Mass drug administration for malaria

  • Protocol
  • Intervention



This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To assess the impact of antimalarial MDA on population asexual parasitaemia, gametocytemia, anaemia, clinical illness, mortality, and MDA-associated adverse events.


Description of the condition

Malaria is the most common and deadly parasitic infection. It causes an estimated 243,000,000 clinical episodes and 863,000 deaths annually, primarily among young children in sub-Saharan Africa (WHO 2009). Four main species of the malaria parasite infect humans: Plasmodium falciparum, P. vivax, P. ovale, and P. malariae. P. falciparum and P. vivax cause the majority of infections, while P. falciparum is responsible for most cases of severe and potentially fatal malaria.

In nature, malaria parasites spread by successively infecting two types of hosts: humans and female Anopheles sp. mosquitoes. All four species of malaria parasites are transmitted to humans via a mosquito bite. Malaria parasites invade hepatocytes, then multiply in the liver as liver schizonts, before being released into the blood stream, invading erythrocytes and developing into asexual blood schizonts. This cycle of erythrocyte invasion and destruction often leads to clinical disease manifested as fever, chills, and flu-like illness symptoms. Left untreated, persons may develop severe complications and die. During this erythrocytic phase, some parasites differentiate into sexual erythrocytic stages, called gametocytes. These gametocytes may be ingested by a mosquito during a blood meal, potentially perpetuating another cycle of growth and multiplication in the mosquito. Additionally, P. vivax and P. ovale can develop dormant liver stage parasites (hypnozoites), which can reactivate and cause relapsing malaria several months or years after initial infection. 

Malaria is both preventable and treatable. Prevention efforts have focused on vector control strategies to reduce adult mosquito populations and human-mosquito contact, and to eradicate mosquito breeding grounds. They include the use of insecticide treated bednets, indoor residual spraying, larviciding, and environmental management. In addition, treatment strategies in endemic areas frequently combine case management, the diagnosis and treatment of malaria patients, with disease prevention by administering antimalarial drugs for prevention of clinical disease to particularly vulnerable population groups such as pregnant women, infants, and non-immune travellers to endemic areas. Success in malaria control using these existing tools has led to renewed interest in the possibility of malaria elimination in some countries or regions.  While the malaria eradication agenda of the mid-twentieth century was ultimately abandoned, current calls for elimination stress the need for new technologies (insecticide delivery systems, new drugs and insecticides, and candidate vaccines) and the revitalization of older strategies (indoor residual spraying).  Mass drug administration (MDA) was a component of many malaria elimination and eradication programmes in the 1950s, but is not currently recommended due to concerns about efficacy, logistical feasibility, sustainability, and the risk of accelerating drug resistance.  However, evidence to guide this recommendation, particularly in light of the development of new antimalarial drugs, is lacking (WHO 2007).

Description of the intervention

Over the past 20 years, annual MDA has been a key strategy for controlling and eliminating highly prevalent neglected tropical diseases such as lymphatic filariasis, soil transmitted helminthes, onchocerciasis, schistosomiasis, and trachoma. The simultaneous administration of essential medicines to target high-prevalence neglected tropical diseases has two main functions: to treat prevalent infection and subsequently to reduce further transmission within the population (Hotez 2009). Mass antimalarial drug administration, defined as the empiric administration of a complete therapeutic course of an antimalarial regimen to an entire population or well-defined sub-population at the same time, has been used for malaria control since the early 1930s and was advocated by the World Health Organization (WHO) in the 1950s as a tool in situations where other more conventional control measures had failed (von Seidlein 2003). While not widely used, there are several examples where MDA, in combination with other malaria control measures, had some success. For instance, MDA with sulfalene-pyrimethamine combined with indoor residual spraying achieved high levels of malaria control during the Garki project in Northern Nigeria in 1969 (Molineaux 1980). In addition, the use of MDA combined with other malaria control interventions succeeded in permanently interrupting malaria transmission on the island of Aneityum in Vanuatu (Kaneko 2000). Primaquine, the only registered drug that can eliminate gametocytes, was also given in combination with chloroquine to an estimated 70% of Nicaragua's population in 1981, preventing an estimated 9200 cases of malaria (Garfield 1983).  In these instances, the entire population was simultaneously treated with a therapeutic dose of an antimalarial both to reduce malaria burden and potentially to interrupt transmission.

