Lice infestation is a problem in local communities, probably because reservoirs remain undetected. Wet combing (combing systematically through wet, well conditioned hair with a fine toothed comb) has been presented as a cheap, ecological, self sufficient, and feasible technique for diagnosis and treatment of head lice.1–3 Compared with traditional scalp inspection it uses five elements to make living lice more visible, to better distinguish them from dandruff, and to assess the maturity of the infestation: water, conditioner, a fine toothed comb, a systematic sweep of the scalp, and a magnifying glass (10×). However, its efficacy as a diagnostic tool and as a therapeutic intervention has not been proved; hence it is not evidence based.
Subjects, methods, and results
We did an observational study comparing detection of head lice using traditional scalp inspection and wet combing. After ethical approval had been obtained, all 260 pupils, aged 2-12 years, of a primary school in a socially deprived urban area in Ghent, Belgium, were invited for a screening test during a three day campaign to detect head lice in November and December 1999. We obtained informed consent from parents. All children at school during the screening period were inspected consecutively and independently by two teams of six trained screeners. The first team did traditional scalp inspection, the second team did wet combing. The results of the first screening team were not communicated to the children, the school staff, or the second screening team. All children found to have head lice by the wet combing technique were given a number of treatment options, which were to be given at home by parents. All children found to have head lice by either of the two methods were reinspected 14 days later.
Association between the results of the two screening techniques was obtained using the kappa statistic. The positive and negative predictive value of traditional scalp inspection (criterion validity) was estimated, using wet combing as the gold standard.4
We screened 224 children (99 (44%) were 2-5 years old and 92 (41%) were female). Forty nine children (22%) were found to have head lice with the wet combing method (of whom 17 (8%) had been found not to have lice using the traditional scalp inspection) and 175 (78%) were found not to have head lice (of whom 14 (6%) were said to have lice using the traditional inspection method) (table). These 14 children were reinspected after 15 days. One of them reported symptoms and was indeed infected. There were no spontaneous reports of infestation among children who were not found to have lice using either technique. Of the 49 children found to have head lice by wet combing (and treated using a variety of products or by combing), 53% no longer had lice at reinspection.
The point prevalence of lice measured with the wet combing method was 21.9% (95% confidence interval 16.5% to 27.3%). We found a poor association between the results of the two tests (κ=0.59, 0.46 to 0.72). Compared with wet combing, the positive predictive value of the traditional scalp inspection method is 0.70 (0.54 to 0.82) and the negative predictive value is 0.90 (0.85 to 0.94).
Comment
Traditional scalp inspection is a poor technique for detecting head lice, as 30% of its “positive” results and 10% of its “negative” results are false (provided that wet combing is indeed the best method of detecting head lice). High values for false positives and false negatives call into question a test's screening efficiency, especially when the prevalence of the disease exceeds 1%.5 Too many lice-free children receive unnecessary treatment, and too many infestations escape detection, jeopardising the control of an epidemic. The gold standard character of wet combing for detection of head lice needs confirmation to legitimise the extra logistic effort of screening campaigns that use wet combing.
Table.
Classical scalp inspection | Wet combing
|
||
---|---|---|---|
Not infected | Infected | Total | |
Not infected | 161 | 17 | 178 |
Infected | 14 | 32 | 46 |
Total | 175 | 49 | 224 |
Acknowledgments
We thank the board of the participating school for allowing data collection, the members of the screening teams for their participation, and the children for their patience. We thank the project leaders of the health community centres of Ghent (L Gijssels and C Bracke) for their permission to integrate this study into their pilot project. B Vincke gave secretarial assistance with data entry.
Footnotes
Funding: Funded by the participating community health centres and by the City of Ghent (SIF contract Action 42).
Competing interests: None declared.
References
- 1.Bingham P, Kirk S, Hill N, Figueroa J. The methodology and operation of a pilot randomized control trial of the effectiveness of the bug busting method against a single application of insecticide product for head louse treatment. Public Health 2000 (in press). [DOI] [PubMed]
- 2.Ibarra J, Hill N. Towards the establishment of bug busting in the public calendar. London: King's Fund; 1994. [Google Scholar]
- 3.Figueroa J, Hall S, Ibarra J. Primary health care guide to common UK parasitic diseases. London: Community Hygiene Concern; 1998. pp. 1–16. [Google Scholar]
- 4.Abramson JH. Making sense of data. A self-instruction manual on the interpretation of epidemiologic data. New York: Oxford University Press; 1988. [Google Scholar]
- 5.Sackett DL, Haynes RB, Guyatt GH, Tugwell P. Clinical epidemiology. A basis science for clinical medicine. Boston: Little, Brown; 1985. [Google Scholar]