Comparison of phenothrin mousse, phenothrin lotion, and wet-combing for treatment of head louse infestation in the UK: a pragmatic randomised, controlled, assessor blind trial

Participants

We recruited participants (children and adults) to the study from respondents to an information letter distributed through schools or via general practitioners associated with Bedfordshire Health. Prospective participants or their parents/carers telephoned the study co-ordinator to make an appointment for a home visit by a trained agency nurse. Most visits were within 24 hours unless requested at a different time. Nurses followed a standard approach to check for living lice using a plastic detection comb (Albyn of Stonehaven Ltd, Stonehaven, Scotland). If moving head lice were found, and the individual was 4 years or over, they were invited to join the study. Prospective participants were conducted through a standard consent procedure in which the content of information sheet was explained verbally and the parent/carer confirmed that they understood the function, processes, and commitments of the study before signing the consent form, which was witnessed by an independent adult. Participants were individually assigned a randomised treatment. All other household members were offered examination and, if found to be infested with head lice, were given the opportunity to join the study.

Prior to inclusion in the study, all participants provided baseline data including: age, gender, ethnicity, hair characteristics including length, thickness, degree of curl, and previous pediculicide use. Inclusion criteria required availability for up to 28 days to accommodate each of the possible treatment regimens and a suitable adult available to perform or assist in application of the treatment. Candidates excluded from this study were: pregnant or nursing mothers; anyone who had bleached, colour-treated or permanently waved their hair; used pediculicide, had been treated with antibiotics, or had participated in a clinical trial during the 4 weeks prior to this trial. In addition anyone with sensitivity to any pyrethroid insecticide or chrysanthemums; receiving treatment for asthma; suffering from a persistent skin disorder of the scalp (other than head lice); or had already participated in this study; was also excluded.

Household members with lice who either did not wish to participate or who failed to satisfy the inclusion criteria were given advice about appropriate treatment methods. Any participants found with lice after completion of the study period were supplied with 0.5% malathion lotion (Suleo-M lotion, Seton Healthcare Group plc, Oldham, UK) in conformity with the Bedfordshire Health policy for treatment. No payment was offered for participation.

Ethics

Ethical approval was granted by both the South Bedfordshire and North Bedfordshire Local Research Ethics Committees of Bedfordshire Health, one of the conditions of which was that the data would be published in the public domain. The study was registered with the Current Controlled Trials database (ISRCTN73201839). See Supplementary files for the study protocol.

The study was conducted in compliance with Good Clinical Practices, and in conformity with the principles of the Declaration of Helsinki and European Standard, EN540: Clinical investigation of medical devices for human subjects. Written and witnessed informed consent was obtained from the participants and parents or guardians of children under 18 years of age.

Treatments

The nurses explained how to apply the treatment to the parent or carer and also gave a printed copy of these instructions. The parent then applied the treatment in the presence of the investigator. The nurses were instructed to answer any questions, and to note these, but not to intervene if any error was observed during treatment. This was intended to represent the clinical situation.

We supplied one group of participants with 0.5% d-phenothrin mousse in 50ml butane pressurised containers, with a canula fitting to allow direction of the mousse during delivery (Full Marks Mousse, Seton Healthcare Group plc, Oldham, UK). This also contained citrate buffered water, ethanol, and emulsifying wax. We supplied the second group with 0.2% d-phenothrin water/isopropanol lotion in 50ml glass bottles with a dropper aperture (Full Marks Lotion, Seton Healthcare Group plc). Carers applied the products to dry hair to saturate the hair and scalp. We made available as many containers of product necessary to comply with the instructions. People using mousse washed it from the hair with shampoo after 30 minutes. Those treated with the lotion left it on the hair for 2 hours before shampooing. We supplied both groups with a non-medicated shampoo (L’Oréal Children’s Shampoo, L’Oréal (UK) Ltd, London, UK) for this. Both groups received a single application of treatment.

We supplied the third group of participants with a “Bug Buster” pack (Community Hygiene Concern, London, UK) for performing the wet-combing technique. We also supplied a bottle of the same non-medicated toiletry shampoo and four 60ml bottles of conditioner rinse (one for each treatment day) (L’Oréal Children’s Shampoo and L’Oreal Conditioner, L’Oréal (UK) Ltd,), and a diary card. Parents/carers were instructed to wash the participant’s hair with the shampoo, massage in a generous amount of conditioner, and comb through the hair systematically from scalp to tip with the louse removal comb provided in the pack. They wiped the comb on a paper towel between strokes. Combing was repeated at regular intervals for 2 weeks (days 0, 4, 8, 12 of the study). We asked carers to fix any lice found during the combing onto the diary card using clear cellulose adhesive tape and to record how much time was spent combing.

