areas of included studies, possibly due to a combination of high levels of migration and low effective MDA coverage (i.e. the proportion of the population that received at least one round of MDA).

In an area of moderate to high transmission, Eisele 2020 ZMBb found an unexpected increase in parasitaemia prevalence in the MDA arm compared to no‐MDA control at one to three months, following four rounds of MDA. Although the effect estimate was very imprecise, it is worth noting that parasitaemia prevalence decreased overall in the entire study area (both the lower‐ and higher‐transmission strata) and specifically in the moderate‐ to high‐transmission control clusters, from above 50% at baseline to below 10% following four rounds of MDA. Other interventions such as ITNs, IRS and enhanced community case management were implemented at high coverage in intervention and control arms at the start of the study and may have contributed to the overall decline in malaria transmission in the study area. This finding may also highlight the importance of implementing MDA as a component of a package of malaria control interventions to achieve reductions in malaria transmission.

Although not an important cause of heterogeneity in our analysis, the short‐term effects of MDA in studies conducted in very low‐ to low‐ transmission settings differed by geography. Trials conducted in Southeast Asia generally found a large reduction in parasitaemia prevalence at one to three months after MDA (Landier 2017 MMRa; McLean 2021 MMR; Pongvongsa 2018 LAO; Tripura 2018 KHM; von Seidlein 2019 VNM), while the two studies in Africa showed a smaller or no reduction (Eisele 2020 ZMBa; Morris 2018 TZA). All studies administered MDA with dihydroartemisinin piperaquine (primaquine was included in most Southeast Asia trials and the Zanzibar study), implemented multiple rounds of MDA, and had similar levels of coverage. Trials in Africa included substantially larger MDA target populations (> 10,000 in Zanzibar; > 37,000 in Zambia) compared to Southeast Asia (< 5000) and implemented additional co‐interventions (mainly, IRS) across arms. Although not measured directly, differences in underlying malaria epidemiology, population movement and risk of re‐introduction, and implementation of the intervention (more resource‐intensive in larger trials) could have contributed to variations in the short‐term effects by geography.

Of note, the certainty of evidence for outcomes in this review was based on a small number of included studies that, after stratification by endemicity and post‐MDA time period, often resulted in limited evidence. We assessed studies from very low‐ to low‐endemicity settings as high risk of bias for several criteria, and the certainty of evidence from studies from all endemicities was affected by imprecision. This highlights the need for additional studies with high‐quality evidence (both randomized and non‐randomized designs) and with sufficient sample sizes to account for the correlation due to clustering. Finally, we excluded many studies due to reasons such as imbalanced co‐interventions across arms or insufficient time points before and after MDA to conduct interrupted time series (ITS) analyses. Although these studies did not meet our criteria for inclusion, they may provide information useful for policymakers.

Quality of the evidence

In settings of moderate to high transmission, we judged the two trials as having low risk of bias (Figure 2), and summary of findings Table 1 describes our assessment of the certainty of evidence available. No pooled effect estimates were provided for outcomes in moderate‐ to high‐transmission settings due to a single study reporting after stratification by post‐MDA time period.

We judged most studies from very low‐ to low‐endemicity settings to be at high risk or unclear risk of bias for some criteria (Figure 2), and our assessment of the certainty of evidence available for important short‐term outcomes is provided in summary of findings Table 2, summary of findings Table 3, summary of findings Table 4, and summary of findings Table 5. Studies in this setting were generally underpowered and there were only a few malaria events across arms, resulting in imprecise estimates of effect. Many outcomes were downgraded by one or two levels due to imprecision after adjustment for clustering. All parasitaemia prevalence outcomes at longer‐term post‐MDA time periods were downgraded by one or two levels due to risk of bias, mainly due to baseline imbalance or biased sampling approaches for cross‐sectional parasitaemia surveys.

Potential biases in the review process

Although we sought to examine short‐ and longer‐term effects of MDA over a range of follow‐up time periods, there were few studies available within each follow‐up time period. Therefore, we were unable to examine whether variables, such as type of antimalarial drug, MDA coverage or number of rounds, and co‐interventions, modified the effect of MDA compared to no MDA.

Agreements and disagreements with other studies or reviews

Previous reviews on this topic described substantial but temporary reductions in malaria burden immediately following MDA (in both low‐ and high‐transmission settings) (Greenwood 2004; Newby 2015; Poirot 2013; von Seidlein 2003). Our review provides evidence of short‐term reductions in parasitaemia incidence in both transmission settings and evidence of short‐term reductions in parasitaemia prevalence in very low‐ to low‐endemicity areas, but not in moderate‐ to high‐endemicity areas. This difference is possibly due to the limited number of studies in moderate‐ to high‐transmission settings included in this review (two trials), different study design, or differences in malaria context between older and newer studies. Many older studies included in previous reviews were conducted in settings prior to scale‐up of vector control measures (e.g. ITNs), improvements in case management (i.e. RDTs and ACTs), and potentially less population movement to limit risk of re‐introduction. This may have resulted in a larger effect of MDA, which is supported by our results from non‐randomized studies that showed large short‐term reductions in parasitaemia prevalence in older studies conducted in settings of moderate to high endemicity. Additionally, it is possible that our inclusion of more rigorous study designs provided less biased comparisons for evaluating the effect of MDA.

The 2013 Cochrane Review on this topic highlighted a few studies on small islands or in highland settings in which malaria transmission was interrupted by MDA (Poirot 2013), but these studies did not meet inclusion criteria for our updated review. Although MDA has been attributed to successful elimination of malaria in Vanuatu (Kaneko 2000 VUT), we excluded this study due to an imbalance in co‐interventions, since MDA was administered in combination with ITNs and larvivorous fish, while the no‐MDA control arm received delayed ITNs and no additional interventions. Additionally, more recent studies from island settings did not meet our inclusion criteria: ITNs were introduced concurrently with MDA in a recent study conducted in Grande Comore, Comoros (Affane 2012 COM), and there were insufficient data points collected to adequately account for trends in malaria seasonality using the ITS design in a study conducted on Anjouan island, Comoros (Deng 2018 COM).

Finally, the previous Cochrane Review also highlighted a paucity of data to assess whether a reduction by MDA was sustained (Poirot 2013). Our findings address some of the knowledge gaps from the previous review with the availability of longer‐term follow‐up data post‐MDA for parasitaemia prevalence, and confirmed malaria illness incidence from several studies in very low‐ to low‐transmission settings.

