Larvivorous fish for malaria prevention

  • Protocol
  • Intervention

Authors


Abstract

This is the protocol for a review and there is no abstract. The objectives are as follows:

To determine the ability of larvivorous fish to prevent malaria by evaluating the impact on:

a) confirmed malaria cases in humans; and/or
b) the density of adult malaria vectors.

If research evidence for an effect on confirmed malaria cases and density of adult vectors is equivocal or insufficient, we will examine the evidence for potential efficacy from field studies meeting minimum design criteria that examine the effects of larvivorous fish on the density of vector larvae or the prevalence of breeding sites positive for larvae of the vector.

Background

Description of the condition

Malaria is the most common vector-borne disease in the world and is endemic in 109 countries. It is responsible for an estimated 247 million cases and nearly a million deaths per year, predominantly in children under five years in sub-Saharan Africa (WHO 2008). The primary strategies for preventing and controlling malaria are:

  1. early diagnosis and effective treatment of malaria cases, including intermittent presumptive treatment of pregnant women in highly endemic areas (Garner 2006);

  2. prevention through mosquito control, using indoor residual spraying (Tanser 2007) of insecticides approved by the WHO Pesticide Evaluation Scheme (WHOPES), and the use of insecticide-treated mosquito nets (Lengeler 2004), including long-lasting insecticide treated nets.

The two primary malaria vector (anopheline mosquitoes that transmit malaria) control strategies, indoor residual spraying and insecticide-treated nets, were developed against the most effective malaria vectors, which share the attributes of feeding late at night, being anthropophilic (preferring to feed on humans), endophagic (preferring to feed indoors), and endophilic (preferring to rest indoors). However, many vectors, particularly in Asia and South America (but also in Africa), prefer animals to humans for their blood meals (are zoophilic), or feed either early in the evening or outside of houses where they will be less likely to encounter either residual spray or insecticide-treated nets. Thus the two main vector control strategies are likely to be less effective in regions where the malaria vectors have these behavioural attributes. These concerns have led some agencies and governments to propose other strategies for vector control, including efforts against the larval stages of the mosquito. 

Description of the intervention

Both insecticide-treated nets and indoor residual spraying directly affect malaria transmission as these strategies attack the adult mosquitoes which transmit the disease to man. Interventions directed against the larval stages indirectly attempt to control malaria by seeking to reduce the size of the adult vector population.  Among these strategies are environmental management, biological control (introducing fish (Pyke 2008; Walton 2007), frogs (Raghavendra 2008), and invertebrate predators such as dragonfly nymphs) or the use of chemical and microbial larvicides against the aquatic immature mosquito stages (Lacey 1990). Biological control methods are promoted to avoid the environmental impact of insecticides.

One of the most widespread of the biological control strategies is the use of fish which attack the mosquito larvae and pupae (Chandra 2008). The potential of one such larvivorous fish Gambusia to ingest large numbers of mosquito larvae led to a series of laboratory-based studies on mosquito larval prey preferences as well as the optimization of systems to propagate these fish. Subsequently field evaluations of Gambusia for impact on larval prevalence and density in mosquito larvae breeding sites were undertaken. Gambusia spp are native to the south-eastern United States but were transported and released in multiple countries around the world, leading to this species being one of the most geographically dispersed freshwater fishes in the world (Pyke 2008).

However, concerns regarding the predation of Gambusia on native fishes when introduced into new areas led to the evaluation of native fish species for larval control. Approximately 300 species of fish have been considered for mosquito control with the most promising belonging to the genera Aphanius, Valencia, Aplocheilus, Oryzias, Epiplatys, Aphyosemion, Roloffia, Nothobranchius, Pachypanchax, Rivulus, Fundulus, Cynolebias, and Cyprinodon (Walton 2007).

How the intervention might work

As malaria is transmitted by the adult Anopheles mosquito, the intensity of transmission relates to a) the proportion of Anopheles with salivary glands infected with the sporozoite stage of the malaria parasite; and b) the number of vectors taking blood meals on humans. The product of these two factors known as the entomological inoculation rate is a direct measurement of the intensity of malaria transmission and is generally expressed as the number of bites by sporozoite infected mosquitoes on a human per year.