Outcome measures

We made mousse and lotion treatment follow ups on days 4, 7, 10, and 14 and those for wet-combing on days 14, 21, and 28 after commencement of treatment. Day 14 was used as the point for measure of primary outcome for the insecticide groups. However, in the wet-combing group it was possible there could be viable eggs present on day 14 and therefore follow up examinations were also conducted on days 21 and 28 to detect any emerging nymphs or other lice missed at day 14. During each of the follow up assessment examinations, any lice found by detection combing were removed and fixed to the case record using clear cellulose adhesive tape. These were later examined in the laboratory to determine the gender or development stage of each insect.

Nurses collected the containers of mousse and lotion after treatment so that the quantities used could be measured. Bottles of conditioning rinse from the wet combing group were also collected for measurement after the final assessment on day 28, although some of these were mislaid by participants.

Sample size

We estimated sample sizes to show a difference between wet combing and phenothrin lotion treatment with 95% confidence, 90% power, and equivalence between the two phenothrin groups to within 20%. For this calculation we conservatively estimated that the phenothrin products would exhibit approximately 80% effectiveness and wet combing 50% success. Sample size calculations were made by the sponsor’s consultant statistician who estimated a minimum sample size of 104 participants in each phenothrin treated group and 58 participants treated with wet combing (266 evaluable participants) would satisfy this probability with greater than 90% power.

Randomisation and blinding

A computer generated list, prepared by the sponsor’s statistician, was used for randomisation of treatments, made up of balanced blocks of 133 treatment allocations with a relative frequency for each of the treatments of 52:52:29. Randomisation was by individual so that different members of a household could receive different treatments. Nurses involved in recruitment were supplied with envelopes in batches of ten and asked to issue them sequentially. Investigators, who were unaware of which treatment had been used, made follow up examinations using plastic detection combs to check for the presence of living lice. On day 4, 7, and 10 these assessors were a different group of agency nurses. On days 14, 21, and 28 the assessing investigators were from the Medical Entomology Centre (IFB and CMB).

Statistical analysis

We analysed for differences between groups based on the intention to treat (ITT) population and tested equivalence only using the per-protocol (PP) population. We calculated differences in cure rate using a chi-squared test and equivalence to within 20% based on the 95% confidence limits derived from the normal approximation to the binomial distribution. The initial analyses for the sponsor were performed by the consultant statistician (PN Lee Statistics and Computing Ltd, Sutton, UK) using bespoke software. Post hoc analyses performed by the investigators employed Epi-Info version 6, OXSTAT II version 1.11, and purpose built spreadsheet calculators. Differences between groups in baseline characteristics, safety, acceptability, and efficacy were tested using Fisher’s exact test for yes/no variables and the Mann-Whitney U test for ranked variables.

Tests for resistance

During the course of the study a high level of treatment failure was observed in the insecticide groups. We collected samples of lice at the final assessment from five participants from different parts of the study area for a bioassay evaluation for sensitivity to phenothrin. The insects were placed on treated or control filter papers using a method previously described for tests of permethrin sensitivity³. Each treated filter disc was impregnated with 500 µl of 2% d-phenothrin solution, giving an insecticide deposition rate of 157 μg cm^-2. The mortality outcomes of the tests were compared, with a baseline sensitivity obtained using laboratory-reared, insecticide sensitive, body/clothing lice by means of log-probit analysis using LDP Line software.

Participants

The study was conducted between June 1997 and March 1998, during which informed consent was obtained for 228 children and 50 adults to participate (Figure 1). Two people were excluded from further analysis as, upon inspection, no live lice had been found. The recruitment case record form for one other participant was lost so this case was also excluded from the study.

Figure 1. Flowchart of participants through the study.

The 275 participants were randomly assigned to one of the treatment groups: 105 received phenothrin mousse, 107 phenothrin lotion, and 63 were allocated to the wet-combing treatment. Of these participants, 246 (89.5%) (100 mousse, 100 lotion, and 46 wet-combing) completed the trial with adequately complete follow-up data sets (Figure 1 and Dataset 1). From the original study group, 5 participants (2 treated with lotion and 3 wet-combing) were non-compliant and excluded from the per-protocol analyses. Non-compliance involved additional combing or other unauthorised treatments (2 receiving lotion and 2 wet-combing) and one participant on wet-combing shaved his head. There were 24 other withdrawals: 2 people on wet-combing dropped out; 20 were lost to follow-up (5 from the mousse group, 5 from lotion, and 10 from wet-combing); and 2 people from the wet-combing group were not allocated Day 21 or Day 28 appointments in error following a communication failure between the study coordinators and the investigators conducting final assessments. Two of those lost to follow up were due to bereavement and the two drop outs chose not to continue in the study. The rate of protocol violation/withdrawal was significantly higher (p < 0.001) in the wet-combing group than the phenothrin-treated groups.