How the intervention might work

Malaria transmission is dependent on mosquito vector dynamics, the proportion of humans with peripheral gametocytemia, and the infectiousness of circulating gametocytes to mosquitoes.  MDA of antimalarials might reduce malaria burden by its direct effect on individuals who receive a treatment dose of antimalarials, as well as by reducing rates of transmission. Antimalarial MDA could reduce transmission in a number of ways. First,  MDA could reduce parasitemia prevalence and potentially reduce malaria transmission via inhibition of the liver or asexual intraerythroctytic stages of the parasite, thus reducing the number of parasites that can progress to form gametocytes. Second, the  antimalarial drug could have a  direct effect on gametocytes. Third, the antimalarial drug could have a sporonticidal effect and inhibition of the sporogonic cycle in the mosquito.  If all persons in a given population are treated by antimalarial MDA then one would expect an immediate reduction in asexual parasite prevalence in the population, and possibly a sustained reduction in the population parasite prevalence if there was a concomitant reduction in transmission.

Most antimalarial drugs target the asexual blood stage of the parasite, as this stage is responsible for the symptomatic disease.  Blood schizonticidal drugs reduce asexual parasitemia and early stage gametocytes in P. falciparum by preventing the development of mature gametocytes, without having a direct effect on circulating mature gametocytes.  Some antimalarial drugs, such as the artemisinins and 8-aminoquinolines (eg primaquine), have known gametocytocidal activities and have the potential to reduce transmission by reducing circulating gametocytemia.  In addition, the 8-aminoquinolines also inhibit the hypnozoite stage of P. vivax and P. ovale species, thus reducing the chance of relapse (White 2008).

Why it is important to do this review

Several examples illustrate the use of MDA as a malaria control tool.  Since its popularity in the 1950s and 1960s, however, it has fallen into disfavour due to concerns regarding its efficacy, logistical feasibility, sustainability as a malaria control tool, and risk of accelerating drug resistance.  But with a renewed interest in malaria elimination and the development of new gametocytocidal antimalarials such as artemisinins, MDAs are once again being considered as a tool for malaria control (Feachem 2009). Given this renewed interest in conducting MDAs, it is important to review the currently available literature to assess the potential of this strategy to reduce malaria burden and transmission.  In addition, a systematic review of the literature will allow us to define the gaps in our understanding of the potential benefits and risks of this strategy, such as the risk of adverse drug events in an asymptomatic population. This information could guide both future antimalarial MDA interventions and their evaluation.


To assess the impact of antimalarial MDA on population asexual parasitaemia, gametocytemia, anaemia, clinical illness, mortality, and MDA-associated adverse events.


Criteria for considering studies for this review

Types of studies

In this review, we will include MDA studies that report estimates of any one of our outcomes of interest. Due to the nature of the intervention, only studies that are carried out on entire populations or well-defined sub-populations will be included; we will exclude individually randomized trials. We will include randomized and non-randomized studies, including non-randomized controlled trials, controlled before-and-after studies, prospective cohort studies, retrospective cohort studies, historically controlled trials, nested case-control studies, case-control studies, cross-sectional studies, and before-and-after comparisons. Although we will use the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system to assess the quality of all studies (Guyatt 2008), we will include studies of high, moderate, low and very low quality.

Types of participants

Children and adults living in malaria endemic areas pre- and post-MDA.

Types of interventions

For the purposes of this review, an MDA will be defined as the empiric administration of a complete therapeutic course of an antimalarial regimen targeted to an entire population or well-defined sub-population at the same time. The MDA can be conducted with or without other malaria and non-malaria co-interventions, such as the use of insecticide treated nets, indoor residual spraying, 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

Primary outcomes
  1. Asexual parasitaemia

Secondary outcomes
  1. Gametocytemia

  2. Anemia

  3. Clinical illness

  4. Mortality

  5. Adverse events related to MDA using WHO definitions (Edwards 2000)

Search methods for identification of studies

Electronic searches

Search strategy for identification of studies

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


We will search the following databases: Cochrane Infectious Disease Group Specialized Register; Cochrane Central Register of Controlled Trials (CENTRAL), published in The Cochrane Library; MEDLINE+; EMBASE; CABS Abstracts; and LILACS, using the search strategy detailed in the Appendix.  

