IntroductionPublic health is a main factor of development in each community . Despite the promotions in health and medical education, external parasitic infestation is a threat for community health development which is still considered as a public health concern . Head lice are insects living as external parasites on humans and animals body. Human infestations occur as head lice (the most common type), body lice, and pubic lice .Head lice infestation is a main public health problem and a great threat for personal hygiene involving children aged between 5 years and 13 years . The disease is widespread around the world but is more common in developing countries, among school age children, and in crowded places with low socio-economic and poor hygiene conditions 5, 6, 7. Indicators concerning the infestation of head lice are used to assess the health, cultural, and economic situation of rural and urban communities .Epidemic relapsing fever and epidemic typhus can be transmitted by lice. Because of frequent blood feeding and saliva injection, fatigue, psychological irritability, paranoia, and weakness can also occur. Other related complications are as follows: depression, hyperthermia, headache, feeling of heaviness in the body, muscle rigidity, attention deficit in the class and educational failure, insomnia, lack of social status, developing secondary infections, hair loss, and allergies 7, 8, 9.Different prevalences of head lice among primary school children in the world have been reported. According to the results of a study , head lice prevalence was estimated as 4.8% (The Netherlands), 35% (Brazil), 1.2% (Turkey), 28.8% (Venezuela), and 29.7% (Argentina). Such variations have been observed in different provinces of Iran such as 4% in Urmia , 13.5% in Hamedan , 1.8% in Kerman , and 4.7% in Sanandaj .The initial search results showed that different studies have been carried out to estimate the prevalence of head lice infestation among Iranian primary school children. There are remarkable variations in the results of these studies limiting their application for decision making and policy making. To provide reliable evidence, combining the findings of different studies using systematic review and meta-analytic methods is considered as an appropriate solution. This study aims to estimate the prevalence of pediculosis among Iranian primary school children using meta-analysis techniques.
2.1 Search strategyDifferent national (SID, Iranmedex, Magiran, and Irandoc) and international (PubMed, Google scholar, Scopus, and ScienceDirect) databases were investigated to provide articles published from January 1, 2000 until January 20, 2015. The search strategy was performed using keywords including: “epidemiology”, “prevalence”, “infestation”, “head louse”, “head lice”, “Pediculus humans capitis”, “pediculosis”, “primary school students”, “primary students”, “school children”, and “Iran”, as well as their Farsi equivalents.Searching was conducted from January 21, 2015 to February 10, 2015. To increase the search sensitivity and to investigate more evidence, references of published articles were hand searched. Two researchers independently reviewed the studies for a possible inclusion in the review using an eligibility form. In addition, to find the results of unpublished studies, we interviewed some related experts and research centers.
2.2 Study selectionFirst the titles, then the abstract, and finally the full text of the articles were consulted to retrieve the studies fit for the review. The relevant studies were selected independently by two researchers. It should be noted that in order to prevent reprint bias, the findings of all studies were carefully reviewed and any duplicated research was excluded.
2.3 Quality assessmentSelected articles were assessed for quality using a standard checklist applied in previous studies . Based on the components used in the STrengthening the Reporting of OBservational studies in Epidemiology checklist , a checklist was designed and used in the current systematic review including 12 questions about different methodological aspects of a study such as sample size estimation and selection, type of study, study population, data collection method and instrument, variable definition, diagnostic method of pediculosis infestation, statistical tests, objectives of the study, appropriate illustration, and presentation of findings according to the objectives. Studies obtaining at least eight quality scores  were entered in the final meta-analysis.
2.4 Inclusion criteriaAll Persian and English written papers that passed the necessary assessment steps, received the required quality scores, and also estimated the infestation rate for all students or boys or girls, were entered in the meta-analysis.
2.5 Exclusion criteriaStudies without prevalence results for all students or according to both sex, those without definite sample size, abstracts presented in congresses without full text, and case–control studies and clinical trials that could not report a correct estimate of prevalence were excluded from the meta-analysis step.
2.6 Data extractionAll necessary information such as title, first author name, date of the study conduction, sample size (total, boys and girls), type of the study, sampling method, study area, language, prevalence of infestation according to gender, p value for the association between infestation and potential risk factors (educational levels and parents' profession, hair length, shared use of personal cosmetics, family size, easy access to bathroom or having bathroom in the house, and number of bathing occasions per week) were extracted from each study and entered into an excel spreadsheet. Meanwhile, two researchers independently extracted the data.