The anopheline vectors of malaria lay their eggs in water, in which larvae develop. Breeding site preferences may be very specific (for example, the mosquito may lay eggs only in one or two types of water sources) or a species may use a wide variety of breeding sites (such as temporary ground water pools including footprints and ditches, as well as more permanent water sources such as swamps and wells). Hence the abundance of adult mosquitoes is dependent, in part, on the number and size of potential breeding sites in which eggs can be laid, the density of the larval stages in the available breeding sites, and a number of other ecological/environmental factors such as temperature, rainfall patterns, availability of humans, and other blood meal sources.

The larger the mosquito population, the larger the potential number of bites by vectors on humans unless people take measures to avoid mosquito bites, such as sleeping under a mosquito net. As the density of vectors is directly related to the entomological inoculation rate, the greater the mosquito man-biting rate, the greater the malaria transmission and therefore the higher the incidence of malaria. If the size of the malaria vector population is limited by interventions that reduce the number of breeding sites or the density of vector larvae per breeding site, then transmission of malaria to humans (all other factors remaining the same) might potentially be reduced. The potential impact of larvivorous fish on malaria might therefore be evaluated indirectly by outcomes in the larval stages. However, impacts on transmission can be measured more directly based on the density of adult mosquitoes or by malaria incidence in the human population.

Why it is important to do this review

The use of larvivorous fish has been promoted as an environmentally friendly alternative to insecticide-based interventions for the control of malaria for at least 35 years. A World Health Organization (WHO) sponsored inter-regional conference on malaria control in 1974 reported that "the utilization of larvivorous fish, mainly Gambusia or suitable local species, is the only practical measure that can be recommended where applicable, as in lakes, ponds, pools, wells, rice fields, etc" (WHO 1974). A 2001 regional meeting in Kazakhstan recommended that more trials on larger numbers of local larvivorous and phytophagous fish be undertaken in different eco-epidemiological settings in that region as well as continuing to search for effective larvivorous fish (WHO 2001). Subsequently, momentum has gathered in the efforts to eliminate malaria, resulting in the 2006-2015 WHO-EURO strategy to eliminate malaria, which included larval control by larvivorous fish (WHO 2006). Currently, the use of fish is also included among the recommended vector control strategies for malaria elimination in low to moderate endemic countries (WHO in press).

Despite these global WHO recommendations, the use of fish remains a highly contentious issue due to both the uncertain efficacy for malaria control and concerns around potential detrimental environmental impacts. A large reduction in numbers of larvae by fish may not significantly reduce the size of the adult population, since it merely replaces the usual density dependent regulation of the larval population. We therefore decided to systematically review all reliable research about the efficacy of larvivorous fish to reduce the transmission of malaria.

Objectives

To determine the ability of larvivorous fish to prevent malaria by evaluating the impact on:

a) confirmed malaria cases in humans; and/or
b) the density of adult malaria vectors.

If research evidence for an effect on confirmed malaria cases and density of adult vectors is equivocal or insufficient, we will examine the evidence for potential efficacy from field studies meeting minimum design criteria that examine the effects of larvivorous fish on the density of vector larvae or the prevalence of breeding sites positive for larvae of the vector.

Methods

Criteria for considering studies for this review

Types of studies

Studies to measure the impact of fish on malaria transmission need to be robust and without confounding or biases which give spurious results. For this review, we will first establish robust, experimental and quasi- experimental study designs for inclusion in this review, and make clear how these are monitored to measure changes in parasitology and entomology outcomes.We will adapt the inclusion criteria for such studies from the WHO recommendations for field trials of larvivorous fish (Ungureanu 1981). We have distilled from the WHO recommendations, the minimal attributes for a study of acceptable scientific rigour. Studies must include a control group with baseline measurements taken in both the control and intervention groups. We will require that control and intervention sites have a) baseline information, b) contemporaneous data collection, c) same locality, d) comparable resident populations in relation to ethnic group, housing, and wealth, (e) similar intensities of malaria transmission, and (f) of sufficient geographic size to minimize immigrating vectors masking the impact of the intervention.

