The importance of socio-economic status and individual characteristics on the prevalence of head lice in schoolchildren

ARTICLE

Auteur(s) : Sara WILLEMS*¹, Hilde LAPEERE^*2, Nele HAEDENS¹, Inge PASTEELS¹, Jean-Marie NAEYAERT², Jan De MAESENEER¹

¹Ghent University, Department of General Practice and Primary Health Care, Campus UZ- 1K3, De Pintelaan 185, B-9000 Gent, Belgium
²Ghent University, Department of Dermatology, Campus UZ-1P6, De Pintelaan 185, 9000 Ghent, Belgium

accepté le 10 Juin 2005

Materials and method

Population and sample

Data collection

A child was found positive if a living louse was found. The presence of nits was also recorded, but no distinction was made between viable or dead eggs or empty eggshells. The positive children were given a letter for their parents with treatment advice and were screened again two weeks later. For every child, the SHD staff collected data about demographic characteristics (sex, date of birth, class and school), properties of the hair (length, colour and type) and characteristics of the family (number of children in the family and socio-economical status). Hair length was defined as very short (< 2 cm), short (≥ 2 cm but less than shoulder length), medium (= shoulder length) or long (> shoulder length). Hair colour was divided into 4 categories (black, brown, red or fair) (( figure 1 )) and hair type into 3 categories (straight, curly or frizzy).

The socio-economic status of the family was based on the person with the highest occupational status in the household and retrieved from the children’s school file. Occupation was classified according to the Standard Occupational Classification, published by the Office of Population Censuses and Surveys, using 4 categories: unemployed, manual worker, non-manual worker and professional [14].

The nationality of the child was also recorded but not included in the analysis since there were too many different nationalities and recording this information into a dichotomous variable (Belgian versus other) led to an important loss of data. All parents of positive children were given treatment advice and the treatment chosen was recorded for 25% of the positive children. No information was obtained on whether the treatment was applied correctly. Positive children were screened again 14 days after baseline screening. Results from the second screening are available for 87% of the children found to have lice at the baseline.

Statistical analysis

Results

Bivariate analysis demonstrated a significant association between head lice and the child’s sex, educational level, hair length and colour, number of children in the family and SES (table 1)( Table 1 ).

Multilevel analysis showed that the variance at school level is 1.93 (P = 0.011). The same parameter at class level within schools is 2.58 (P < 0.001) and the residual level (level of the different children within classes) is 0.59 (P < 0.001).

This shows that a child’s school and in particular his or her class have a greater impact on the risk of head lice than individual characteristics. The impact at child level can be attributed to SES (P = 0.017), the number of children in the family (P < 0.001), the length of the child’s hair (P = 0.028) and the hair colour (P = 0.021) (table 2)( Table 2 ).

Being from a family with a lower SES, more children and having longer hair tend to result in a higher risk of getting head lice (respectively OR: 0.80,95% CI: 0.68-0.96; OR: 1.2,95% CI: 1.10-1.32; OR: 1.2, 95% CI: 1.02-1.43). For example, for every child that a family has, the probability ratio of getting head lice increases by 1.2 times. Having black hair entails a lower risk than having brown hair (OR: 0.6,95% CI: 0.43-0.89).

Head lice were still found in 41.4% of the children screened fourteen days after baseline screening.

According to the multilevel analysis, only the hair colour (P = 0.011) and SES (P = 0.031) were statistically significant. The lower the SES, the higher the chance of still being positive on the second screening (OR: 0.6, CI: 0.37-0.95). Children with black hair have a higher chance of being lice-free at the screening than children with brown hair (OR: 0.3, CI: 0.11-0.74).
Table 1 Sex, age, educational level, hair characteristics, number of children in the family and socio-economic status (SES) of the family: bivariate results at baseline and two weeks after baseline screening

Discussion

The strength of this study is its methodology. No similar survey applying the same diagnostic criteria and screening method and including a comprehensive set of confounding factors has ever been performed on such a large sample. Furthermore, the use of multilevel analysis allowed the unique impact of each independent variable to be determined. A source of concern in the present study could be the lack of a randomised sample. However, a good representative sample of the population was achieved with only a slight overrepresentation of larger kindergartens and primary schools. Furthermore, the prevalence might be influenced by seasonal trends in head lice infestations as this study was carried out between January and June. Some authors found a difference in prevalence according to the month when a screening was performed [17, 18]. Finally, information on SES was not always recorded in detail in the child’s school record, resulting in a loss of data.