Study flow diagram

Figures and Tables -
Figure 1

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 study

Figures and Tables -
Figure 2

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

Navigate to figure in ReviewOpen in new tab

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies

Figures and Tables -
Figure 3

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies

Navigate to figure in ReviewOpen in new tab

Comparison 1: MDA versus no MDA in moderate to high endemicity (cRCTs) on P falciparum outcomes, Outcome 1: Parasitaemia prevalence (P falciparum)

Figures and Tables -
Analysis 1.1

Comparison 1: MDA versus no MDA in moderate to high endemicity (cRCTs) on P falciparum outcomes, Outcome 1: Parasitaemia prevalence (P falciparum)

Navigate to figure in ReviewOpen in new tab

Comparison 1: MDA versus no MDA in moderate to high endemicity (cRCTs) on P falciparum outcomes, Outcome 2: Parasitaemia incidence (P falciparum)

Figures and Tables -
Analysis 1.2

Comparison 1: MDA versus no MDA in moderate to high endemicity (cRCTs) on P falciparum outcomes, Outcome 2: Parasitaemia incidence (P falciparum)

Navigate to figure in ReviewOpen in new tab

Comparison 1: MDA versus no MDA in moderate to high endemicity (cRCTs) on P falciparum outcomes, Outcome 3: Confirmed malaria illness incidence (P falciparum)

Figures and Tables -
Analysis 1.3

Comparison 1: MDA versus no MDA in moderate to high endemicity (cRCTs) on P falciparum outcomes, Outcome 3: Confirmed malaria illness incidence (P falciparum)

Navigate to figure in ReviewOpen in new tab

Comparison 1: MDA versus no MDA in moderate to high endemicity (cRCTs) on P falciparum outcomes, Outcome 4: Gametocytaemia prevalence (P falciparum)

Figures and Tables -
Analysis 1.4

Comparison 1: MDA versus no MDA in moderate to high endemicity (cRCTs) on P falciparum outcomes, Outcome 4: Gametocytaemia prevalence (P falciparum)

Navigate to figure in ReviewOpen in new tab

Comparison 1: MDA versus no MDA in moderate to high endemicity (cRCTs) on P falciparum outcomes, Outcome 5: Malaria‐specific mortality

Figures and Tables -
Analysis 1.5

Comparison 1: MDA versus no MDA in moderate to high endemicity (cRCTs) on P falciparum outcomes, Outcome 5: Malaria‐specific mortality

Navigate to figure in ReviewOpen in new tab

Comparison 2: MDA versus no MDA in very low to low endemicity (cRCTs) on P falciparum outcomes, Outcome 1: Parasitaemia prevalence (P falciparum)

Figures and Tables -
Analysis 2.1

Comparison 2: MDA versus no MDA in very low to low endemicity (cRCTs) on P falciparum outcomes, Outcome 1: Parasitaemia prevalence (P falciparum)

Navigate to figure in ReviewOpen in new tab

Comparison 2: MDA versus no MDA in very low to low endemicity (cRCTs) on P falciparum outcomes, Outcome 2: Parasitaemia incidence (P falciparum)

Figures and Tables -
Analysis 2.2

Comparison 2: MDA versus no MDA in very low to low endemicity (cRCTs) on P falciparum outcomes, Outcome 2: Parasitaemia incidence (P falciparum)

Navigate to figure in ReviewOpen in new tab

Comparison 2: MDA versus no MDA in very low to low endemicity (cRCTs) on P falciparum outcomes, Outcome 3: Confirmed malaria illness incidence (P falciparum)

Figures and Tables -
Analysis 2.3

Comparison 2: MDA versus no MDA in very low to low endemicity (cRCTs) on P falciparum outcomes, Outcome 3: Confirmed malaria illness incidence (P falciparum)

Navigate to figure in ReviewOpen in new tab

Comparison 3: MDA versus no MDA in very low to low endemicity (cRCTs) on P vivax outcomes, Outcome 1: Parasitaemia prevalence (P vivax)

Figures and Tables -
Analysis 3.1

Comparison 3: MDA versus no MDA in very low to low endemicity (cRCTs) on P vivax outcomes, Outcome 1: Parasitaemia prevalence (P vivax)

Navigate to figure in ReviewOpen in new tab

Comparison 3: MDA versus no MDA in very low to low endemicity (cRCTs) on P vivax outcomes, Outcome 2: Confirmed malaria illness incidence (P vivax)

Figures and Tables -
Analysis 3.2

Comparison 3: MDA versus no MDA in very low to low endemicity (cRCTs) on P vivax outcomes, Outcome 2: Confirmed malaria illness incidence (P vivax)

Navigate to figure in ReviewOpen in new tab

Comparison 4: Supplemental analysis: post‐hoc subgroup analysis by continent, Outcome 1: Plasmodium falciparum parasitaemia prevalence post‐MDA 1‐3 months

Figures and Tables -
Analysis 4.1

Comparison 4: Supplemental analysis: post‐hoc subgroup analysis by continent, Outcome 1: Plasmodium falciparum parasitaemia prevalence post‐MDA 1‐3 months

Navigate to figure in ReviewOpen in new tab
Summary of findings 1. MDA compared to no MDA for Plasmodium falciparum malaria (moderate to high endemicity, short‐term follow‐up)

Patient or population: People of all ages living in an area with moderate to high endemicity of P falciparum malaria (≥ 10% prevalence)
Setting: Moderate to high endemicity defined as ≥10% prevalence of P falciparum
Intervention: MDA
Comparison: Control (no MDA or placebo)

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with no MDA

Risk with MDA

Follow‐up: 1 to 3 months

Parasitaemia prevalence

5 per 100

9 per 100
(3 to 27)

RR 1.76
(0.58 to 5.36)

786
(1 RCT)

⊕⊕⊝⊝
LOWa,b,c

Due to imprecision

At 1‐3 months post‐MDA, parasite prevalence may increase in MDA compared no MDA. However, the effects vary and it is possible that MDA makes little or no difference on parasitaemia prevalence.

Parasitaemia incidence

68 events per 100 person‐years

42 events per 100 person‐years
(27 to 63)

Rate ratio 0.61
(0.40 to 0.92)

739
(1 RCT)

⊕⊕⊕⊝
MODERATEa,b,d

Due to imprecision

At 1‐3 months post‐MDA, there is probably a reduction in parasitaemia incidence in MDA compared to no MDA.