Ideally we will include randomized controlled trials, but will also include non-randomized trials including before-and-after controlled trials with at least one site in both the intervention and control areas. We may include controlled interrupted time series designs that meet the inclusion criteria for this systematic review. In all studies, the intervention and control groups must have been comparable in access to anti-malaria interventions, with the exception of the introduction of larvivorous fish in the intervention area. This is straightforward for the human outcomes (malaria cases), but is more complicated for the entomological outcomes. Controlled before-and-after studies must have measured outcomes (larval density, prevalence of breeding sites with larvae, adult vector density and/or malaria incidence) before and after the intervention with at least one control area (Figure 1). These studies will need to have a series of measures of the impact of fish on vector populations in which adult and larval vector populations were monitored before the introduction of the fish.  Because of the natural variation in abundance (seasonality) of malaria vectors in both the larval and adult stages (partly related to season and climatic conditions and partly random), the outcomes of these studies need to be monitored for one or more full years prior to the introduction of the intervention, followed by multiple assessments over 12 or more months thereafter in both intervention and control areas. 

Figure 1.

Experimental designs that have been used to attempt to evaluate the impact of fish on the larvae of malaria vectors in malaria endemic countries are shown in Figure 1.

Types of participants

Children and adults living in rural and urban malaria areas.

Types of interventions

Interventions

Introduction of larvivorous fish of any species, either adults or juveniles, to mosquito breeding sites, with or without other malaria co-interventions.

Controls

Control areas will have no fish introduced; malaria co-intervention(s) in the control arm of the studies will not differ from the intervention arm.

Types of outcome measures

Primary outcomes

The primary direct outcomes of interest are the effects on the incidence of diagnostically confirmed cases of malaria, or the density of the adult malaria vector population. As the density of biting adult vector mosquitoes is directly related to the entomological inoculation rate, reducing the adult vector population should correspond comparably to a reduction in the number of malaria cases in humans.

The primary outcomes are direct measures of changes in the intensity of malaria transmission:

  1. the malaria incidence in the human population; and

  2. the density of adult vector mosquitoes (which is directly related to the entomological inoculation rate).

Confirmed malaria cases

Malaria infections, defined as laboratory confirmed malaria cases (malaria parasitaemia detected by microscopy or rapid diagnostic tests in either active or passive case detection). We will use confirmed malaria cases to calculate the incidence of malaria.

Preventing mosquito bites

Estimates of the density of the malaria vectors measured by a technique previously shown to be appropriate for the vector. This includes measuring adult mosquito density by counts of vectors either landing on exposed body parts of humans acting as baits or collected resting inside buildings using knockdown spray catches.

Secondary outcomes

Secondary outcomes include changes in the density of vector larvae or the prevalence of breeding sites positive for the larvae of the malaria vector(s). Impacts on the larval population are considered a secondary outcome due to the uncertainty of predicting how a change in the larval vector population in the sampled breeding sites will impact changes in the adult vector population and hence, malaria transmission to humans.  Significant reductions in either the densities of larvae in breeding sites, or the prevalence of breeding sites positive for the vectors, would suggest that the intervention has the potential to reduce the adult vector population and thereby reduce transmission of malaria to humans. If the secondary outcomes identified in this review do not demonstrate an effect, then it is unlikely that larvivorous fish are an intervention that can significantly reduce the transmission of malaria. If the secondary outcomes are significantly diminished, then the intervention has the potential to be beneficial in reducing malaria. In these circumstances, what is required before recommending this strategy for implementation is to demonstrate an effect on transmission; whether this is density of the adult vector population, or an effect on confirmed malaria episodes in humans.

Secondary outcomes are those which have the potential to have an impact on malaria transmission. The secondary outcomes will be:

  1. the density of immature vector stages; and

  2. the prevalence of breeding sites positive for a malaria vector.

Adverse effects

Dissemination of larvivorous fish as a malaria control strategy has the potential for adverse effects on local ecosystems by reducing or eliminating indigenous fish and amphibians as well as other invertebrates (Walton 2007). Whilst we would not anticipate many data sources regarding adverse impacts from controlled trials in this regard, we will examine studies for any measure of these potential harms. We will note if studies examine or measure effects on indigenous fish species or other native organisms.

Search methods for identification of studies

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

Electronic searches

We will search the following databases using the search terms detailed in Appendix 1: the Cochrane Infectious Diseases Group Specialized Register; the Cochrane Central Register of Controlled Trials (CENTRAL), published in The Cochrane Library; MEDLINE; EMBASE; CABS Abstracts; and LILACS. We will also search the metaRegister of Controlled Trials (mRCT) using 'malaria' and 'larvicide* or fish' as search terms.

We will also search the literature database of the Armed Forces Pest Management Board using the terms: ('frogs' and 'fish') and 'malaria' and examine the Tropical Diseases Bulletin and the archives of the World Health Organization.