This study confirms the results of other studies in the field, yet on some points it nuances, refutes or explains earlier findings. The clustering of head lice within classes was observed in a prevalence study on 735 pupils from one school [4]. Our study, performed on a larger sample of children from different schools, confirms the observation of the latter study. Concerning hair characteristics, as reported earlier by Mumcuoglu et al. [3], we found higher prevalence rates in children with brown hair. One can assume that brown-coloured lice contrast more in fair, red and black hair and are therefore detected earlier. Earlier treatment of these children may lead to lower prevalence at formal cross-sectional screening sessions. Hair length can probably be considered as an important transmitting factor, as head lice need specific conditions to move from one head to another [16], which occur more frequently in longer hair. It also seems that the infestation rate rises with the number of children in the family, possibly because children in large families have a higher risk of being infested by their siblings (or parents). The influence of the family’s socio-economic status on the presence of head lice has been investigated before, using indicators such as educational level or profession of the parents, family income, recourse to social security and socio-economic class of the school. A positive correlation has been found [1, 6, 11, 19] in some reports but not in others [7, 8]. However, in most surveys, only bivariate analysis has been used. This study shows that when adjusting for a range of confounding factors, SES (defined as the highest occupational class in the household) is significantly related to the prevalence of head lice. In several studies, a difference was found in prevalence rates between girls and boys [1, 4, 6, 7, 20]. It is believed that gender-related behaviour differences affect transmission rates, e.g. difference in personal grooming, close contact, hairstyle changes and the use of hair accessories [21]. Counahan et al. supported the gender difference based on a large study that used multilevel statistical techniques and included type of residence, class, gender and hair length as confounding factors [20]. However, when building a model that adjusts to a more detailed set of determinants including other hair characteristics, no difference was found. We found that the commonly assumed risk factor of age was not associated with active infestation [6, 10]. Probably, this could be explained by the high association with school class which was introduced as a level in the model. This study did not confirm earlier findings of lower infestation rates in children with curly hair [3]. Nymphal stages are easily missed in curly or long hair when using the dry inspection method, possibly leading to an underestimation of head lice infestation in children with curly or long hair [22]. When applying wet combing, even the small lice are found in straight as well as in curly hair [23].

A striking result from the screening after treatment advice is that 41.4% of the children found to be positive in the baseline screening were still positive, meaning they were not adequately treated. This is rather high, considering that every parent received treatment advice. Vander Stichele et al. also found a low treatment effectiveness of 51% [9]. Several factors could be responsible for the high rate of treatment failure. It cannot be excluded that some parents, regardless of the advice given, did not treat their children. It is more likely, however, that there was resistance to the chosen pediculicide or that a lot of children were not treated correctly. Pediculicides are often expensive and should be applied correct in order to have optimum results. The wet combing method was also discussed as a treatment option. This method is relatively cheap but requires stringent implementation and is time-consuming. Wet combing should be repeated meticulously once every 3 to 4 days over a period of 14 days to break the life cycle of the head louse [9] and is not effective if the instructions are not followed correctly.

Attention should be paid when interpreting the results of this study. A significant association was found between head lice and several factors at child level, yet this does not imply a causal relationship. Presumably, these factors lead indirectly to higher prevalence rates because they hamper the detection (e.g. hair colour) or treatment (e.g. SES) of lice.

This survey illustrated that the clustering of children in classes and schools is a more important determinant of the prevalence of head lice than personal characteristics. Head lice infection is therefore difficult to avoid, making its early detection and effective treatment all the more important. Although earlier studies indicated it was feasible to screen larger groups of children for head lice using the wet-combing technique, the feasibility of organising large screening campaigns in schools several times per year is questionable. Probably, it is more advisable to introduce screening in the daily routine of the family, e.g. as part of the weekly bathing ritual. Schools and school health departments could play an important supporting role for the parents by organising information campaigns and “practice sessions” on the screening and treatment of children.

The current study showed that children from families with a lower SES are particularly at risk of getting head lice. Previous studies showed that the lives of people from the lowest socio-economic classes are often characterised by a short-term “survival” perspective. “Compliance” is, from a socio-culturally point considered, a middle-class concept because in lower social classes, the need for an adequate follow-up of the treatment instructions could be overruled by acute social problems related to income, housing conditions… [24]. Therefore special emphasis should be put on the need to establish specific support systems (e.g. by school nurses and primary care nurses) for children from low SES families with lice infestations. As Whitehead has indicated, efforts to decrease social inequalities in health must adopt a multi-level approach and focus on empowering families, strengthening communities (by educational efforts towards caregivers) and encouraging macro-economic and cultural change [25].

Further research is needed to confirm the findings of this study using a randomised sample and more detailed measurement scales, e.g. for SES. A cost-effectiveness study should be performed in which the benefits and costs of wet combing versus detection-combing of dry hair are assessed.