Confirmed malaria illness incidence

28 per 1000 population

11 per 1000 population
(1 to 122)

Rate ratio 0.41
(0.04 to 4.42)

144,422
(1 RCT)

⊕⊕⊝⊝
LOWa,b,c

Due to imprecision

At 1‐3 months post‐MDA, there may be a reduction in confirmed malaria illness incidence in MDA compared to no MDA.

Follow‐up: 4 to 6 months

Parasitaemia prevalence

55 per 100

65 per 100
(49 to 86)

RR 1.18
(0.89 to 1.56)

1414
(1 RCT)

⊕⊕⊕⊝
MODERATEa,b,d

Due to imprecision

At 4‐6 months post‐MDA, there is probably little or no effect on parasitaemia prevalence in MDA compared to no MDA

Parasitaemia incidence

129 events per 100 person‐years

118 events per 100 person‐years
(71 to 194)

Rate ratio 0.91
(0.55 to 1.50)

1376
(1 RCT)

⊕⊝⊝⊝
VERY LOWa,b,d,e

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia incidence at 4‐6 months post‐MDA

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: confidence interval; cRCT: cluster‐randomized controlled trial; MDA: mass drug administration; RCT: randomized controlled trial; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect

aNot downgraded for inconsistency; the comparison presented is reported from a single study.
bNot downgraded for indirectness; outcome was evaluated in all ages or assessed in children (considered the most appropriate population to measure malaria transmission in moderate‐ to high‐endemicity areas).
cDowngraded 2 levels for imprecision due to very wide CIs.
dDowngraded 1 level for imprecision due to wide CIs.
eDowngraded 2 levels for risk of bias since malaria cases in outcome were defined as fever plus parasitaemia > 5000, which excludes all afebrile and low density infections and results in an underestimate of the outcome.

Figures and Tables -
Summary of findings 1. MDA compared to no MDA for Plasmodium falciparum malaria (moderate to high endemicity, short‐term follow‐up)
Navigate to table in Review
Summary of findings 2. MDA compared to no MDA for Plasmodium falciparum malaria (very low to low endemicity, short‐term follow‐up)

Patient or population: People of all ages living in an area with very low to low endemicity of P falciparum malaria (< 10% prevalence)
Setting: Very low to low endemicity defined as < 10% prevalence of P falciparum
Intervention: MDA
Comparison: Control (no MDA or placebo)

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with Control

Risk with MDA

Follow‐up: < 1 month

Parasitaemia prevalence
follow up: range < 1 month

12 per 100

1 per 100
(0 to 6)

RR 0.12
(0.03 to 0.52)

1232
(1 RCT)

⊕⊕⊝⊝
LOWa‐d

Due to risk of bias and imprecision

At < 1 month post‐MDA, there may a reduction in parasitaemia prevalence in MDA compared to no MDA.

Follow‐up: 1 to 3 months

Parasitaemia prevalence

3 per 100

1 per 100

(0 to 1)

RR: 0.25

(0.15 to 0.41)

17,454
(7 RCTs)5

⊕⊕⊝⊝
LOWe‐i

Due to risk of bias

At 1‐3 months post‐MDA, there may a reduction in parasitaemia prevalence in MDA compared to no MDA.

Parasitaemia incidence

15 events per 100 person‐years

5 events per 100 person‐years
(3 to 10)

Rate ratio 0.37
(0.21 to 0.66)

736
(1 RCT)

⊕⊕⊕⊝
MODERATEb,d,,j

Due to imprecision

At 1‐3 months post‐MDA, there is probably a reduction in parasitaemia incidence in MDA compared to no MDA.

Confirmed malaria illness incidence

6 per 1000 population

4 per 1000 population
(1 to 17)

Rate ratio: 0.58

(0.12 to 2.73)

130,651
(2 RCTs)

⊕⊝⊝⊝
VERY LOWa,j‐l

Due to risk of bias and imprecision

We do not know if MDA has an effect on confirmed malaria illness incidence at 1‐3 months post‐MDA compared to no MDA.

Follow‐up: 4 to 6 months

Parasitaemia prevalence

5 per 100

3 per 100

(2 to 6)

RR: 0.63

(0.36 to 1.12)

5670

(4 RCTs)

⊕⊝⊝⊝
VERY LOWc,d,f,l

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia prevalence at 4‐6 months post‐MDA compared to no MDA.

 

Confirmed malaria illness incidence

4 per 1000 population

4 per 1000 population
(0 to 53)

Rate ratio 0.93
(0.07 to 12.43)

23,251
(1 RCT)

⊕⊝⊝⊝
VERY LOWa,b,k,l

Due to risk of bias and imprecision

We do not know if MDA has an effect on confirmed malaria illness incidence at 4‐6 months post‐MDA compared to no MDA.

Follow‐up: 7 to 12 months

Parasitaemia prevalence

5 per 100

4 per 100

(3 to 6)

RR: 0.86

(0.55 to 1.36)

7760

(5 RCTs)

⊕⊝⊝⊝
VERY LOWc,d,l,m

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia prevalence at 7‐12 months post‐MDA compared to no MDA.

Confirmed malaria illness incidence

11 per 1000 population

5 per 1000 population (2 to 12)

Rate ratio 0.47

(0.21 to 1.03)

26,576
(3 RCTs)

⊕⊝⊝⊝
VERY LOWd,f,j,l

Due to risk of bias and imprecision

We do not know if MDA has an effect on confirmed malaria illness incidence at 7‐12 months post‐MDA compared to no MDA.

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: confidence interval; cRCT: cluster‐randomized controlled trial; MDA: mass drug administration; RCT: randomized controlled trial; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect

aDowngraded 1 level for risk of bias due to several criteria scored as high or unclear risk of bias.
bNot downgraded for inconsistency; the comparison presented is reported from a single study.
cNot downgraded for indirectness; all ages are at similar risk of malaria transmission, given the local epidemiology of malaria in this study setting and the outcomes were assessed by a highly sensitive diagnostic method (ultrasensitive PCR).
dDowngraded 1 level for imprecision due to wide CIs.
eEight included studies reported parasitaemia prevalence during the 1‐3 month post‐MDA follow‐up period; however, one study did not contribute data in the meta‐analysis due to no events at baseline before MDA or at any follow‐up time points.
fDowngraded 2 levels for risk of bias due to several criteria scored as high or unclear risk of bias, including baseline imbalance and high risk of contamination in several studies.
gNot downgraded for inconsistency based on I2 statistic; however, reasons for heterogeneity explored in post‐hoc sub‐group analysis by continent (sub‐Saharan Africa and Southeast Asia; Analysis 4.1).
hNot downgraded for indirectness; all ages are at similar risk of malaria transmission, given the local epidemiology of malaria in this study setting so there is no concern with assessing this outcome in different age groups across studies.
iNot downgraded for imprecision due to appreciable benefit of pooled effect as reported by seven studies.
jNot downgraded for indirectness since outcome was assessed in all ages and by routine detection methods.
kDowngraded 2 levels for imprecision due to very wide CIs.
lNot downgraded for inconsistency based on I2 statistic.
mDowngraded 2 levels for risk of bias due to several criteria scored as high or unclear risk of bias, including baseline imbalance, high risk of contamination, and a large unexplained increase in sampled population in the MDA group at this time point.
 