2.7 Statistical AnalysisAll collected data were entered into STATA version 11 software (Stata Corporation, College Station, TX, USA). Standard error of the prevalence in each study was calculated according to binomial distribution. Cochrane test (Q) and I2 indices were applied to determine the heterogeneity of the study results. Due to the significant heterogeneity, a random effect model was used to estimate the total and gender specific prevalence's of pediculosis. In addition, to minimize the random variations between the point estimates, results of the studies were adjusted using Bayesian analysis methods. Finally, we assessed the potential effects of different variables on the mentioned heterogeneity using met regression methods. Having conducted a sensitivity analysis, studies influencing the heterogeneity were determined. The forest plot was designed to illustrate the point prevalence's of the selected studies as well as their 95% confidence intervals (CIs). In this graph, the size of each box indicated the study weight, while the CI was illustrated by the lines crossing the boxes.
ResultsIn the first phase of study search, 12,541 records were identified. Restricting the search strategy, the relatively relevant studies were reduced to 1,056. In the next steps, 323 duplicate papers and 618 irrelevant articles (during the title or abstract review) were excluded. During the quality assessment, 69 studies were removed, while four papers were added after review of references. Finally, 40 studies 4, 7, 8, 11, 12, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49 were determined eligible to enter into the meta-analysis (Figure 1; Table 1).Selected studies for meta-analysis included a total of 200,306 primary school students, 49,559 boys and 107,321 girls. Type of studies were determined for 31 studies (16 descriptive and 15 descriptive-analytic). Among the 32 studies reporting the sampling method, 19 of which had randomization, five had consensus, and eight had multistage cluster sampling. Publication dates varied between 2000 and 2014.The total prevalence of infestation varied from 0.47% in the Doroodgar et al  study with a sample size of 3,589 conducted in Isfahan (center of Iran) in 2013 to 27% in the study carried out by Alempour-Salemi et al among 918 students in Sistan-Baloochistan province (south-east of Iran). As illustrated in Table 1, infestation prevalences among boys differed from zero in the Salehi et al  study to 11.1% in the Soleimani et al  study. The corresponding figures for girls were 0.1% (Doroodgar study ) and 35.3% (Soleimani study ) respectively. Using Bayesian analysis method and adjustment of the effect of studies, the total prevalence modified to 0.5% (95% CI: 0.3–0.7) in the Doroodgar study  and 23.1% (95% CI: 20.5–25.6) in the Alempour-Salemi study.According to the results of the sensitivity analysis, the Motevalli-Haghi et al  study and the Motevalli-Emami et al  study carried out among 45,237 and 26,421 students respectively affected the amount of heterogeneity more than other studies. However, excluding these studies from the meta-analysis did not remove the heterogeneity (p < 0.0001). Because the heterogeneity did not change after exclusion of six articles, meta-analysis was carried out for all 40 studies using a random effect model to prevent more reduction of the primary studies.The forest plot (Figure 2) shows the total prevalence of head lice per study and a pooled estimate. Using a random effect model, the total prevalence of head lice among Iranian primary school children was estimated as 7.4% (95% CI: 6.6–8.2). The corresponding estimates for boys and girls were 1.6% (95% CI: 1.2–2.05) and 8.8% (95% CI: 7.6–9.9), respectively (Table 2).The effects of publication date and the climate of the study area on the pooled estimate of prevalence were investigated using met regression models. It was noted that only the effect of publication date was statistically significant, i.e., there were four unit decreases per year in the prevalence of pediculosis (Table 3).The association between hair length and pediculosis was investigated in 14 studies, eight of which reported a significant effect of long hair. Of 15 studies assessing the relationship between pediculosis and shared use of personal hygiene items, eight studies observed significant associations between use of a common comb and developing head lice infestation.Twenty-seven studies investigated the influence of the level of mother's education, all of which showed that the risk of infestation was significantly higher among children whose mothers had low levels of education. In the case of education of fathers, 21 out of 27 studies reported significant positive associations between low levels of father's education and pediculosis.The associations between the jobs of mothers and fathers were investigated in 19 studies and 23 studies, respectively. Students whose mothers were housewives (5 studies) as well as those whose fathers were worker or jobless (14 studies) were significantly at higher risk of pediculosis.Twenty-two studies assessed the relationship between prevalence and family size, 17 of which reported a higher risk of infestation in students living in bigger size families. The effect of having a bathroom in the house on pediculosis was significant in nine out of 14 studies investigating these associations. Of these, seven papers showed higher frequencies of infestation among children without access to a bathroom within their homes and also eight out of 14 studies reported higher rates of infestation among students with a lower frequency of bathing per week.
