Malaria affects more than 250 million people and causes more than a million deaths each year (WHO 2005). One important control strategy against this and other mosquito-borne diseases is mosquito control, which aims to reduce human-mosquito contact. Different control measures are used routinely against mosquitoes and their larvae, including chemical (eg insecticide), biological (eg larvivorous fish or pathogenic fungi), environmental (eg land filling or drainage), and personal protection (eg mosquito repellents formulated as pills, coils, ointments, lotions, and sprays; and insecticide-treated or untreated bed nets).
Electronic mosquito repellents (EMRs) are marketed in response to a huge demand from the public for convenient, safe, and effective antimosquito products. Female Anopheles mosquitoes transmit malaria by sucking blood from humans, and these small handheld, battery-powered EMRs are intended to repel them by emitting a high frequency buzz almost inaudible to the human ear. They can be used both indoors and outdoors, and are claimed to repel mosquitoes within a range of up to 2.5 metres (Kutz 1974; Helson 1977). No adverse effects have been reported in the literature. Mobile phone companies also market a ring tone that is claimed to repel mosquitoes within a one-metre radius (BBC 2003).
Some of the EMRs seem to be based on known aspects of mosquito behaviour, while others have no scientific data to substantiate their claims. Manufacturers have put forward at least two reasons to explain the alleged repellent action of sound against mosquitoes. One reason is that the flight sound of males repels females once they have been inseminated (Foster 1985); hence, whatever mimics the males' flight sound may repel females. However, research has shown that male mosquitoes are actually the ones attracted by the female flight sound and females normally have a very weak sensitivity for sound compared with the males (Wigglesworth 1965; Chapman 1982; McIver 1985; Michelsen 1985). Another reason is that mosquitoes avoid the ultrasonic cries of bats (Foster 1985). Although both explanations may be conceivable, there is no published scientific information to support either idea.
Different brands of EMRs have been examined for their efficacy under laboratory conditions, none of which showed any effects for the devices tested (Singleton 1977; Curtis 1982; Iglisch 1983; Foster 1985; Jensen 2000; Andrade 2001; Cabrini 2006). There are review articles concluding that the EMRs are ineffective in repelling mosquitoes (Coro 1998; Coro 2000). Scientific skepticism over the last 30 years and a successful prosecution of EMR sellers under the UK Trade Description Act in 1980s (Curtis 1994; BBC 2005) seems to have done little to deter manufacturers marketing EMRs and the people who buy them. This is a concern because it is likely to lead to consumers not using other protective methods that are proven to work. This could result in an increased risk of infection with mosquito-borne diseases, especially malaria (Jensen 2000).
Despite the scientific view and research findings, EMRs are still widely promoted and used by the public. We therefore decided to systematically review all reliable research about the effects of high-pitched sounds in preventing mosquito bites and, hence, to assess whether there is any evidence that EMRs have any potential in preventing malaria in the field setting. We included only field studies since laboratory studies do not reflect influences on mosquito behaviour, including climate, mosquito density, and composition of different species in the same locality.
To assess whether EMRs prevent mosquito bites, and to assess evidence of impact on malaria infection.
Criteria for considering studies for this review
Types of studies
Preventing mosquito bites
Field entomological studies that control for geographic site (conducted in same locality), time (conducted at same time), and attractiveness of human participants (by rotating participants between the experiments with and without the EMR).
Preventing malaria infection
Randomized and quasi-randomized controlled trials.
Types of participants
Adults or children.
Types of interventions
- EMRs with any operational wavelength.
- Dummy EMRs, inoperable EMRs, EMRs switched off, or no EMRs. We excluded other repellents and treated or untreated bed nets as control.
- If used, malaria chemoprophylaxis must be identical in both the intervention and control groups.
Types of outcome measures
- The number of mosquitoes of any species landing on exposed body parts of humans acting as baits. Time period defined by entomological collection procedures.
- Malaria infection, defined as clinical malaria (fever with malaria parasitaemia detected by microscopy or rapid test); or asymptomatic malaria parasitaemia.
Search methods for identification of studies
We attempted to locate all relevant studies regardless of language or publication status (published, unpublished, in press, and in progress).
The following databases were searched using the search terms and strategy described in Appendix 1 : Cochrane Infectious Diseases Group Specialized Register (March 2009); Cochrane Central Register of Controlled Trials (CENTRAL), published in The Cochrane Library (2009, Issue 1); MEDLINE (1966 to March 2009); EMBASE (1974 to March 2009); LILACS (1982 to March 2009); Cambridge Scientific Abstracts (CSA) (1982 to March 2009); and Science Citation Index (SCI) (1945 to March 2009).