Searching other resources

Reference lists 

We will also check the reference lists of all studies and articles identified by the above methods and of previously published reviews, as well references listed in review articles (von Seidlein 2003; Greenwood 2004).

Conference proceedings

We will search the following recent proceedings for relevant abstracts: Fifth MIM Pan-African Malaria Conference (Nairobi, Kenya, November 2009) and the 58th Annual American Society of Tropical Medicine and Hygiene Conference (Washington, D.C., November 2009).

Researchers and organizations

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

Data collection and analysis

Selection of studies

All studies that match our inclusion criteria will be included in the review.

Inclusion criteria:


  • Empiric administration of a complete therapeutic course of an antimalarial regimen (no screening or diagnostic testing prior to antimalarial administration).

  • Intended to be given to an entire population or well-defined sub-population.

  • At the same time.

Study design

  • Measurement of at least one outcome of interest.

  • Randomized and non-randomized studies.

Exclusion criteria:

  • Individually randomized trials.

  • Studies of intermittent preventive treatment in pregnant women or infants, where antimalarials are given to the individual at specified time points in the individual's life (eg during the second trimester of pregnancy or at the ninth month of life after a vaccination) rather than to the whole population or sub-population at a specified point in calendar time (eg in September before the start of the malaria transmission season).

Two authors will independently screen the titles and abstracts of our search results for potentially relevant studies. We will retrieve the full report of any study identified by at least one author during the initial screen. Two authors will then review the full reports of all retrieved studies and independently assess eligibility using an eligibility form based on the above inclusion and exclusion criteria. A third author will resolve any discrepancies between the first two authors. We will attempt to contact study investigators if eligibility is unclear after review by the third author. All study reports will be scrutinized to ensure that multiple publications from the same study will be included only once. We will report the excluded studies and the reasons for their exclusion.

Data extraction and management

Using a pre-tested data abstraction form, two authors will independently abstract information on the study characteristics, including the parasite species of interest, study design, setting, MDA coverage, duration of follow-up, methods for ensuring comparability between sites, and outcomes. For cluster randomized trials, we will try to obtain an estimate of the population size in each intervention or control group, the number of clusters in the intervention and control groups, and the unit of clustering. We will resolve disagreements between the two primary authors by consulting with a third author. We will attempt to contact study investigators for missing data.

We will extract dichotomous data (asexual parasitaemia, gametocytemia, adverse events), continuous data (anaemia), and count and rate data (asexual parasitaemia, gametocytemia, clinical illness, mortality). For dichotomous data, we will extract the number of persons experiencing the event and the number of persons in each group. For continuous data, we will extract the mean and standard deviation in each group. For count data, we will extract the number of events in the treatment and comparison group, the total person time at risk in each group or the rate ratio, and a measure of variance.

For cluster randomized studies, we will determine whether there has been any adjustment for clustering and will aim to extract the measure of effect and the adjusted measure of variance. For non-randomized studies, we will determine whether any adjustments have been made in the analysis to reduce confounding, and if any adjustments have been made we will extract the adjusted measure of effect and the adjusted measure of variance.

Assessment of risk of bias in included studies

Two authors will independently assess the risk of bias for both randomized and non-randomized studies with a comparison group using an assessment form. We will assess all studies on sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, recruitment bias, baseline imbalance, loss of clusters, correct analysis, and other potential threats to validity. Each element will be assessed separately and classified as 'yes', 'no' or 'unclear'; details will be presented in a separate risk of bias table for each study. In addition, we will present a risk of bias summary and a risk of bias graph. We will resolve disagreements between the two primary authors by consulting with a third author. We will attempt to contact study investigators if risk of bias is unclear after review by the third author.

Measures of treatment effect

We will report the findings in a summary of findings table for all outcomes of interest.  We will assess the treatment effect of an MDA using risk ratios for dichotomous data, mean difference for continuous data and the rate ratio for count data.  We will use standard error estimates reported in the studies if appropriate methods were used to derive those estimates (eg clustering accounted for in reports from cluster randomized trials). 