Searching other resources

Reference lists 

We will also check the reference lists of all studies identified by the above methods and of previously published reviews as well references listed in review articles (Beltran 1973; Chandra 2008; Pyke 2008; Walker 2007) and previously compiled bibliographies (Gerberich 1968) as well as the references lists in articles identified by the above searches.

Conference proceedings 

We will search the conference proceedings of the MIM Pan-African Malaria Conferences, the American Society of Tropical Medicine and Hygiene, the American Mosquito Control Association and the Society for Vector Ecology for relevant abstracts.

Researchers and organizations

We will also contact heads of malaria control and prominent researchers in countries with active or former programmes using larvivorous fish to request access to both published and unpublished manuscripts describing controlled trials. Individuals to be contacted will include Dr Yohannes Ambachew (WHO), Dr Steven Bjorge (WHO-SEARO), Dr Mikhail Ejov (WHO-EURO), Dr Andrew Beljaev (WHO, retired), Dr RR Abeyasighe (Director, Anti-malaria Campaign, Sri Lanka), Dr Supratman Sukowati (NIH, Indonesia), Dr Rita Kusriastuti (Indonesia National Malaria Control Program), Dr Arbani Roos Poerwokoesoemo (Indonesian National Malaria Control Program, retired), Dr SK Ghosh (National Institute of Malaria Research, India), and Dr Craig Stoops (United States Navy).

Data collection and analysis

Selection of studies

We will divide the search results equally amongst the review authors (AA, GP, RW, and RB) with two review authors screening each abstract from the search for potentially relevant studies. TB will retrieve the corresponding full articles from these identified studies and AA, GP, RW, and RB will assess eligibility using an eligibility form. Two authors will independently screen each search result and assess each article. Any discrepancies between the eligibility results of the two review authors will be resolved by discussion with a third co-author (TB). If the eligibility is unclear we will write to the study authors for clarification. We will scrutinize the study reports to ensure that multiple publications from the same study are included only once. We will list the excluded studies and the reasons for their exclusion in the 'Characteristics of excluded studies' table.

Data extraction and management

AA, GP, RW, and RB will extract data from the study reports. Two co-authors will independently extract data from each study report into a pre-designed data extraction form. Any discrepancies between the results of the two review authors will be resolved by discussion with a third co-author (TB). If relevant data are unclear or not reported we will write to the trial authors for clarification. 

We will extract information on the study characteristics and study methods, including setting, methods for ensuring comparability between sites, types of fish, and outcomes and how these were measured. 

We will extract the number of confirmed malaria patients, the denominator populations for estimating incidence, breeding sites, sentinel sites for measuring densities of adult mosquitoes, and clusters (i.e. communities or villages etc.) allocated to each intervention group for each study. We will extract co-interventions and we will examine whether both control and intervention arms experience the same co-interventions. Co-interventions include access to and use of insecticide-treated nets (including long-lasting nets), indoor residual spraying, larvicides (including bio-larvicides and insect growth regulators), polystyrene beads, source reduction activities, and environmental management. 

Intervention and control arms of included studies should also not differ in access to or use of interventions directed against the parasite including intermittent presumptive treatment (of either pregnant women or infants/children), mass treatment with anti-malaria drugs, active and passive case detection with presumptive malaria treatment for fever, and/or radical treatment of laboratory-confirmed cases. Furthermore, monitoring for malaria cases should be comparable in both intervention and control arms of the studies included in this review. 

Cluster-randomized studies

For trials randomized using clusters, we will record the number of clusters in the trial, the average size of clusters, and the unit of randomization (e.g. household or community). We will document the statistical methods used to analyse the trial if possible and examine the methods for adjustments for clustering or other covariates. When reported, estimates of the intra-cluster correlation (ICC) coefficient for each outcome should be recorded. We will contact authors to request missing information. For cluster trials, where results have been adjusted for clustering, we will extract the point estimate and report the 95% confidence interval. If the results are not adjusted for clustering, we will extract event rates for dichotomous outcomes and means or medians for continuous outcomes as described above.

  • For the dichotomous outcome of infections with confirmed malaria in people, we will aim to extract the number of participants with malaria and the number of individuals in each arm of the study experiencing the event and the number of patients allocated to each intervention group. If these data are not reported we will extract percentages or measures of effect, such as risk differences etc.

