Reason for withdrawal from publication
The Editor withdrew this review as of Issue 4, 2006. This review will be reinstated following a substantive update.
The editorial group responsible for this previously published document have withdrawn it from publication.
Feedback
CHC 'Bug Buster'
Summary
Community Hygiene Concern (CHC) has many years' experience of working to solve the head louse problem, supporting parents to deal with treatment failures, re‐infestation, confusion and frustration. The social stigma associated with head lice still persists and there is growing anxiety about the cost and health implications of repeated treatment applications.
As a public interest charity, we welcomed the announcement that the highly respected Cochrane Collaboration would review interventions for treating head lice. We had high expectations that full examination would be made of the best evidence available on the natural history of Pediculus capitis, the epidemiology of pediculosis capitis, the reliability of different detection methods, the validity of treatment instructions/claims and the sustainability of control strategies. We anticipated the resulting review would be helpful in providing support, an informed choice, and assist in the monitoring of outcomes. CHC has studied the full Cochrane review, as published electronically (Dodd, 1999, 2001a) and an editorial based on its findings published in the BMJ (Dodd 2001b). We have been so disappointed that it is our intention to campaign for a new review. We find this body of work, which concentrates on the UK situation, is shallow, unbalanced, and comes to misleading conclusions.
There is a well‐known paucity of randomised controlled trial evidence to support treatment protocols. In the circumstances, we expected that this review would carefully distinguish between speculation, expert opinion and proven fact, but its recommendations are based on a trial rejected as seriously flawed by an earlier systematic review (Vander Stichele et al, 1995). It omits key papers which provide evidence on the appropriate diagnostic techniques for different circumstances, and clinical evaluations of treatments which has led to a false conclusion that no insecticide has been shown to be more effective than another. In the case of the trial by Roberts et al (2000), Roberts himself has protested at the Cochrane misinterpretation of the trial results (bmj.electronic response, 2001). Moreover, Dodd does not call for implementation of the Medicine Control Agency ruling in February 2000 that 'natural' treatments must pass safety and efficacy licensing tests to trade legally.
Contrary to Cochrane principles there has been no consumer participation in this review and the review author declares a serious potential conflict of interest. In our view, it damages the gold standard reputation of the Cochrane reviews in general and should be withdrawn.
We feel that an independent, suitably qualified panel should undertake a new Cochrane review.
Reply
REPLIES TO SPECIFIC COMMENTS
Comment 1 This review accepts, with no adequate explanation, the RCT conducted by Burgess et al. (1994) which was rejected by Vander Stichele et al. (1995) as seriously flawed.
Reply The review by Vander Stichele et al (1995) was criticised in other publications for choosing inappropriate quality criteria by which to evaluate trials (Burgess 1995, Stallbaumer and Ibarra 1995). Some of the trials included in the Vander Stichele review, upon closer inspection, were found to be seriously flawed in their methodology. The authors agreed that a subsequent Cochrane Review (Vander Stichele and Bogaert 1995)was appropriate.
The trial by Burgess et al (1994) was excluded by Vander Stichele because it failed to meet a number of the quality criteria set for the review, and according to the authors had a high risk of bias. According to Vander Stichele, the trial by Burgess fails on the following criteria; randomization procedure, concealment allocation to patients, definition of exclusion criteria, ethical procedures, appropriateness of conclusions, documentation of prior exposure to pesticides, history of prior head lice treatment, definition of current head lice infection, prevalence of lice in study area, contact tracing, documentation of: pediculicidicity, residual activity and cure rate. In fact the majority of this information is detailed in the paper or was provided by the author when the trial was being assessed for inclusion in the Cochrane Review.
