Introduction

Around 55 of the 106 countries with ongoing malaria transmission in 2000 are still on track to meet the World Health assembly target of a 75% reduction in malaria incidence by 2015 [1]. Effective vector control via large-scale use of long lasting insecticidal nets (LLINs) or indoor residual spraying (IRS) in combination with prompt and effective management of malaria cases, contributed largely to the malaria decline worldwide. However, residual malaria transmission caused by early evening and outdoor biting vectors is not covered by present vector control interventions [2, 3]. Therefore, additional vector control tools such as topical repellents need to be evaluated to further advance malaria elimination [4, 5].

A large-scale cluster randomized trial (MalaResT project) was set-up to evaluate the epidemiological impact of topical repellents in addition to the existing vector control measures (LLIN only) in the province of Ratanakiri in Cambodia [6]. To achieve community protection, mass distribution and use of repellents was promoted in the intervention arm aiming to reduce the residual malaria transmission. For over 50 years, DEET has been the most popular insect repellent worldwide. More recently, a synthetic component, namely picaridin, has been developed and is now also endorsed by the World Health Organization (WHO) [7]. In this study, picaridin was chosen as the repellent for its safety profile and excellent efficacy [7, 8]. Moreover, compared to DEET,[9, 10] it has a higher acceptability in terms of odor perception and does not affect plastic materials [11, 12].

Besides a randomized trial showing no adverse neurologic, gastrointestinal, or dermatologic effects among pregnant women applying daily DEET and no adverse effects on survival, growth, or development at birth or at one year among their babies [13], safety information of commonly used repellents mainly comes from studies conducted in small groups of healthy volunteers from high income countries when used for personal protection [11, 14]. We found no single study providing detailed safety information of repellent use when this is massively introduced in the community as part of a public health intervention. In contrast with earlier studies whereby short term exposure was assessed, the use of a topical repellent as a malaria control intervention implies daily skin application during an entire malaria season. Importantly, the tolerance and safety of the employed repellents might critically determine adherence to their use and hence also the effectiveness of the intervention. The aim of this study was to report on safety issues during the mass introduction of the picaridin repellent. Specifically, we report on the frequency of adverse reactions (ARs) reported individually in an expedited manner, and of perceived side effects reported per family during the two-weekly bottle exchanges.

Materials and methods

Study design

The randomized cluster trial was composed of two arms, with clusters consisting of villages, conducted over two years. In the control arm (49 clusters; 56 villages; 5,287 families; 23,789 inhabitants) all individuals were provided with a LLIN (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Flow chart describing the cluster randomized trial, evaluating the epidemiological efficacy of the introduction of insect repellents at the village level on malaria prevalence.

Safety data reported in the present study come from the intervention arm.

https://doi.org/10.1371/journal.pone.0172566.g001

In the intervention arm (49 clusters; 57 villages; 5,642 families; 25,051 inhabitants), besides the LLIN, all people aged over 2 years were additionally provided with the picaridin repellent to use during the early evening and morning when not protected by their bed nets. Children from 2 to 10 years were given picaridin 10% (as a milky lotion); all individuals above 10 years received picaridin 20% (in spray formulation). Both formulations applied at a dosage of 0.5ml/1000cm² were shown to provide an effective protection during at least five hours [15]. No placebo repellent was used in the control arm as its use would have created a false perception of protection.

Picaridin repellent, distribution & promotion

A chemical analysis of both formulations by an independent laboratory (Walloon Agriculture Centre, Belgium, a WHO collaborative center) showed agreement with the WHO specification [16] for the active ingredient content and the detection of impurities (sec-butyl chloroformate, sec-butyl carbonic anhydride) [17] (see S1 Text and S2 Text)

The distribution of picaridin was done from April to December in 2012 and from March to December in 2013, covering the period of malaria transmission. A 100 ml bottle of picaridin (spray or lotion) is sufficient to cover the consumption of an adult and a child for two and four weeks respectively. A two-weekly bottle exchange program, performed by 135 village distributors, was put in place to ensure continuous access to the repellent. Distributors were supervised by nine supervisors from the health centers. In 2013, besides the health center supervisors, 13 additional supervisors were recruited to improve the repellent distribution, promote its use and facilitate data collection. All field activities were under the management of a field coordinator (who is a medical doctor) with two assistants, and all the AEs were given medical care by the field coordinator and another medical doctor based in the Ratanakiri provincial referral hospital (Fig 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Information flow on AEs, perceived side effects and patient management tree.

https://doi.org/10.1371/journal.pone.0172566.g002

The actual monthly repellent consumption was calculated per family expressed as a percentage of the expected consumption in 2012–2013. Repellent distribution started in 2012 in April, was interrupted during the dry season (January- February) and starting again in March 2013.