Of the 275 people known to satisfy the inclusion and exclusion criteria, 62 (23%) were male (Table 1). The percentage distribution of males was similar for the two phenothrin groups, 18% and 21% respectively, but higher in the wet-comb group (33%). This difference in proportion of males between wet-comb group and mousse-treated group was statistically significant (p < 0.05) and the difference between wet-comb group and the lotion-treated group was nearly significant (0.05 < p < 0.1).

Outcomes

Post-treatment examinations at day 14 showed that there were 20/107 successful treatments and 3 cases of reinfestation after cure (an overall success rate of 21.5%) using phenothrin lotion and 18/105 successes and 3 cases of reinfestation for phenothrin mousse (giving 20.0% overall success). This made mousse users 1.07 times more likely to have lice after completion of the treatment (95% confidence interval (CI) 0.63 to 1.81; odds ratio (OR) 1.10, 95% CI 0.56 to 2.13). In the case of wet-combing with conditioner there were 12/63 (19.1%) successful treatments and no cases of reinfestation. Participants treated with combing were, therefore, 1.13 times more likely to have lice (95% CI 0.61 to 2.11; OR 1.16, 95% CI 0.53 to 2.54) than if they had been treated with phenothrin lotion. People treated with wet-combing were also 1.05 times more likely to have lice than those participants receiving phenothrin mousse (95% CI 0.56 to 1.99; OR 1.06, 95% CI 0.48 to 2.34).

In both insecticide treated groups the majority of lice at post-treatment assessments were juveniles, of which 712 first were stage nymphs that could only have originated from eggs not killed by insecticide. However, it was not possible to properly analyse the full effect of ovicidal failure due to participants being withdrawn early in the study on grounds of lack of efficacy. Nevertheless from the data available it was possible to determine that failure to kill louse eggs was a major contributing factor in the low rate of outcome success.

Intention to treat analysis found no statistical difference (p < 0.05) between the lotion and mousse. Similarly, no statistical difference was found between wet-combing and either of the insecticides. Success in curing the infestation was also not significantly associated with gender or hair type, thickness, or length, after adjustment for any randomisation anomalies. However, success rates were significantly (p < 0.01) higher in people who had previously used a head louse treatment successfully, by an estimated factor of 1.88 (95% CI 1.13 to 3.11). Success also significantly (p < 0.01) increased with age.

Analysis based on the per-protocol population, without taking into account the failure of randomisation, showed no significant difference in success rates between the two phenothrin groups (lotion 17.8%, mousse 13.7%), which showed equivalence to within 20% (mousse-lotion difference, -4.1%, 95% CI -6.1% to 14.4%).

Adverse events

We found several clearly defined treatment-related adverse events in people treated using the phenothrin products: 12 adverse events in 11 people using lotion (9 scalp irritation, 3 irritation of the respiratory system); 10 adverse events in six people treated with mousse (5 scalp irritation, 3 dry skin, 1 bullous reaction, 1 paraesthesia of the scalp). There were no similarly defined adverse events for the wet-combing group but five carers reported children expressing signs of stress while being combed, one person reported discomfort during combing, and backache or arm/shoulder aches for three carers were also reported, but not formalised as reported adverse events. All events were considered mild and resolved rapidly except for two cases. The case of paraesthesia, which was classified as moderate, persisted for two days after treatment, and one case of dry skin persisted for some time after treatment but may have been an exacerbation of a pre-existing problem. Stinging of the hands and paraesthesia-like reactions were also reported by some of the carers while applying the phenothrin-based products. Paraesthesia has been reported from use of other pyrethroid preparations^8–10 and would likely be exacerbated by the presence of alcohol in the product.

Tests for resistance

Lice from different participants showed marked differences when tested for sensitivity to phenothrin (Table 2). In all cases the lice were taken from people who had experienced treatment failure during the study so it was not surprising that the majority had insects that were resistant to the insecticide. However, one person had apparently also been reinfested with sensitive lice from a contact, as shown by the mixed sensitivity of the insects. All lice from another participant, treated using wet-combing, were phenothrin susceptible. Output data from the LDP Line analyses of the observations of head lice, in comparison with susceptible laboratory reared body lice, showed a resistance ratio (RR) of 54.74 when exposed to the insecticide. From the log-probit analyses the estimated time required to kill 50% of the insects (LT₅₀) was 502 minutes based on a mortality curve with a slope of 1.0096 ± 0.1324 (chi-squared 19.9681, p = 0.0005). In contrast the estimated LT₅₀ for the body lice was 24.74 minutes (slope = 5.1932 ± 0.5086, chi-squared = 11.6217, p = 0.0404). Log-probit analysis also suggested that the insecticide sensitive head lice (LT₅₀, 44 minutes; LT₉₅, 95 minutes) were approximately twice as tolerant of phenothrin as laboratory reared lice, although the number of insects involved was too small to provide a clear distinction.