Figures and Tables -
Summary of findings 2. MDA compared to no MDA for Plasmodium falciparum malaria (very low to low endemicity, short‐term follow‐up)
Navigate to table in Review
Summary of findings 3. MDA compared to no MDA for Plasmodium falciparum malaria (very low to low endemicity, long‐term follow‐up)

Patient or population: People of all ages living in an area with very low to low endemicity of P falciparum malaria ( < 10% prevalence)
Setting: Very low to low endemicity defined as < 10% prevalence of P falciparum
Intervention: MDA
Comparison: Control (no MDA or placebo)

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with Control

Risk with MDA

Follow‐up: 13 to 18 months

Parasitaemia prevalence

4 per 100

4 per 100
(1 to 14)

RR 0.82
(0.20 to 3.34)

1537
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐d

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia prevalence at 13‐18 months post‐MDA compared to no MDA.

Confirmed malaria illness incidence

17 per 1000 population

13 per 1000 population
(3 to 51)

Rate ratio 0.77
(0.20 to 3.03)

23,251
(1 RCT)

⊕⊝⊝⊝
VERY LOWb,d‐f

Due to risk of bias and imprecision

We do not know if MDA has an effect on confirmed malaria illness incidence at 13‐18 months post‐MDA compared to no MDA.

Follow‐up: 19 to 24 months

Parasitaemia prevalence
follow‐up: range 19 to 24 months

3 per 100

1 per 100
(0 to 6)

RR 0.34
(0.06 to 1.97)

1393
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐d

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia prevalence at 19‐24 months post‐MDA compared to no MDA.

Follow‐up: 25 months and above

Parasitaemia prevalence
follow‐up: range 25 to 30 months

3 per 100

3 per 100
(1 to 12)

RR 0.89
(0.22 to 3.62)

1521
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐d

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia prevalence at 25‐30 months post‐MDA compared to no MDA.

Parasitaemia prevalence
follow‐up: range 31 to 36 months

3 per 100

4 per 100
(1 to 19)

RR 1.25
(0.25 to 6.31)

1679
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐d

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia prevalence at 31‐36 months post‐MDA compared to no MDA.

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; cRCT: cluster‐randomized controlled trial; MDA: mass drug administration; RCT: randomized controlled trial; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect

aDowngraded 2 levels for risk of bias due to several criteria scored as high or unclear risk of bias, including baseline imbalance, high risk of contamination, and a large unexplained increase in sampled population in the MDA group at this time point.
bNot downgraded for inconsistency; the comparison presented is reported from a single study.
cNot downgraded for indirectness; all ages are at similar risk of malaria transmission, given the local epidemiology of malaria in this study setting and the outcomes were assessed by a highly sensitive diagnostic method (ultrasensitive PCR).
dDowngraded 2 levels for imprecision due to very wide CIs.
eDowngraded 1 level for risk of bias due to several criteria scored as high or unclear risk of bias.
fNot downgraded for indirectness since outcome was assessed in all ages and by routine detection methods.

Figures and Tables -
Summary of findings 3. MDA compared to no MDA for Plasmodium falciparum malaria (very low to low endemicity, long‐term follow‐up)
Navigate to table in Review
Summary of findings 4. MDA compared to no MDA for P vivax malaria (very low to low endemicity, short‐term follow‐up)

Patient or population: People of all ages living in an area with very low to low endemicity of P vivax malaria (< 10% prevalence)
Setting: Very low to low endemicity defined as < 10% prevalence of P vivax
Intervention: MDA
Comparison: Control (no MDA or placebo)

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with Control

Risk with MDA

Follow‐up: < 1 month

Parasitaemia prevalence
follow up: range < 1 month

27 per 100

5 per 100
(2 to 11)

RR 0.18
(0.08 to 0.40)

1232
(1 RCT)

⊕⊕⊝⊝
LOWa‐d

Due to risk of bias and imprecision

At < 1 month post‐MDA, there may a reduction in parasitaemia prevalence in MDA compared to no MDA.

Follow‐up: 1 to 3 months

Parasitaemia prevalence

12 per 100

2 per 100

(1 to 3)

RR: 0.15

(0.10 to 0.24)

6896

(5 RCTs)

⊕⊕⊝⊝
LOWc,e‐g

Due to risk of bias

At 1‐3 months post‐MDA, there may a reduction in parasitaemia prevalence in MDA compared to no MDA.

Follow‐up: 4 to 6 months

Parasitaemia prevalence

11 per 100

9 per 100

(7 to 10)

RR: 0.78

(0.63 to 0.95)

 

5670

(4 RCTs)

⊕⊝⊝⊝
VERY LOWc,d,f,h

Due to risk of bias and imprecision

We do not know if MDA reduces parasitaemia prevalence at 4‐6 months post‐MDA compared to no MDA.

Follow‐up: 7 to 12 months

Parasitaemia prevalence

9 per 100

11 per 100

(9 to 13)

RR: 1.12

(0.94 to 1.34)

7760

(5 RCTs)

⊕⊝⊝⊝
VERY LOWc,d,h,i

Due to risk of bias and imprecision

We do not know if MDA has an effect on parasitaemia prevalence at 7‐12 months post‐MDA compared to no MDA.

Confirmed malaria illness incidence

41 per 1000 population

57 per 1000 population (40 to 80)

Rate ratio: 1.38

(0.97 to 1.95)

3325

(2 RCTs)

⊕⊝⊝⊝
VERY LOWd,f,h,j

Due to risk of bias and imprecision

We do not know if MDA has an effect on confirmed malaria illness incidence at 7‐12 months post‐MDA compared to no MDA.