To answer each point in turn: Patients were individually randomised into the test groups, but details of allocation concealment were not given. However as two different pediculicides were used, each of which has a different formulation, with different application procedures, it is impossible to conceal which treatment is being provided to which patient. Patients were allocated to treatment groups in blocks of three, 1 permethrin: 2 pyrethrin. Appropriate exclusion criteria were chosen and reported in the trial. Although prior exposure to pesticides was not documented, the study was carried out in an area where no products were commercially available. Participants were only included in the trial if they exhibited a current infection with lice determined by observing live lice on the head/scalp. The study was conducted on two sites where the prevalence of lice was 75% and 100%. Contacts were not treated or traced in this study. Participants were followed up at 2,4,6,8,14 days post treatment and assessed for the presence of live lice by detection combing. Overall cure rate was determined at each follow up, but the number of participants assessed at each follow up, except the first, was poor and a direct comparison between the treatment groups could not be made at day 14 as no data was available for one of the groups. Residual activity of the preparations was not considered, although permethrin has been shown to exhibit some residuality. However, most insecticides do not have a residual action, which according to Vander Stichele's inclusion criteria would dismiss the majority of compounds from being considered effective pediculicides. However, in the long term it is not advantageous to use a product with a residual activity, as the residuum slowly diminishes to sub‐lethal levels that can lead to the development of resistant strains of lice (Burgess 1995), which now have widespread occurrence.
Despite there being some difficulties with the trial by Burgess et al (1994), the design met with the inclusion criteria chosen for the Cochrane Review and a detailed discussion of the appropriateness of these criteria can be found in the review. The quality criteria by which Vander Stichele judged the trials are not necessarily relevant to addressing the effectiveness of a product for the curative treatment of head lice. It is clear by the discrepancies between their assessment of the Burgess paper and that made in the Cochrane Review, that their judgement of the trials is doubtful and inconsistent as has been highlighted in previous reports (Burgess 1995, Stallbaumer and Ibarra 1995).
Comment 2 The clinical evaluations by Downs et al.(1998, 1999) made in Bath and Bristol are completely omitted from this review. These show an 87% failure rate for pyrethroids and a 64% failure rate for malathion. The lapse leads Dodd to the false conclusion that no insecticide has been shown to be more effective than another.
Reply Neither of the papers listed above by Downs et al (1999a,b) are RCTs. Downs et al (1999a) assesses the prevalence of head lice in schoolchildren and looks at resistance in vitro. If treatment was applied to patients directly, the methodology is not detailed. Downs et al (1999b) tests the susceptibility of field collected lice for resistance to different pediculicides, in vitro, by exposing the lice to insecticide impregnated filter papers. In vitro studies were excluded from the Cochrane Review because the methods used and the results obtained are not indicative of the effectiveness of pediculicide when used in the field to treat people infected with head lice. In the Downs et al (1999b) study, a group of children were also treated with one of two commercially available pediculicides. However, one of the preparations was not used according to manufacturers instructions and neither treatment was reapplied 7 days after the initial application to account for the presence of nymphs hatched from eggs that were laid prior to the initial treatment. This is important as neither preparation is 100% ovicidal. Assessment of cure was conducted 24 and 72 hour post treatment and the presence of nymphs post treatment was not taken as treatment failure. This in itself is not incorrect as any first and second stage nymphs found up to one week following treatment are hatched from surviving eggs. Adults and third instar nymphs can only be from reinfection unless there is complete treatment failure, when they should have been found at the first follow up investigation (Burgess 1995). However, even if the product is not reapplied after 7 days, a final assessment at 14 days post treatment should be taken as the definitive evaluation of pediculicidal and ovicidal activity, as eggs that were not killed by treatment can take up to ten days to hatch (Taplin et al 1986).
It is true that in the field, some pediculicides are more effective than others, with treatment success being reflected by local resistance patterns. However, there are currently no RCTs of suitable quality which address these questions and therefore the results cannot be included in a meta‐analysis to determine which products are the most effective. In order to provide a meaningful answer to this question, a nation‐wide study would have to be conducted, looking at the different local resistance patterns across the country.
The results obtained by Downs et al (1999a,b) could be incorporated into the background section of the review, but the papers do not meet the inclusion criteria to be included in the analysis.