  • For the continuous outcomes (larval and adult mosquito densities, number of breeding sites), we will aim to extract means and standard deviations for each treatment group together with the numbers of patients in each group. If medians are presented, we will extract them and aim to also extract ranges. 

Non-randomized (controlled before-and-after) studies

We will extract details of study design methods. When the studies have adjusted for a covariate in the analyses and reported an adjusted measure of effect, we will extract the measure of effect and its standard error and record the variable for which the analyses were adjusted. When the study reports a number of different adjusted measures of effect, we will extract each result.

Assessment of risk of bias in included studies

TB, RB, and/or RW will assess the risk of bias for studies. Two co-authors will independently assess the risk of bias for each study using a risk of bias assessment form. Any discrepancies between the results of the risk of bias analysis of the review authors will be resolved by discussion with a third co-author (PG). If relevant data are unclear or not reported we will write to the trial authors for clarification.

Cluster-randomized studies

For cluster-randomized trials we will assess six main components: sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, and other biases (recruitment bias, baseline measurements, missing clusters, no adjustment, and herd effect).

For sequence generation and allocation concealment we will describe the methods used. For blinding we will describe who was blinded and the blinding method.  For incomplete outcome data we will report the percentage and proportion lost to follow up (the number of patients for which outcomes are measured/the number randomized). For selective outcome reporting we will state any discrepancies between the methods and the results in terms of the outcomes measured and the outcomes reported.

Judgments will be classified as “yes”, “no”, or “unclear” indicating a low, high, or unclear/unknown risk of bias respectively. We will record the results of the assessment using the 'Risk of bias' summary and 'Risk of bias' graph in addition to the 'Risk of bias' tables. Where our judgement is unclear we will attempt to contact the trial authors for clarification.

Non-randomized (controlled before-and-after) studies

For controlled before-and-after studies we will assess sequence generation, allocation concealment, incomplete outcome data, blinding, selective outcome reporting, and other biases (baseline outcome comparability, baseline characteristics comparability, contamination).

We will assess risk of bias in controlled before-and-after studies in relation to (1) baseline measurements (were these measured prior to the intervention, with no substantial difference between study groups); (2) characteristics and comparability of the control sites; (3) blinding of the main outcome; (4) whether the control group was unlikely to benefit from the fish or be influenced by the fish; (5) reliability of the main outcome measure; and (6) coverage of surveys assessing impact.

For controlled before-and-after studies, we will check that the data are collected at the same time, that the control site(s) are comparable in relation to malaria endemicity and vector, and there is at least one control and one intervention site. The seasonality of malaria transmission in many areas means that malaria in the human population and the density of adult and larval vector populations should be monitored for a period of at least a year before and a year after introduction of fish. 

Measures of treatment effect

For dichotomous outcomes (infection with malaria), we will present the risk ratio (RR). For continuous outcomes that are summarized by arithmetic means and standard deviations, we will report the mean difference (MD). We will present all results with 95% confidence intervals (CIs).

For all outcomes, we will carry out meta-analysis if sufficient data are available.

Unit of analysis issues

When the analyses have not adjusted for clustering, we will attempt to adjust the results by multiplying the standard errors of the estimates by the square root of the design effect, where the design effect is calculated as DEff=1+(m-1)*ICC. This requires information to be reported, i.e. the average cluster size (m) and the intra-cluster correlation coefficient (ICC). 

For measures of malaria in humans, we will correct cluster-randomized controlled trials for intra-cluster correlation where possible. However, if these data are not available we will present as averages by cluster. If we do find such studies that have been carried out carefully, but analysed without correction, we may conduct a sensitivity analysis assuming a variety of ICC values.

Dealing with missing data

We will report whether patients or communities/villages have been lost to follow up during the time period of the study. We will analyse data according to a complete case analysis.

Assessment of heterogeneity

When cluster-randomized trials are combined in meta-analysis, we will inspect the forest plots to detect overlapping confidence intervals, apply the Chi² test with a P value of 0.10 used to indicate statistical significance, and also use the I² statistic with a value of 50% used to denote moderate levels of heterogeneity.

Assessment of reporting biases

If there are sufficient cluster-randomized trials (approximately 10) we will construct funnel plots to look for evidence of publication bias.

Data synthesis

TB and PG will analyse the data using Review Manager 5 (RevMan 2008). We will stratify the analysis by study design, i.e. non-randomized studies, cluster-randomized trials that adjust for clustering, and cluster-randomized trials that do not adjust for clustering. We will also stratify analyses by the control intervention. We will split the tables by the stratification features or subgroup the forest plots.