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; cRCT: cluster‐randomized controlled trial; MDA: mass drug administration; RCT: randomized controlled trial; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect

aDowngraded 1 level for risk of bias due to several criteria scored as high or unclear risk of bias.
bNot downgraded for inconsistency; the comparison presented is reported from a single study.
cNot downgraded for indirectness; all ages are at similar risk of malaria transmission, given the local epidemiology of malaria in this study setting and the outcomes were assessed by a highly sensitive diagnostic method (ultrasensitive PCR).
dDowngraded 1 level for imprecision due to wide CIs.
eNot downgraded for imprecision due to appreciable benefit of pooled effect as reported by five studies.
fDowngraded 2 levels for risk of bias due to several criteria scored as high or unclear risk of bias, including baseline imbalance and high risk of contamination.
gNot downgraded for inconsistency despite the large value of the I2 statistic since the direction of effect was consistent with large imprecision.
hNot downgraded for inconsistency based on I2 statistic.
iDowngraded 2 levels for risk of bias due to several criteria scored as high or unclear risk of bias, including a large unexplained increase in sampled population in the MDA group at this time point.
jNot downgraded for indirectness since outcome was assessed in all ages and by routine detection methods.

Figures and Tables -
Summary of findings 4. MDA compared to no MDA for P vivax malaria (very low to low endemicity, short‐term follow‐up)
Navigate to table in Review
Summary of findings 5. MDA compared to no MDA for P vivax malaria (very low to low endemicity, long‐term follow‐up)

Patient or population: People of all ages living in an area with very low to low endemicity of P vivax malaria (< 10% prevalence)
Setting: Very low to low endemicity < 10% prevalence of P vivax
Intervention: MDA
Comparison: Control (no MDA or placebo)

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with Control

Risk with MDA

Follow‐up: 13 to 18 months

Parasitaemia prevalence

17 per 100

14 per 100
(8 to 25)

RR 0.81
(0.44 to 1.48)

1537
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐d

Due to risk of bias and imprecision

We do not know if MDA reduces parasitaemia prevalence at 13‐18 months post‐MDA compared to no MDA.

Follow‐up: 19 to 24 months

Parasitaemia prevalence
follow‐up: range 19 to 24 months

11 per 100

9 per 100
(4 to 20)

RR 0.84
(0.38 to 1.83)

1393
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐d

Due to risk of bias and imprecision

We do not know if MDA reduces parasitaemia prevalence at 19‐24 months post‐MDA compared to no MDA.

Follow‐up: 25 months and above

Parasitaemia prevalence
follow‐up: range 25 to 30 months

11 per 100

9 per 100
(4 to 21)

RR 0.89
(0.41 to 1.94)

1521
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐d

Due to risk of bias and imprecision

We do not know if MDA reduces parasitaemia prevalence at 25‐30 months post‐MDA compared to no MDA.

Parasitaemia prevalence
follow‐up: range 31 to 36 months

6 per 100

7 per 100
(3 to 20)

RR 1.20
(0.44 to 3.29)

1679
(1 RCT)

⊕⊝⊝⊝
VERY LOWa‐c,e

Due to risk of bias and imprecision

We do not know if MDA reduces parasitaemia prevalence at 31‐36 months post‐MDA compared to no MDA.

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: confidence interval; cRCT: cluster‐randomized controlled trial; MDA: mass drug administration; RCT: randomized controlled trial; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect

aDowngraded 2 levels for risk of bias due to several criteria scored as high or unclear risk of bias, including a large unexplained increase in sampled population in the MDA group at this time point.
bNot downgraded for inconsistency; the comparison presented is reported from a single study.
cNot downgraded for indirectness; all ages are at similar risk of malaria transmission, given the local epidemiology of malaria in this study setting and the outcomes were assessed by a highly sensitive diagnostic method (ultrasensitive PCR).
dDowngraded 1 level for imprecision due to wide CIs.
eDowngraded 2 levels for imprecision due to very wide CIs.

Figures and Tables -
Summary of findings 5. MDA compared to no MDA for P vivax malaria (very low to low endemicity, long‐term follow‐up)
Navigate to table in Review
Table 1. Description of studies

Study ID (Design)

Year(s) of study

Malaria endemicitya

Plasmodium species

Antimalarial drug resistance

MDA group

Control group

Co‐intervention(s)b

Outcomes reported (months of follow‐up post‐MDAc)

Drug

Rounds, interval, and duration implemented

Population targeted (coverage)

Eisele 2020 ZMBa (cRCT)

2014‐2017

Low

P falciparum

Widespread resistance to CQ and SP, but no evidence of resistance to artemisinin

DHAp

4 rounds administered at start of rainy season, during rainy season, during dry season, and at start of rainy season over 15 months

37,694

(79% in round 1; 63% in round 2; 76% in round 3; 66% in round 4)

No drug and no placebo

IRS, ITNs, and enhanced standard of care

  • Parasitaemia prevalence (2)

  • Parasitaemia incidence (2)

  • Confirmed malaria case incidence (2)

  • Adverse effects

Eisele 2020 ZMBb (cRCT)

2014‐2017

High

P falciparum

Widespread resistance to CQ and SP, but no evidence of resistance to artemisinin

DHAp

4 rounds administered at start of rainy season, during rainy season, during dry season, and at start of rainy season over 15 months

45,442

(79% in round 1; 63% in round 2; 76% in round 3; 66% in round 4)

No drug and no placebo

IRS, ITNs, and enhanced standard of care

  • Parasitaemia prevalence (2)

  • Parasitaemia incidence (2)

  • Confirmed malaria case incidence (2)

  • Adverse effects

Escudie 1962 BFA (CBA)

1960‐1961

High

P falciparum,

P ovale,

P malariae

ND

AQ‐PQ or CQ‐PQ

(Low frequency MDA)

7 rounds administered 28 days apart over 7 months

1890

(75% to 91% per round)

No drug and no placebo

None (IRS arms excluded)

  • Parasitaemia prevalence (3)

  • Gametocytaemia prevalence (3)

(High frequency MDA)

15 rounds administered 14 days apart over 7 months

2560

(84% to 97% per round)

  • Parasitaemia prevalence (3)

  • Gametocytaemia prevalence (3)

Landier 2017 MMRa (cRCT)