Comment 3 In 1999, the Bug Busting method is inaccurately described and dismissed as "unfeasible" for treatment in advance of the publication of RCT evidence. No consideration is given to its value as a detection method. In 2001, no change has been made inspite of publication in the interim of a review calling for the standardization of detection methods for diagnosis and epidemiological purposes (Figueroa, 2000), a pilot trial which provides evidence on the appropriate diagnostic techniques for different circumstances (Bingham et al, 2000) and an observational study (De Maeseneer et al, 2000).
Reply In the outcome measures of the review, it is stated that pediculicidal effect should be determined by detection combing, therefore further discussion of detection methods is not relevant for the review.
In answer to the second point, the Bug Busting method of louse control requires the hair to be wet combed with conditioner for at least 30 minutes every 3‐4 days using a fine toothed louse detection comb. The use of this method as a control measure was described as unfeasible because "too few people have sufficient motivation or skill" to implement the method effectively and success is most likely "if an infection is found in the earliest stages when few lice are present" (Burgess 1997). Roberts et al (2000) report that Bug Busting requires significant time and effort for it to work as a treatment for head lice. They suggest that when children that had used pediculicide in the 4 weeks before entry on to their trial were excluded from the analysis, the cure rate in the Bug Busting group increased. They propose that an explanation for this could be that children in this subgroup were less likely to report recurrent treatment failure and less socially disadvantaged, and therefore more able to use Bug Busting successfully. They continue by suggesting that the results of their trial indicate that policies advocating the use of Bug Busting as a first line treatment for head lice infection are inappropriate for the general population. The highlighted difficulties with compliance for this treatment method support the notion that Bug Busting is an unfeasible first line treatment method for head lice. That does not mean however, that Bug Busting is unsuitable under all circumstances. It may still be used in cases where all else has failed, or in families who do not wish, for what ever reason, to use insecticides, in which case motivation to comply and persevere with treatment is likely to be high.
It is also worthy of note that the diagnostic methods used in the Bingham study were not those "approved" by CHC, i.e. they first used parting the hair and looking, and then used dry combing alone. However, CHC only approve wet combing in their literature, and criticised dry combing, because lice move rapidly away from disturbance in dry hair (Stallbaumer and Ibarra 1995), it allows the build up of static electricity and causes pain if sore patches are present (Ibarra 2001).
Comment 4 In 2001, Dodd includes a RCT conducted by Roberts et al. (2000) in North Wales comparing Bug Busting (using the first, more laborious Bug Buster comb) with two applications of malathion a week apart which recorded a 38% and 78% success rate respectively, but inaccurately reports that it found that Bug Busting is ineffective as a curative measure. Roberts himself has protested at the misinterpretation (bmj.electronic response, 2001).
Reply When choosing a product/method for treating a head lice infection the cure rate is an important outcome measure to consider. The desired outcome of treatment is complete cure of infection, as continuance of infection is possible with only one gravid female or a single male and non‐gravid female. It is therefore a matter of interpretation whether one considers a treatment method which has been shown to have a cure rate of 38% to be effective and as described by Roberts et al (2000) "inappropriate for the general population" as a first line treatment against head lice. Downs et al (2001) criticised as "unethical", Roberts' use of malathion lotion knowing moderate resistance to be present in the study area. If the use of a proven treatment, achieving a 78% treatment success rate is to be considered unethical, then surely the situation pertaining to the use of an untried and apparently unsuccessful combing method should, by these standards, be totally unethical or at the very least of dubious ethics. Roberts himself, in his BMJ response (2001), lists the Bug Buster Kit to have poor effectiveness against head lice, together with the pyrethroid pediculicides Full Marks and Lyclear, products which are not usually considered effective for treating head lice.