The following conference proceedings were searched for relevant abstracts: XV International Congress of Tropical Medicine and Malaria, Cartagena, Colombia, August 2000; First MIM Pan-African Malaria Conference, Dakar, Senegal, 6 to 9 January 1997; Second MIM Pan-African Malaria Conference, Durban, South Africa, 15 to 19 March 1999; Third MIM Pan-African Malaria Conference, Arusha, Tanzania, 17 to 22 November 2002; Fourth MIM Pan-African Malaria Conference, Yaoundé, Cameroon, 13 to 18 November 2005; International Conference on Entomology, Brisbane, Australia, 15 to 21 August 2004; and Medicine and Health in the Tropics, Marseille, France, 11 to 15 September 2005.
Researchers, organizations, and manufacturers
We contacted some corresponding authors and field and clinical experts (Professor Chris Curtis, London School of Hygiene and Tropical Medicine; Dr Morteza Zaim, WHO Pesticide Evaluation Scheme (WHOPES), Geneva, to enquire about other published or unpublished relevant studies (September 2006). We also contacted EMR manufacturers (Isotronic, Lentek International Inc., Electronic Pest Controls Ltd.) for unpublished and ongoing trials or studies (September 2005).
We also checked the reference lists of all studies identified by the above methods.
Data collection and analysis
Selection of studies
AAE scanned the results of the literature search for potentially relevant studies and then retrieved the full articles. AAE and PG independently assessed the potentially relevant studies using an eligibility form based on the inclusion criteria; disagreements were resolved through discussion.
Data extraction and management
We independently extracted data from all included studies using a data extraction form and resolved any disagreements in the extracted data by referring to the original paper and through discussion. We described the devices tested and the number of observations made, and assessed the quality of the studies in relation to whether they controlled for study locality, time of day or night, participants (used same people to bait mosquitoes), and whether the observers were blinded. We also assessed the number of times observations were repeated to gain some quantitative measure of quality and grouped this with an arbitrary cut off into adequate (20 or more) or inadequate (less than 20).
Assessment of risk of bias in included studies
We summarized the results of the risk of bias assessment in Table 1.
The number of landings on which the rates were calculated varied considerably according to different ecological and geographical situations, mosquito species, and season and time of day of the tests. If possible, we would have tested for a difference using the original data to measure a mean difference between arms within one study and to calculate 95% confidence intervals. Had a difference been shown, we would also have examined the effects of EMR by a variety of factors: EMR frequency (< and ≥ 20 kHz); mosquito population density; malaria endemicity (< and ≥ entomological inoculation rate of 1/person/night); and mosquito species.
Description of studies
We identified 18 potentially relevant studies of EMR to prevent mosquito bites and included 10 (see 'Characteristics of included studies' for study details); none were randomized or quasi-randomized controlled trials that used EMR to prevent malaria. We excluded eight studies because they were only laboratory based or because they did not provide any data or did not control for locality, time, and blinding (see 'Characteristics of excluded studies'). The papers for the 10 included studies contained 22 experiments, of which 15 were field experiments that met the reviews inclusion criteria; the excluded seven experiments were only laboratory based or used chemical repellents.
Seven studies were carried out in the North America, three in Canada (Helson 1977; Belton 1981; Lewis 1982) and four in the USA (Gorham 1974; Kutz 1974; Garcia 1976; Schreck 1977). Two studies were done in Africa, in Gabon (Sylla 2000) and in The Gambia (Snow 1977). One study was undertaken in Russia (Rasnitsyn 1974).
Seven studies gave the commercial name of the EMRs tested (Rasnitsyn 1974; Helson 1977; Schreck 1977; Snow 1977; Belton 1981; Lewis 1982; Sylla 2000); five studies gave some information about the ultrasound frequencies used, which ranged from 125 Hz to 74,600 Hz (Kutz 1974; Belton 1981; Rasnitsyn 1974; Snow 1977; Sylla 2000). The other studies gave no commercial name and no details of the frequencies used.
All studies counted mosquitoes landing on the bare body parts (mostly arms, legs and/or feet) of the human participants for definite time periods with the EMR switched on or off, or, in some studies, with or without a functional EMR as case and control. None of the field studies performed in North America and Russia were on Anopheles mosquitoes; they used Aedes, Culex, Culiseta, and Mansonia mosquitoes. The two studies in Africa were on Anopheles as well as other mosquitoes. The lowest number of observers was one (with 7 observations; Gorham 1974) and the highest was 18 (with 324 observations; Sylla 2000). Also, timing and length of collections varied, ranging from one minute (Kutz 1974) to over a 12-hour period (Sylla 2000).
Risk of bias in included studies
As shown in Table 1, all studies made some attempts to control for locality (geography) of the study area (wooded area, forest, plain, beach) as they measured landing rates with and without the EMR in the same geographical area. Different individuals may have different attraction for the mosquitoes, and the studies controlled for this, usually by swapping the EMR between the participants so that the same individuals acted as both case and control.
The density of mosquitoes and the intensity with which they attempt to bite varies throughout day and night. Our inclusion criteria required studies to attempt to control for time. This was clearly described in all but one study (Garcia 1976).