Acknowledgements

References

2 Badiaga S, Menard A, Tissot Dupont H, Ravaux I, Chouquet D, Graveriau C, et al. Prevalence of skin infections in sheltered homeless of Marseilles (France). Eur J Dermatol 2005; 5: 382-6.

3 Mumcuoglu KY, Miller J, Gofin R, Adler B, Ben-Ishai F, Almog R, et al. Epidemiological studies on head lice infestation in Israel. I Parasitological examination of children. Int J Dermatol 1990; 29: 502-6.

4 Speare R, Buettner PG. Head lice in pupils of a primary school in Australia and implications for control. Int J Dermatol 1999; 38: 285-90.

5 Downs AMR, Stafford KA, Stewart GH, Coles GC. Factors that may be influencing the prevalence of head lice in British school children. Ped Dermatol 2000; 17: 72-4.

6 Amr ZS, Nusier MK. Pediculosis capitis in northern Jordan. Int J Dermatol 2000; 39: 919-21.

7 Sim S, Lee IY, Lee KJ, Seo JH, Im KI, Shin MH, et al. A survey on head lice infestation in Korea (2001) and the therapeutic efficacy of oral trimethoprim/sulfamethoxazole adding to lindane Shampoo. Korean J Parasitol 2003; 41: 57-61.

8 Kokturk A, Baz K, Bugdayci R, Sasmaz T, Tursen U, Kaya TI, et al. The prevalence of pediculosis capitis in schoolchildren in Mersin, Turkey. Int J Dermatol 2003; 42: 694-8.

9 Vander Stichele RH, Gyssels L, Bracke C, Meersschaut F, Blokland I, Wittouck E, et al. Wet combing for head lice: feasibility in mass screening, treatment preference and outcome. J R Soc Med 2002; 95: 348-52.

10 Mumcuoglu KY, Friger M, Ioffe-Uspensky I, Ben-Ishai F, Miller J. Louse comb versus direct visual examination for the diagnosis of head louse infestations. Pediatr Dermatol 2001; 18: 9-12.

11 Menan E. N‘Guessan G, Kiki-Barro PC, Nebavi NG, Adjetey TA, Kone M. Scalp pediculosis in school environment in the city of Abidjan: prevalence and influence of socio-economic conditions. Sante 1999; 9: 32-7.

12 Burgess IF. Human lice and their management. Adv Parasitol 1995; 36: 271-342.

13 De Maeseneer J, Blokland I, Willems S, Vander Stichele R, Meersschaut F. Wet combing versus traditional scalp inspection to detect head lice in schoolchildren: observational study. BMJ 2000; 321: 1187-8.

14 Anonymous. Office of Population Censuses and Surveys. Standard Occupational Classification. Vols 1 and 2. London, England: Her Majesty’s Stationery Office, 1990.

15 Brown H, Prescott R. Applied mixed models in medicine. Wiley Chichester., 1999.

16 Canyon DV, Speare R, Muller R. Spatial and kinetic factors for the transfer of head lice (Pediculus capitis) between hairs. J Invest Dermatol 2002; 119: 629-31.

17 Mimouni D, Ankol OE, Gdalevich M, Grotto I, Davidovitch N, Zangvil E. Seasonality trends of Pediculosis capitis and Phthirus pubis in a young adult population: follow-up of 20 years. J Eur Acad Dermatol Venereol 2002; 16: 257-9.

18 Combescot C. Epidémiologie actuelle de la pédiculose à Pediculus capitis. Bull Acad Natle Med 1990; 174: 231-7.

19 Gbakima AA, Lebbie AR. The head louse in Sierra Leone: an epidemiological study among school children, in the Njala area. West Afr J Med 1992; 11: 165-71.

20 Counahan M, Andrews R, Buttner P, Byrnes G, Speare R. Head lice prevalence in primary schools in Victoria, Australia. J Paediatr Child Health 2004; 40: 616-9.

21 Downs AMR, Stafford KA, Coles GC. Head lice: prevalence in schoolchildren and insecticide resistance. Parasitol Today 1999; 15: 1-4.

22 Downs AM, Oxley J. Head lice infestations in different ethnic groups. Int J Dermatol 2001; 40: 237-40.

23 Ibarra J, Hall DM. Head lice in schoolchildren. Arch Dis Child 1996; 75(6): 471-3.

24 Willems S, De Roo L, Anthierens S, Derese A, De Maeseneer J. Towards a new model for understanding the complexity of health services utilisation patterns by the poor and underserved. So Sci Med (Submitted).

25 Whitehead M. Tackling inequalities: a review of policy initiatives. In: Benzeval M, Judge K, Whitehead M, eds. Tackling inequalities in health. London: King’s Fund, 1995: 22-51.