Comment 5a (5) Correspondence about the trial has been ignored, but these letters raise many important points relevant to the sensible use of resources, including:
(a) that the N Wales study found Bug Busting 2% more effective than malathion was proved in Bath and Bristol (Hill, 2000)
Reply The results from the Bath/Bristol study (Downs et al 1999b) were based on a trial involving 16 participants (Bristol) and 14 participants (Bath). In order to be able to detect a 20% difference between treatments at 80% power with 95% confidence you need around 60‐65 per group using conventional clinical statistical methods. In fact, with the exception of Taplin's two placebo controlled studies, all RCT's currently available are underpowered to produce the results required. The results of the Bath/Bristol investigation (Downs et al 1999b) are only pertinent for the locality from which the lice were taken. It is known that resistance patterns vary across the country and even within a locality. If a pediculicide has a cure rate of 36% it should surely be deemed ineffective for the curative treatment of head lice within that locality, as the outcome of treatment should be complete cure of all the individuals treated. In the Taplin et al (1982) trial, the cure rate at day 7 was 44.68% in the control group, in which treatment was with the vehicle alone.
Comment 5b (b) that skill in managing Bug Busting grows with familiarity, whereas exposure of the louse population to repeatedly applied insecticide increases insect resistance (Ibarra et al, 2000)
Reply Burgess (in Gross 2001) states that wet combing with conditioner results in foam and slime making it difficult to see the lice and in addition, components such as panthenol and surfactants in conditioners may lead to adverse reactions in some patients. In the same article he commented on wet combing, saying that the problem is not that the method does not work, but that people do not know how to do it. "You need to appreciate what you are doing and why and to continue until you find no more lice," he said. He recommended that, if necessary, wet combing needs to be continued for up to two hours at a time, depending on the number of lice found and the thickness of the hair, and carried out four times over a period of at least two weeks. This period may need to be extended until all lice have been removed.
The prophylactic use of insecticides should not occur as it will not prevent infection and may promote resistance and increases the risk of toxicity. The use of repellents is not recommended for the same reason (Aston et al 1998). Treatment with a pediculicide should only be undertaken if live lice are found. It is recommended that pediculicides (and combing) should be used in a mosaic strategy. This involves applying a particular insecticide/treatment and repeating application seven days later. If live lice are still found then another insecticide from a different chemical class is used and so on. This method avoids continual repeated use of a single product and is designed to delay the onset of resistance. However, it is inevitable that insects will eventually develop some degree of resistance to insecticides.
However, some of the currently unlicensed products that appear to work contain nothing more than conditioning agents found in conventional hair conditioners e.g. Nice n' Clear. This product appears to kill lice just by surfactant activity over a relatively extended application time, in some cases not much exceeding that used for Bug busting (Burgess pers comms 2002). It is possible that quite a lot of the "activity" observed by CHC when Bug Busting was introduced, was due to the killing effects of conditioner as much as the effects of combing, but over time, however, strains of lice more tolerant of the conditioning agents have been selected and so Bug Busting is likely to be less effective now as a result (Burgess pers comms 2002). We can see this in the use of Nix/Lyclear. Taplin's placebo controlled study found a reduction of cases in the placebo group at the first assessment even using only 10 minutes application. However, these days the lice are more tolerant of both the permethrin and the vehicle so these effects are no longer experienced (Burgess pers comms 2002).
Comment 5c (c) that the N Wales trial, in common with all others cited in the review, lacks information on long‐term outcome (Richards, 2000) which Dodd does not remark upon; this is vital when families face multiple episodes of infestation
Reply It is believed that time spent curing an individual is wasted unless infectious contacts are also treated and traced (Dodd 2001). By doing this, the risk of reinfection to the patient is reduced, as will be the degree of transmission of lice on a wider scale. As most infections have existed for weeks rather than days before they are detected, contacts over the last month should be traced (Dodd 2001). Contact tracing is essential to prevent reinfection. All proven cases should be treated at the same time with the same insecticide (Aston et al 1998). Prevention is best achieved by regular detection combing and early intervention, although in practical terms the solution lies as much in education as in treatment. With regard to the long‐term implications of resistance management, see the comments above. In cases of consistent treatment failure and/or multiple episodes of infestation, Bug Busting it not necessarily the 'cure‐all' that it claims to be, some families are using it almost constantly and not achieving any effect (Burgess pers comms 2002). We should not forget that resistance operates at two levels, lice and humans!