One study blinded the observers to whether they were measuring during a control or experimental phase (Sylla 2000). The use of blinding was unclear or not used in the other studies.
In order to test significance, we intended to consider rates per person in participants in control or intervention areas. As landing rates were not given per person, we were unable to compare these data statistically. We recorded the number of times observations were repeated to gain some insight into data quality, arbitrarily defining this as adequate (with 20 or more repeated observations or human participants) and inadequate (less than 20 observations). Three studies were of adequate quality by this criterion.
Effects of interventions
The number of mosquitoes landing per collection with and without EMR are presented in Table 2. All 10 studies reported that the landing rates with and without the EMR were little different and that the EMRs failed to repel mosquitoes. These results occurred regardless of the study location, mosquito density, mosquito genera, or time of study (ie day or night with day-biting and night-biting mosquitoes).
No trials were found to assess the effects of EMRs on malaria infection.
The included studies were of good quality, had controlled for locality, and all but one had explicitly controlled for time of day or night, and the human bait for the mosquitoes. The results of this review provide clear evidence from field-based studies that there is no hint that these devices have any effect on mosquito landing rates. The studies reported here examined the effectiveness of the EMRs with different methods, settings, mosquito species (since they may respond differently to the high-pitched sounds emitted by the EMRs), frequencies of the sound emitted by the EMRs (since mosquitoes may respond to a particular sound wavelength), and times of day (since day-biting and night-biting mosquitoes may behave differently to the sound emitted by the EMRs), and mosquito density (since this may affect EMR efficacy), but none of them supported the claims of the EMRs' effectiveness. Although we did not conduct a meta-analysis of the included studies, there was no suggestion of difference in landing rates between cases and controls in any trial. In 12 of the 15 experiments, the landing rates in the groups with functioning EMR was actually higher than in the control groups. The absolute number of mosquitoes landing on the human participants during the experiments while the EMR was functioning was too high to consider the EMR a repellent.
EMRs are claimed to be effective by mimicking the sound waves produced by the beating of male mosquitoes' wings, especially during swarms. Female mosquitoes, which bite humans, are claimed to be repelled by this sound since they mate only once in their lives. Hence the repellent mechanism should be based on the hearing mechanism in females. However, this theory is implausible since the hearing ability of the females is relatively weak (Wigglesworth 1965; Chapman 1982; Michelsen 1985). It is the hearing system of males that is relatively strong, and the presence of numerous sound and vibration receptors (known as Johnston organ) on their plumose antennae enables them to detect the vibration in the environment as well as the sound of female mosquitoes (Chapman 1982). Thus it is not surprising that the included studies did not produce any evidence that EMRs act as repellents.
Implications for practice
EMRs are not effective in repelling mosquitoes and should not be recommended or used.
Implications for research
There is no evidence of an effect of EMRs on landing rates. Thus there is no evidence that these EMRs could potentially be useful in preventing malaria in humans. Given these findings from 10 carefully conducted studies, it would not be worthwhile to conduct further research on EMRs in preventing mosquitoes biting or in trying to prevent the acquisition of malaria.
AA Enayati developed the protocol of the review during the Fellowship Programme organized by the Cochrane Infectious Diseases Group in March 2005. The Department for International Development (DFID) UK supports this Programme through the Effective Health Care Research Programme Consortium at the Liverpool School of Tropical Medicine.
This document is an output from a project funded by the DFID for the benefit of developing countries. The views expressed are not necessarily those of DFID.
Data and analyses
This review has no analyses.
Appendix 1. Search methods: detailed search strategies
Last assessed as up-to-date: 9 March 2009.
Protocol first published: Issue 3, 2005
Review first published: Issue 2, 2007
Contributions of authors
AA Enayati developed and wrote the review. P Garner helped develop the review, developed and applied inclusion and quality criteria, extracted data, and helped write the review. J Hemingway initiated the review and helped with technical issues. AA Enayati is the guarantor.
Declarations of interest
Sources of support
- Liverpool School of Tropical Medicine, UK.
- Department of International Development, UK.
2012, Issue 4: Status: Current question – no update intended. Further research unlikely to change conclusions.
As of 15 February 2010, this Cochrane Review is no longer being updated, as there is high-quality evidence that electronic mosquito repellents are not effective, meaning further research is unlikely to change our confidence in the estimate of effect.
The review status is a pilot system used by the Cochrane Infectious Diseases Group to help the reader understand whether the review is concerns a current question, and is up to date.
We report on:
1. The question the review addresses. Is it a:
- Historical question, where the intervention or policy has been superseded by new medical developments (such as a new drug); or a
- Current question, which is still relevant to current policy or practice.
2. Whether the review is up to date. Is the review:
- Up to date;
- Update pending; or
- No update intended.
We then provide comment on the review status, to help explain the categories selected.
Medical Subject Headings (MeSH)
MeSH check words
Animals; Female; Humans