Every two weeks, all distributors met their respective supervisor at the health center to report on the activities and replenish the stock for the subsequent bottle exchange round [18]. During the entire study period, all users and care takers (for children) were informed about the importance and the appropriate use of the repellent, and how to react in case of AEs.

To promote compliance, a series of health education campaigns were conducted each year. In 2013, more intensive and continuous campaigns were performed.

Results

AEs individually reported

Over the two years period a total of 41 individual AEs were individually reported by phone. 22 were classified as ARs and 11 as repellent abuse (Table 1 and Table 2). The remaining eight were judged not related to the repellent (details in S2 Table). All ARs were considered mild. Of the cases of repellent abuse, three were considered moderate and two as life-threatening.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. ARs notified by phone and their classification according to the terminology of the study protocol for reporting, management and treatment of AEs (N = 22).

https://doi.org/10.1371/journal.pone.0172566.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Repellent abuse (oral ingestion) notified by phone and their classification according to the terminology of the study protocol for reporting, management and treatment of AEs (N = 11).

https://doi.org/10.1371/journal.pone.0172566.t002

All 22 ARs occurred in six villages in 2012 in different families belonging to six villages (Fig 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Monthly counts of ARs and repellent abuses (oral ingestion), 2012–2013.

https://doi.org/10.1371/journal.pone.0172566.g003

Most cases were female having skin manifestations (see details in Table 1 and S2 Table). Three cases were children under or equal to ten years old (S2 Table). In fourteen cases, the event occurred during the first two weeks of application. Seventeen cases (77.3%) were advised by the study physician to permanently stop using the repellent. All cases completely recovered, of which 14 (63.6%) without treatment (Table 1). The total incidence rate per 1,000 inhabitants of ARs for all intervention villages was 0.84 in 2012 but ranged from 1.05 to 17.63 in the affected 6 villages (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Incidence rate per 1000 inhabitants of ARs notified by phone, 2012.

https://doi.org/10.1371/journal.pone.0172566.t003

There were five cases of repellent abuse in 2012 and six in 2013, involving nine villages (see details in Table 2, Fig 3 and S3 Table). On six occasions the oral ingestion was accidental, but on five it was considered a suicide attempt. All suicide attempts were women ranging from 13 to 26 years of age. Of the six accidents, four were male children between 2 and 7 years of age.

The mild and moderate cases typically presented with nausea, vomiting, headache and fatigue. One mild case involved a 4 months pregnant woman who attempted to commit suicide by drinking about 50 ml (a half bottle) of 20% picaridin. Fortunately, no negative impact on fetus and new born was noted. One moderate case consisted of convulsions in a 25-year old lady with a history of childhood epilepsy. The two life-threatening suicide attempt cases, in girls aged 19 and 13 years old, were unconscious for about two hours after drinking 200–300 ml (two to three full bottles). Besides unconsciousness, the two patients always had normal vital signs recorded in the health facilities. The treatment of the first life-threatening case included administration of intravenous fluids and atropine. The second life-threatening case was treated in the village with administration of intravenous fluids, cortico-steroids and antibiotics. After treatment, all cases completely recovered. The total incidence rate of repellent abuse per 1000 inhabitants was 0.2 in both years for all intervention villages, while the incidence for villages ranged from 1.9 to 3.5 and 1 to 5.7 in 2012 and 2013 respectively (Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Incidence rate per 1000 inhabitants of repellent abuses, 2012–2013.

https://doi.org/10.1371/journal.pone.0172566.t004

Perceived side effects reported by the families to the distributors during routine visits

More than 95% of all families received the repellent at least once (Table 5). Around 20% of the families receiving picaridin reported perceived side effects each year, half of them occurring within the first two to three months of use (Table 5 & Fig 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Monthly trends of families reporting perceived side effects and average of the actual monthly repellent consumption.

https://doi.org/10.1371/journal.pone.0172566.g004

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Families reported perceived side effects and repellent consumption through distributor reports, 2012–2013.

A percentage above 100% of families receiving repellents in 2013 is due to new arrivals after the population census done early 2013 were given repellent and recorded but were not updated in the census data.

https://doi.org/10.1371/journal.pone.0172566.t005

Of those families, most (68.6%) reported side effects only once in 2012, while in 2013 about half reported two times or more. In both years, families reporting perceived side effects consumed less repellent. In 2012, itching/irritation/rash (12.6%), headache (8.8%) and dizziness (7.6%) were the top three reported side effects. In 2013, headache (11.6%), bad smell (10.2%) and itching/irritation/rash (10.1%) were most common (Table 5).