Comment 5d (d) the judgement that it is unethical to sell products for use on children where widespread insect resistance has been demonstrated (Downs et al, 2000b) on which Dodd gives no opinion.
Reply See previous comments.
Comment 6 Review of key materials that influence parental choice has been omitted. This divorces the review from reality. There is no acknowledgement that pharmaceutical companies advertise and market to the general public their pyrethroid preparations, claiming instant success ‐ 10 minutes for permethrin (Warner Lambert, 2000) and 30 minutes for phenothrin (SSL International, 2000) ‐ and all their products, with the exception of shampoo formulations, state that a single application kills not only lice but their eggs.
Reply This is fair comment, but the review was not designed to be a critical appraisal of the marketing strategies of pharmaceutical companies, rather an objective and impartial review of the best evidence relating to the effectiveness of the available treatments. The review states:
"The way in which these insecticides are formulated into the marketed products may have a pronounced effect on their efficacy and acceptability. In some cases the vehicle for the pediculicide may also have some degree of pediculicidal effect which will enhance the performance of the insecticide. This must be taken into account when comparing products, which may have the same active ingredient (insecticide), but different formulation. Formulation can also affect the acceptability of the pediculicide to the public. The products which become the most popular and hence used most frequently, are those which in general tend to be the easiest to use, with the shortest application time and the most pleasing cosmetic attributes (such as smell). This can lead to single product domination of the market which in turn may mean an increased rate of resistance development."
The review also recommends "a mosaic model in which the same product is used for a course of treatment (2 applications spaced 7 days apart) and then a different insecticide from another resistance group is used for a course of treatment if the first insecticide fails."
It should be noted that until a year or so ago advertising directly to the public was not permitted and therefore would not have had a direct influence on public choice. It is true that what they might be seeing currently is potentially biased, but the actual wording of the claims is relatively vague and usually implies that the products "help" to get rid of lice (Burgess pers comms 2002).
Comment 7 Dodd fails to offer guidance to the licensing authorities to prevent history from repeating itself. It was clear when resistance was demonstrated to the first generation of modern insecticides, DDT and lindane, (Maunder,1971) that provision should be made in the licences of new compounds for the inevitable development of resistance. The point was not taken up when malathion, carbaryl or the pyrethroids were launched and we are suffering the consequences now.
Reply This is outside the scope of the review.
Comment 8 This review ignores work on a likely new insecticide for head lice, imidocloprid, (Downs et al, 2000a) belonging to the neonicotinoid group. These compounds are predicted to have only 5 years useful life (Hill, 2001).
Reply The imidacloprid study (Downs et al 2000) is not a RCT and therefore is not suitable for inclusion in the review. It is not the intention of the review to speculate about the development of new compounds for use as pediculicides, only to review the evidence for the effectiveness of currently available treatments.
Comment 9 Dodd does not call for implementation of the Medicine Control Agency ruling in February 2000 that 'natural' treatments must pass safety and efficacy licensing tests to trade legally.
Reply This is outside the scope of the review.
Comment 10 Contrary to Cochrane principles there has been no consumer participation in this review. If CHC had been consulted, we would have suggested that a main objective should be determination of licensing regulations that protect a vulnerable public from expensive, ineffective and possibly unsafe products.
Reply We plan to add a second reviewer for future versions, and this could be a consumer representative.
Comment 11 Serious potential conflict of interests In the review Dodd declares a potential conflict of interest because she has been a research assistant at the Medical Entomology Centre involved in clinical trials of treatments for head lice run by Burgess (results unpublished to date). A second serious potential conflict of interest is declared in the BMJ editorial, which is the author's receipt of a grant from Warner Lambert, the pharmaceutical company which manufactures the permethrin product with a third share of the UK market.