2013‐2015

Low

P falciparum,

P vivax

Artemisinin resistance firmly established

DHAp with PQ

3 rounds administered 1 month apart over 3 months

1434 (66% in round 1, 56% in round 2, and 65% in round 3)

Delayed MDA

ITNs, uninterrupted access to case management

  • Parasitaemia prevalence (7)

  • Confirmed malaria illness incidence (7)

  • Adverse effects

McLean 2021 MMR (cRCT)

2014‐2017

Very low

P falciparum,

P vivax

Artemisinin resistance: Kelch 13 mutation in 57% of samples at baseline

DHAp with PQ

3 rounds administered 1 month apart over 3 months

4622

(86% in round 1, 86% in round 2, 88% in round 3)

No drug and no placebo

ITNs, routine malaria control by village health workers

  • Parasitaemia prevalence (31)

  • Adverse effects

Molineaux 1980 NGA (CBA)

1970‐1975

High

P falciparum,

P malariae,

P ovale

ND

SP

(Low frequency MDA)

9 rounds administered 10 weeks apart over 18 months

14,129

(73% to 92% per round)

No drug and no placebo

IRS

  • Parasitaemia prevalence (0)

  • Gametocytaemia prevalence (0)

(High frequency MDA)

23 rounds administered 2 weeks apart during the wet seasons and 10 weeks apart during the dry seasons over 18 months

1810

(72% to 91% per round)

Morris 2018 TZA (cRCT)

2016‐2017

Very low

P falciparum,

P malariae,

P ovale, and

P vivax

No evidence of resistance to first line treatment AS‐AQ

DHAp with PQ

2 rounds administered 4 weeks apart over 6 weeks

10,944

(91% in round 1, 88% in round 2)

No drug and no placebo

IRS and ITNs

  • Parasitaemia prevalence (0)

  • Confirmed malaria illness incidence (14)

  • Adverse effects

Pongvongsa 2018 LAO (cRCT)

2016‐2017

Low

P falciparum,

P vivax

ND

DHAp with PQ

3 rounds administered 1 month apart over 3 months

1006 (81% in round 1, 80% in round 2, and 82% in round 3)

Delayed MDA

ITNs, uninterrupted access to case management

  • Parasitaemia prevalence (10)

  • Adverse effects

Roberts 1964 KEN (CBA)

1953‐1954

Moderate

P falciparum

ND

Pyrimethamine

2 rounds administered 1 year apart over 13 months

101,000

(95% in round 1, 93% in round 2)

No drug and no placebo

None

  • Parasitaemia prevalence (7)

Shekalaghe 2011 TZA (cRCT)

2008

Very low

P falciparum

ND

SP+AS with PQ

1 round over 16 days

1110

(95%)

Placebo

ITNs, single treatment campaign for trachoma with azithromycin

  • Parasitaemia prevalence (4)

  • Confirmed malaria illness incidence (4)

  • Gametocytaemia prevalence (4)

  • Adverse effects

Tripura 2018 KHM (cRCT)

2014‐2016

Very low

P falciparum,

P vivax

Reduced susceptibility to artemisinins and ACT partner drug resistance

DHAp

3 rounds administered 1 month apart over 3 months

858 (74% in round 1, 60% in round 2, and 71% in round 3)

Delayed MDA

ITNs, uninterrupted access to case management

  • Parasitaemia prevalence (10)

  • Confirmed malaria case incidence (9)

  • Adverse effects

von Seidlein 2003 GMB (cRCT)

1999

High

P falciparum

ND

SP+AS

1 round over 1 month

12,331

(89%)

Placebo

None

  • Parasitaemia prevalence (5)

  • Parasitaemia incidence (5)

  • Gametocytaemia prevalence (5)

  • Malaria‐specific mortality

  • Adverse effects

von Seidlein 2019 VNM (cRCT)

2013‐2015

Very low

P falciparum,

P vivax

No evidence of resistance to DHAp at the start of study, but treatment failure to DHAp has increased following study

DHAp with PQ

3 rounds administered 1 month apart over 3 months

1439 (83% in round 1, 98% in round 2, and 99% in round 3)

Delayed MDA

ITNs, uninterrupted access to case management

  • Parasitaemia prevalence (10)

  • Adverse effects

ACT = artemisinin‐based combination therapy, AQ = amodiaquine, AS = artesunate, CBA = controlled before‐and‐after study, CQ = chloroquine, cRCT = cluster‐randomized controlled trial, DHAp = dihydroartemisinin piperaquine, ITNs = insecticide‐treated bed nets, IRS = indoor residual spraying, MDA = mass drug administration, PQ = primaquine, SP = sulfadoxine‐ (or sulfalene‐) pyrimethamine, NA = not applicable, ND = not described.

aMalaria endemicity classified as very low (> 0% to < 1%), low (1% to < 10%), moderate (10% to < 35%) or high (≥ 35%) (WHO 2017).
bCo‐interventions were balanced across intervention and control groups, as per inclusion criteria.
cPost‐MDA refers to the length of time, in months, after the last round of MDA that the outcome was evaluated.

Figures and Tables -
Table 1. Description of studies
Navigate to table in Review
Table 2. Description of outcomes

Study ID (design)

Parasitaemia prevalence

Parasitaemia incidence

Confirmed malaria illness incidence

All‐cause or malaria‐specific mortality

Gametocytaemia prevalence

Adverse effects

Eisele 2020 ZMBa (cRCT)

Yes

Yes

Yes

No

No

Yes

Eisele 2020 ZMBb (cRCT)

Yes

Yes

Yes

No

No

Yes

Escudie 1962 BFA (CBA)

Yes

No

No

No

Yes

No

Landier 2017 MMRa (cRCT)

Yes

No

Yes

No

No

Yes

McLean 2021 MMR (cRCT)

Yes

No

No

No

No

Yes

Molineaux 1980 NGA (CBA)

Yes

No

No

No

Yes

No

Morris 2018 TZA (cRCT)

Yes

No

Yes

No

No

Yes

Pongvongsa 2018 LAO (cRCT)

Yes

No

No

No

No

Yes

Roberts 1964 KEN (CBA)

Yes

No

No

No

No

No

Shekalaghe 2011 TZA (cRCT)

Yes

No

Yes

No

Yes

Yes

Tripura 2018 KHM (cRCT)

Yes

No

Yes

No

No

Yes

von Seidlein 2003 GMB (cRCT)

Yes

Yes

Yes

Yes

Yes

Yes

von Seidlein 2019 VNM (cRCT)