Unit of analysis issues

We will not report confidence intervals for cluster randomized studies that have not appropriately adjusted for clustering. In addition, we will not combine the results of studies that have and have not appropriately adjusted for clustering.

Dealing with missing data

No imputation measures for working with missing data will be applied.  We will attempt to contact study investigators to obtain missing data.

Assessment of heterogeneity

We will assess heterogeneity by summarizing study and patient characteristics across the studies. If the data can be entered into forest plots, we will assess heterogeneity by inspecting these plots for overlapping confidence intervals and using the Chi2 test and I2 statistic.

Assessment of reporting biases

If sufficient data are available, we will assess reporting/publication bias using funnel plots, which will be created to examine study effects by plotting relative measures of treatment effect on a logarithmic scale against the standard error (and its inverse). Plotting against both will allow us to emphasize differences between studies of smaller size and studies of larger size.

Data synthesis

Data will be analysed using Review Manager 5 by the authors (EP, JH, JS). Data will be summarized in a table format and forest plots. We will stratify analyses according to randomized versus non-randomized studies. We will present this stratified analysis for each outcome in tables with clear headings delineating randomized versus non-randomized studies. Cluster randomized trials that have not adjusted for clustering will be clearly labelled in the tables and not combined with other studies in the meta-analysis. We will only conduct a random-effects or fixed-effect meta-analysis if we identify a sufficient number of high quality studies with both an outcome indicator estimate and a measure of precision. Random-effects models will be used if substantial heterogeneity is identified.

Subgroup analysis and investigation of heterogeneity

We expect substantial heterogeneity. We will assess statistical heterogeneity using the Chi2 test and will not perform a meta-analysis if we identify substantial heterogeneity. In addition, we anticipate the following sub-group analysis: studies that report early outcome measures (< 6 months after MDA) and studies that report late outcome measures (≥ 6 months after MDA).

Sensitivity analysis

To determine the robustness of the results from this meta-analysis, heterogeneity will be explored in the framework of sensitivity analyses. We will perform sensitivity analysis by assessing: high and moderate quality studies only; the use of stand alone MDA for malaria control (ie no other malaria-specific co-interventions); the use of MDA with chloroquine/primaquine for control of P.vivax.


The editorial base is funded by the UK Department for International Development (DFID) for the benefit of low- and middle-income countries. The review is funded by the Centers of Disease Control and Prevention. Eugenie Poirot is funded by a fellowship from the Association of Schools of Public Health.


Appendix 1. Search strategy


A. Anti-Malarials

exp Antimalarials/ or exp Malaria/ or antimalarial* or anti-malarial* or ((schizonticidal* or gametocidal* or hypnozoiticidal* or drug* or treatment) and (malaria*))

B. Mass Administration

((mass or coordinate*) adj5 (administ* or distribut* or applicat* or "use" or therap* or treatment*))


A. Anti-Malarials

exp antimalarial agent/ or exp malaria/ or antimalarial* or anti-malarial* or ((schizonticidal* or gametocidal* or hypnozoiticidal* or drug* or treatment) and (malaria*))

B. Mass Administration

((mass or coordinate*) adj5 (administ* or distribut* or applicat* or "use" or therap* or treatment*))


A. Anti-Malarials

(Must run each MeSH term separately. Ovid syntax used for recording purposes.)

exp Antimalarials/ or exp Malaria/ or antimalarial* or anti-malarial* or ((schizonticidal* or gametocidal* or hypnozoiticidal* or drug* or treatment) and (malaria*))

B. Mass Administration

((mass or coordinate*) near/5 (administ* or distribut* or applicat* or "use" or therap* or treatment*))


A. Anti-Malarials

ti=(antimalarial* or anti-malarial* or ((schizonticidal* or gametocidal* or hypnozoiticidal* or drug* or treatment) and (malaria*))) or ab=(antimalarial* or anti-malarial* or ((schizonticidal* or gametocidal* or hypnozoiticidal* or drug* or treatment) and (malaria*))) or de="antimalarials"

B. Mass Administration

(mass) and (administ* or distribut* or applicat*)


A. Anti-Malarials

antimalarial* or anti-malarial* or ((schizonticidal* or gametocidal* or hypnozoiticidal* or drug* or treatment) and (malaria*))

Declarations of interest

All authors report no conflict of interest.