Reply The only conflict of interest that may be relevant to the review is the one stated within it. The research grant from Warner Lambert was paid to the Medical Entomology Centre for a series of literature reviews evaluating potential active ingredients that could be used in pediculicides. As I have no knowledge of the outcome of the companies decision, or influence over it, then I do not think that this affects my impartiality in any way. None of the work done for Warner Lambert had any relation to their existing products. My involvement in follow‐up investigations for clinical trials whilst working at the Medical Entomology Centre does not affect my impartiality as I had, and still have, no involvement in the outcome of the trials and no knowledge of the treatments that the individual participants received. I have no contact with any pharmaceutical company and now work in a completely different area and therefore I will not be involved in any trials etc. in the future.
References
Aston R, Duggal H et al 1998. Head Lice: Report for Consultants in communicable Disease Control (CCDCs). The Public Health Medicine Environmental Group Executive Committee.
Burgess, IF; CM Brown and NA Burgess. 1994. Synergised pyrethrin mousse, a new approach to head lice eradication: efficacy in field and laboratory studies. Clinical Therapeutics, 16(1): 57‐64.
Burgess, IF. 1995. Clinical efficacy of treatment for head lice: Authors differ on assessment of flaws in trials. British Medical Journal, 311(7016): 1369‐1370.
Burgess, IF. 1997. The management of head lice infections. Surgery OTC Review, 4(5): 15‐17.
Downs, AMR; KA Stafford and GC Coles. 1999. Head Lice: Prevalence in schoolchildren and insecticide resistance. Parasitology Today, 15(1): 1‐4.
Downs, AMR; KA Stafford and GC Coles. 1999. Evidence for double resistance to permethrin and malathion in head lice. British Journal of Dermatology, 141: 508‐511.
Downs, AMR; KA Stafford and GC Coles. 2000. Susceptibility of British head lice, Pediculus capitis, to imidacloprid and fipronil. Medical and veterinary Entomology, 14:15‐17.
Gross, Z. 2001. Pharmacy‐based head lice management. The Pharmaceutical Journal, 267(7164): 317.
Richards SM. 2000. Treatment of head lice. The Lancet, 356: 2007.
Roberts, RJ; D Casey; DA Morgan and M Petrovich. 2000. Comparison of wet combing with malathion for treatment of head lice in the UK: a pragmatic randomised controlled trial. The Lancet, 356: 540‐544.
Stallbaumer, M and J Ibarra. 1995. Clinical efficacy of treatment for head lice. Counting lice be visual inspection flaws trials' results. British Medical Journal, 311(7016): 1369.
Taplin, D; PM Castillero; J Spiegel; S Mercer; AA Rivara and L Schachner. 1982. Malathion for treatment of Pediculus humanus var capitis infestation. Journal of the American Medical Association, 247(22): 3103‐3105.
Vander Stichele, RH and MG Bogaert. 1995. Authors reply. British Medical Journal, 311(7016): 1369‐1370.
Vander Stichele, RH; EM Dezeure and MG Bogaert. 1995. Systematic review of clinical efficacy of topical treatments for head lice. British Medical Journal, 311(7005): 604‐608.
Contributors
Community Hygiene Concern (CHC) Distributors of 'Bug Busting' products January 2002
What's new
| Date | Event | Description |
|---|---|---|
| 9 November 2008 | Amended | Converted to new review format. |
History
Protocol first published: Issue 3, 1998 Review first published: Issue 2, 1999
| Date | Event | Description |
|---|---|---|
| 20 August 2006 | Amended | Review withdrawn as of Issue 4, 2006 |
| 30 November 2004 | Amended | Date new studies found but not yet included/excluded |
| 22 April 2002 | Amended | Response to comment/criticism added |
| 26 January 2002 | Amended | Comment/criticism added |
| 28 February 2001 | New citation required and conclusions have changed | Substantive amendment |
Sources of support
Internal sources
Medical Entomology Centre, UK.
External sources
Department for International Development, UK.
European Commission (Directorate General XII), Belgium.
Liverpool School of Tropical Medicine, UK.
National Health Service R&D Programme, UK.
Withdrawn from publication for reasons stated in the review