Cluster-randomized studies

If the results of the cluster-randomized trials are not adjusted for clustering then we will present the result in a table since the confidence intervals of unadjusted results are artificially narrow and could be misinterpreted in a meta-analysis.

We will combine cluster-randomized trials that do adjust for clustering in meta-analysis. When no statistically significant heterogeneity is detected, we will apply the fixed-effect meta-analysis model. When statistically significant heterogeneity is observed within groups that cannot be explained by subgroup or sensitivity analyses, we will apply a random-effects meta-analysis model to synthesize the data. When substantial heterogeneity is determined from the assessments of heterogeneity, or when a pooled meta-analysis result is considered to be meaningless because of clinical heterogeneity, we will not carry out meta-analysis but will present a forest plot with the pooled effect suppressed and report the I² statistic and P value from a Chi² test.

Non-randomized studies

We will present data from non-randomized studies in tables or in forest plots with the meta-analysis totals suppressed. If multiple measures of effect are reported that are adjusted for different variables then we will report each.

Subgroup analysis and investigation of heterogeneity

If numerous cluster-randomized trials are combined in meta-analysis, we will use subgroup analyses to investigate heterogeneity.

Sensitivity analysis

If numerous cluster-randomized trials are combined in meta-analysis, we will use sensitivity analyses to investigate the robustness of the results.

Acknowledgements

We are grateful to our affiliated Institutions and Organizations, and thank the referees and editors for their comments and encouragement. The editorial base for the Cochrane Infectious Disease Group is funded by the Department for International Development (DFID), UK, for the benefit of developing countries. The findings and conclusions in this report have not been formally disseminated by the Centers for Disease Control and Prevention and should not be construed as representing any agency determination or policy.

Appendices

Appendix 1. Search methods: detailed search strategies

Search setCIDG SRaCENTRALMEDLINEEMBASELILACSCAB ABSTRACTS
1mosquito*mosquito*mosquito*mosquito$mosquito$mosquito*
2

control* OR breeding* OR lar

va* Or predat*

control* OR breeding* OR larva* OR predat*control* OR breeding* OR larva* OR predat*control$ OR breeding$ OR larva$ Or predat$control$ OR breeding$ OR larva$ OR predat$control* OR breeding* OR larva* Or predat*
31 and 21 and 2PEST CONTROL, BIOLOGICALVECTOR CONTROL1 and 21 and 2
4(fish* or frog*)MOSQUITO CONTROL/METHODS2 OR 32 OR 3(fish$ OR frog$)(fish* or frog*)
5larvivorous3 or 41 AND 41 AND 4larvivorouslarvivorous
64 or 5(fish* OR frog*)MOSQUITO CONTROL/METHODS(fish$ OR frog$)4 or 5

“Gambusia” OR

“Poecilia” OR

“Aphanius” OR

“Oreochromis” OR “Tilapia” OR “Aplocheilus”  OR

“Cyprimus” OR

“Ctenopharyngodon” OR “Rasbora” OR

“Aphyocypris”

73 and 6larvivorous  5 OR 6larvivorous  3 and 64 or 5 or 6
86 OR 7(fish* OR frog*)

“Gambusia” OR

“Poecilia” OR

“Aphanius” OR

“Oreochromis” OR “Tilapia” OR “Aplocheilus”  OR

“Cyprimus” OR

“Ctenopharyngodon” OR “Rasbora” OR

“Aphyocypris”

 —3 and 7
9 —5 and 8larvivorous6 or 7 or 8 — —
10 — —

“Gambusia” OR

“Poecilia” OR

“Aphanius” OR

“Oreochromis” OR “Tilapia” OR “Aplocheilus”  OR

“Cyprimus” OR

“Ctenopharyngodon” OR “Rasbora” OR

“Aphyocypris”

 

5 and 9 — —
11 — —8 OR 9 OR 10 — — —
12 — —7 AND 11 — — —

aCochrane Infectious Diseases Group Specialized Register.

History

Protocol first published: Issue 4, 2009

Contributions of authors

All authors provided input to the protocol development.

Declarations of interest

The authors have no conflicts of interest.

Sources of support

Internal sources

  • No sources of support supplied

External sources

  • Department for International Development, UK.

Ancillary