Discussion

This study assessed the safety of the mass introduction of a topical insect repellent in rural communities as part of a public health malaria control intervention. An extensive reporting system was put in place to detect AEs associated with the use of the repellent, applied twice daily for an extended period of time. On a total of about 25,000 individual repellent users from the intervention arm, only 22 –less than one in thousand—ARs related to cutaneous repellent use and 11 cases of repellent abuse were reported.

The ARs related to topical use of the repellent occurred during the first few months after the introduction of the repellent in the population, and mostly during the first two weeks of use (Fig 3). Contact or irritant dermatitis is indeed to be expected within the first days or weeks of use, and individuals with such reactions generally discontinued repellent use. Inappropriate repellent use during the early stages of the project might have contributed to some of the ARs. As observed during field visits, some individuals continued, sometimes encouraged by their distributor to use the repellent despite skin reactions, and this apparently led to progressive worsening. Only 326 (5.6%) of families were recorded by the distributors to discontinue the use of the repellent during the trial.

Repellent abuse (either accidental or suicide attempt) is another concern for community level interventions. Children and particularly young boys were at risk for accidental ingestion of the product. All suicide attempts occurred in young women, which is in line with the fact that within the region, women are more likely to attempt suicide and to use less-violent methods [20, 21]. The insect repellent DEET has been associated with neurotoxic effects such as seizures, although causality was difficult to demonstrate [22–24]. As to picaridin, no neurotoxicity or lethal effect have been reported in animal experiments and individual human experience [8, 25]. Our findings demonstrate that hallucinations and unconsciousness can occur after ingestion of high amounts of picaridin (2–3 bottles, 200–300 ml), but reassuringly this resolved completely with supportive treatment. There was one case of intentional abuse in a pregnant woman, who delivered a healthy baby. In animal experiments, maternal effects such as increased liver weights were observed with application of picaridin, but no toxic effects on the fetuses was detected [25].

In 2013, there was an intensification of the health promotion campaigns, and concurrently more perceived side effects were reported. Similarly, within each year, periods with more repellent use appeared to be associated with more perceived side effects and families reporting side effects also had a lower average repellent consumption. We note however that the increased reporting of side effects during the second year was largely explained by increased reporting of headache, dizziness and bad smell, while the frequency of skin manifestations remained fairly stable across the two years. Importantly, there were no indications of cumulative toxicity in our study, whereby side effects would have started emerging only after prolonged use. This is reassuring for a malaria control intervention which implies daily administration possibly over a period of several years.

Bad smell associated with headache and dizziness, has been reported related to agrochemical products [26]. In a comparative study performed in Malaysia participants felt more comfortable with the odor of picaridin as compared to DEET [27]. As preferences in smell can vary extensively geographically and across societies, adapting the repellent’s perfume to local preferences might be a way to improve adherence.

Our findings need to be interpreted in the specific context of the study. First, the reported findings were observed during a cluster-randomized trial, whereby due efforts were done to carefully inform users about the appropriate use and what to do in case of AEs, and whereby users were closely monitored for intolerance. We also note that we did not collect information on AEs and perceived side effects in the control arm, since no placebo was used. This makes it obviously difficult to causally link the spontaneously reported symptoms to the repellent use, and our findings are thus most likely an overestimation. Finally, the aggregation at the family level of data on perceived side effects precluded more detailed individual level analysis.

In conclusion, we found ARs during mass introduction of the picaridin insect repellent to be uncommon and generally mild. All the 22 ARs reported, amongst over 25,000 individual users occurred during the first few months of the study. Oral repellent abuse was also uncommon but occurred throughout the two years. Hence, control programs employing mass repellent use should be particularly prepared to detect and manage ARs at the time of program onset and cases of repellent abuse remain throughout. While one in five families reported any side effect, few families discontinued repellent use. Our study supports the safety of the picaridin insect repellent for community wide use as part of a malaria control intervention.

Supporting information

Acknowledgments

This study was funded by the Bill and Melinda Gates foundation Global Health Grant OPP1032354. The authors would like to thank all partners and stakeholders involved in the project. Specifically for this study, we thank Johnson & Johnson for the donation of the picaridin repellent; the National Malaria Program for organizing the importation and storage of the repellent; the provincial health department of Ratanakiri for the storage of the repellent; the survey team and especially their drivers who played a key role in getting the repellent where needed; all the village distributors & supervisors for their hard work; all the participating villagers for their efforts; we thank all health centers and Ratanakiri provincial hospital staff for active involvement in the project and the management of any safety events. We are grateful to the Data Safety Monitoring Board for their vital work during the study. Special thanks also to Prof. Lutgarde Lynen for her advice throughout the project.