Yes

No

No

No

No

Yes

CBA = controlled before‐and‐after study, cRCT = cluster‐randomized controlled trial
Figures and Tables -
Table 2. Description of outcomes
Navigate to table in Review
Table 3. Difference‐in‐differences analysis of P falciparum parasitaemia prevalence in non‐randomized studies

Study

Intervention, % (n)

Control % (n)

Difference‐in‐differences, percentage pointsa

Pre‐MDA

During MDA

Post‐MDA

Pre‐MDA

During MDA

Post‐MDA

During MDA

Post‐MDA

1 to 3 months

4 to 6 months

7 to 12 months

1 to 3 months

4 to 6 months

7 to 12 months

1 to 3 months

4 to 6 months

7 to 12 months

Escudie 1962 BFAb

Low frequency MDA with AQ‐PQ or CQ‐PQ

67.6 (190)

21.6 (75)

38.3 (105)

ND

ND

59.4 (129)

74.8 (517)

386 (72.3)

ND

ND

‐61.4

‐42.1

ND

ND

High frequency MDA with AQ‐PQ or CQ‐PQ

33.6 (131)

12.6 (59)

61.4 (286)

ND

ND

59.4 (129)

74.8 (517)

386 (72.3)

ND

ND

‐36.3

14.9

ND

ND

Molineaux 1980 NGAc

Low frequency MDA with SP

41.8 (525)

1.9 (40)

ND

ND

ND

49.1 (493)

32.5 (380)

ND

ND

ND

‐23.2

ND

ND

ND

High frequency MDA with SP

44.9 (754)

7.3 (109)

ND

ND

ND

49.1 (493)

32.5 (380)

ND

ND

ND

‐20.9

ND

ND

ND

Roberts 1964 KEN

MDA with pyrimethamine

8.3 (25)

9 (188)

2.9 (26)

(27) (4.5)

(15) (5)

18 (154)

34.4 (723)

40.7 (366)

37 (222)

26 (78)

‐15.8

‐28.1

‐22.8

‐11.3

AQ = amodiaquine, CQ = chloroquine, MDA = mass drug administration, ND = no data, PQ = primaquine, SP = sulfalene‐pyrimethamine

aCalculated as difference in proportion at the time period of during MDA or post‐MDA minus the proportion at pre‐MDA in the intervention and control separately and the difference in these two proportion differences between the intervention and control groups.
bMDA with AQ‐PQ or CQ‐PQ either every 4 weeks ('low frequency MDA') or every 2 weeks ('high frequency MDA').
cMDA with sulfalene‐pyrimethamine either every 10 weeks (low frequency MDA') or every 2 weeks during the wet season and every 10 weeks during the dry season ('high frequency MDA') to all ages except infants prior to their first malaria episode.

Figures and Tables -
Table 3. Difference‐in‐differences analysis of P falciparum parasitaemia prevalence in non‐randomized studies
Navigate to table in Review
Table 4. Difference‐in‐differences analysis of P falciparum gametocytaemia prevalence in non‐randomized studies

Study

Intervention, % (n)

Control, % (n)

Difference‐in‐differences

percentage pointa

Pre‐MDA

During MDA

Post‐MDA

1 to 3 months

Pre‐MDA

During MDA

Post‐MDA

1 to 3 months

During MDA

Post‐MDA

1 to 3 months

Escudie 1962 BFAb

Low frequency MDA with AQ‐PQ or CQ‐PQ

20.3 (57)

0.9 (3)

38.3 (35)

19.4 (42)

14 (97)

19.1 (102)

‐14.1

18.3

High frequency MDA with AQ‐PQ or CQ‐PQ

8.2 (32)

1.9 (9)

61.4 (107)

19.4 (42)

14 (97)

19.1 (102)

‐1.0

53.4

Molineaux 1980 NGAc

Low frequency MDA with SP

10.1 (127)

0.6 (12)

ND

12.4 (124)

7.9 (92)

ND

‐5.0

ND

High frequency MDA with SP

12.4 (208)

3.2 (48)

ND

12.4 (124)

7.9 (92)

ND

‐4.7

ND

AQ = amodiaquine, CQ = chloroquine, MDA = mass drug administration, ND = no data, SP = sulfalene‐pyrimethamine

aCalculated as difference in proportion at the time period of during MDA or post‐MDA minus the proportion at pre‐MDA in the intervention and control separately and the difference in these two proportion differences between the intervention and control groups.
bMDA with AQ‐PQ or CQ‐PQ either every 4 weeks ('low frequency MDA') or every 2 weeks ('high frequency MDA').
cMDA with sulfalene‐pyrimethamine either every 10 weeks ('low frequency MDA') or every 2 weeks during the wet season and every 10 weeks during the dry season ('high frequency MDA') to all ages except infants prior to their first malaria episode.

Figures and Tables -
Table 4. Difference‐in‐differences analysis of P falciparum gametocytaemia prevalence in non‐randomized studies
Navigate to table in Review
Comparison 1. MDA versus no MDA in moderate to high endemicity (cRCTs) on P falciparum outcomes

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1.1 Parasitaemia prevalence (P falciparum) Show forest plot

2

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

1.1.1 Baseline before MDA

2

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

1.1.2 During MDA

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

1.1.3 Post‐MDA 1‐3 months

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

1.1.4 Post‐MDA 4‐6 months

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

1.2 Parasitaemia incidence (P falciparum) Show forest plot

2

Rate Ratio (IV, Fixed, 95% CI)

Totals not selected

1.2.1 Post‐MDA 1‐3 months

1

Rate Ratio (IV, Fixed, 95% CI)

Totals not selected

1.2.2 Post‐MDA 4‐6 months

1

Rate Ratio (IV, Fixed, 95% CI)

Totals not selected

1.3 Confirmed malaria illness incidence (P falciparum) Show forest plot

1

Rate Ratio (IV, Fixed, 95% CI)

Totals not selected

1.3.1 Baseline before MDA

1

Rate Ratio (IV, Fixed, 95% CI)

Totals not selected

1.3.2 Post‐MDA 1‐3 months

1

Rate Ratio (IV, Fixed, 95% CI)

Totals not selected

1.4 Gametocytaemia prevalence (P falciparum) Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

1.4.1 Baseline before MDA

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

1.4.2 Post‐MDA 4‐6 months

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

1.5 Malaria‐specific mortality Show forest plot

1

Risk Ratio (IV, Fixed, 95% CI)

Totals not selected

1.5.1 Post‐MDA 4‐6 months

1

Risk Ratio (IV, Fixed, 95% CI)

Totals not selected
Figures and Tables -
Comparison 1. MDA versus no MDA in moderate to high endemicity (cRCTs) on P falciparum outcomes
Navigate to table in Review
Comparison 2. MDA versus no MDA in very low to low endemicity (cRCTs) on P falciparum outcomes

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

2.1 Parasitaemia prevalence (P falciparum) Show forest plot

7

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

2.1.1 Baseline before MDA

6

2093

Risk Ratio (M‐H, Fixed, 95% CI)

0.89 [0.62, 1.26]

2.1.2 During‐MDA

2

991

Risk Ratio (M‐H, Fixed, 95% CI)

0.26 [0.07, 0.94]

2.1.3 Post‐MDA <1 month

1

234

Risk Ratio (M‐H, Fixed, 95% CI)

0.12 [0.03, 0.52]

2.1.4 Post‐MDA 1‐3 months

7

5718

Risk Ratio (M‐H, Fixed, 95% CI)

0.25 [0.15, 0.41]

2.1.5 Post‐MDA 4‐6 months

4

3129

Risk Ratio (M‐H, Fixed, 95% CI)

0.63 [0.36, 1.12]

2.1.6 Post‐MDA 7‐12 months

5

3704

Risk Ratio (M‐H, Fixed, 95% CI)

0.86 [0.55, 1.36]

2.1.7 Post‐MDA 13‐18 months

1

243

Risk Ratio (M‐H, Fixed, 95% CI)

0.82 [0.20, 3.34]

2.1.8 Post‐MDA 19‐24 months

1

239

Risk Ratio (M‐H, Fixed, 95% CI)

0.34 [0.06, 1.97]

2.1.9 Post‐MDA 25‐30 months

1

242

Risk Ratio (M‐H, Fixed, 95% CI)

0.89 [0.22, 3.62]

2.1.10 Post‐MDA 31‐36 months

1

246

Risk Ratio (M‐H, Fixed, 95% CI)

1.25 [0.25, 6.31]

2.2 Parasitaemia incidence (P falciparum) Show forest plot

1

Rate Ratio (IV, Fixed, 95% CI)

Totals not selected

2.2.1 Post‐MDA 1‐3 months

1

Rate Ratio (IV, Fixed, 95% CI)

Totals not selected

2.3 Confirmed malaria illness incidence (P falciparum) Show forest plot

4

Rate Ratio (IV, Fixed, 95% CI)

Subtotals only

2.3.1 Baseline before MDA

3

Rate Ratio (IV, Fixed, 95% CI)

0.87 [0.45, 1.69]

2.3.2 Post‐MDA 1‐3 months

2

Rate Ratio (IV, Fixed, 95% CI)

0.58 [0.12, 2.73]

2.3.3 Post‐MDA 4‐6 months

1

Rate Ratio (IV, Fixed, 95% CI)

0.93 [0.07, 12.43]

2.3.4 Post‐MDA 7‐12 months

3

Rate Ratio (IV, Fixed, 95% CI)

0.47 [0.21, 1.03]

2.3.5 Post‐MDA 13‐18 months

1

Rate Ratio (IV, Fixed, 95% CI)

0.77 [0.20, 3.03]
Figures and Tables -
Comparison 2. MDA versus no MDA in very low to low endemicity (cRCTs) on P falciparum outcomes
Navigate to table in Review
Comparison 3. MDA versus no MDA in very low to low endemicity (cRCTs) on P vivax outcomes

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

3.1 Parasitaemia prevalence (P vivax) Show forest plot

5

Risk Ratio (M‐H, Fixed, 95% CI)

Subtotals only

3.1.1 Baseline before MDA

5

3187

Risk Ratio (M‐H, Fixed, 95% CI)

1.02 [0.86, 1.21]

3.1.2 Post‐MDA <1 month

1

234

Risk Ratio (M‐H, Fixed, 95% CI)

0.18 [0.08, 0.40]

3.1.3 Post‐MDA 1‐3 months

5

2673

Risk Ratio (M‐H, Fixed, 95% CI)

0.15 [0.10, 0.24]

3.1.4 Post‐MDA 4‐6 months

4

3299

Risk Ratio (M‐H, Fixed, 95% CI)

0.78 [0.63, 0.95]

3.1.5 Post‐MDA 7‐12 months

5

4406

Risk Ratio (M‐H, Fixed, 95% CI)

1.12 [0.94, 1.34]

3.1.6 Post‐MDA 13‐18 months

1

243

Risk Ratio (M‐H, Fixed, 95% CI)

0.81 [0.44, 1.48]

3.1.7 Post‐MDA 19‐24 months

1

239

Risk Ratio (M‐H, Fixed, 95% CI)

0.84 [0.38, 1.83]

3.1.8 Post‐MDA 25‐30 months

1

242

Risk Ratio (M‐H, Fixed, 95% CI)

0.89 [0.41, 1.94]

3.1.9 Post‐MDA 31‐36 months

1

246

Risk Ratio (M‐H, Fixed, 95% CI)

1.20 [0.44, 3.29]

3.2 Confirmed malaria illness incidence (P vivax) Show forest plot

2

Rate Ratio (IV, Fixed, 95% CI)

Subtotals only

3.2.1 Baseline before MDA

1

Rate Ratio (IV, Fixed, 95% CI)

1.74 [0.67, 4.53]

3.2.2 Post‐MDA 7‐12 months

2

Rate Ratio (IV, Fixed, 95% CI)

1.38 [0.97, 1.95]
Figures and Tables -
Comparison 3. MDA versus no MDA in very low to low endemicity (cRCTs) on P vivax outcomes
Navigate to table in Review
Comparison 4. Supplemental analysis: post‐hoc subgroup analysis by continent

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

4.1 Plasmodium falciparum parasitaemia prevalence post‐MDA 1‐3 months Show forest plot

7

5718

Risk Ratio (M‐H, Fixed, 95% CI)

0.25 [0.15, 0.41]

4.1.1 Africa

2

1033

Risk Ratio (M‐H, Fixed, 95% CI)

0.97 [0.32, 2.98]

4.1.2 Asia

5

4685

Risk Ratio (M‐H, Fixed, 95% CI)

0.19 [0.11, 0.33]
Figures and Tables -
Comparison 4. Supplemental analysis: post‐hoc subgroup analysis by continent
Navigate to table in Review