DiscussionThe main outcome of this meta-analytic study carried out among 200,306 Iranian elementary school children was that the overall head lice infestation rate was 8.8% with girls infested 5.5 fold higher than boys. We also observed that the infestation was more common among students with low educated parents, long hair, big family size, housewife mothers, worker or jobless fathers, using shared combs, lack of bathrooms, or lower frequency of bathing at home.According to Table 4, 14 studies showed the prevalence of infestation from 4.1% in the Oh et al  study in Korea to 40.3% in the Gazmuri et al  study in Chile. Infestation rates in Chile , Egypt , Turkey 1, 52, 53, 55, Thailand , Jordan , Argentina , Malaysia , and Mexico  were more than that in Iran. While, the prevalence of infestation was lower in Korea , Australia , and another Turkish study  in comparison with the prevalence in Iran.Among the above foreign studies, eight studies reported sex specific prevalence estimates. The prevalence among girls in Korea  was lower than that of the current study, while Chile , Turkey 52, 53, 54, Thailand , Jordan , and Argentina  reported more girls contaminated than in Iran. Iranian school boys were less affected than those in Chile , Jordan , and Argentina  and more than boys investigated in Turkey 52, 53, 54 and Thailand . The prevalence of infestation In Iranian boy students was similar to those of Korean students .Comparing the results of different countries with those in Iran showed that the rate of infestation among all Iranian students, as well as among girls, is lower than those in most other countries. Conversely, Iranian boy students were more infested than those in most other countries. Moreover, the reported prevalences in other countries varied greatly. These reports showed that although the infestation rate is distributed worldwide; it is not equal among different countries and seems to be associated with cultural and socioeconomic factors 55, 56, 57, 62. A limitation in this comparison is that the reported prevalence in Iran is based on meta-analytic results while the results of the other countries are according to primary studies. These comparisons could be different if the prevalences of the other countries were estimated with meta-analytic methods.Risk of pediculosis is associated with different factors. Prevalence of infestation is more common in low socioeconomic regions, high density populations, where there is lack of personal hygiene, poor health facilities, and lack of a school health educator. This infestation also occurs when hair and clothes are not regularly washed. In addition, pediculosis is related to educational levels of parents and their attitude to health standards . Our study showed that the frequency of head lice infestation among primary school children is related to different factors. In a study conducted among 2,210 Korean school students, similar to the current study, the infestation was significantly less common among students whose parents were teachers or employees and more common among those with less frequent hair washing per week. However, in contrast to the current study, the prevalence of infestation was contributed to a small family size and no association was observed between infestation and parents' education , while more educated Iranian parents increased the risk of pediculosis according to our meta-analysis.In Egypt , primary school girls, in comparison with school boys, were 25.8% more infested and family size and access to enough water supplies were associated with pediculosis infestation. These findings are in accord with the current situation in Iran. Turkish school girls were 3.1 fold more contaminated than boys , while this ratio was higher for Iranian students.In the other study carried out in Turkey , female sex, living with three or more siblings, and low educational levels of mothers and fathers, increased the risk of infestation approximately 41 folds, two folds, 73% and 45%, respectively. Such findings are in agreement with the current study.According to a third Turkish school survey, head lice infestation prevalence was more common among students with a previous personal and family history of infestation. However, frequency of bathing and hair washing, shared use of a comb, towel and bed, family size, and also the number of family members per room had no influence on pediculosis infestation prevalence. Similar to our findings, the rate of infestation among Turkish students with long hair were more than three times higher than those of students with short hair .Çetinkaya et al , in a study performed in Turkey, found no association between infestation with parents' level of education, type of house, and lack of bathroom within the house. Based on a cross sectional study in Jordan in 2012 , infestation was significantly related to sex (34.7% in girls vs. 19.6% in boys), living in a village (31.2% in rural area vs. 23.5% in urban area), and past history of infestation (57.4% in those with a positive history vs. 11.5% in those without). It was also associated with small family size, low income, long hair, lack of a bathroom, and low frequency of hair washing. In addition, a cross sectional study in Mexico  introduced factors such as low income (odds ratio = 9.9) and low frequency of hair washing (odds ratio = 8) as risk factors for head lice infestation.Comparison of the factors relating to pediculosis infestation demonstrated the common determinant factors were different among countries. Approximately all associated factors had socioeconomic, cultural, behavioral, and health backgrounds. Therefore, we can overcome this global health problem by reasonable planning. The majority of these factors are easily avoidable with minimum cost and simple actions such as keeping the hair neat and short, increase the frequency of hair washing per week, not sharing a personal comb and towel, and effective and prompt treatment of infested students. The main point in the prevention of reinfestation is establishing and maintaining appropriate health behaviors by comprehensive and continued supervision.The current study represented a comprehensive understanding of pediculosis prevalence and its determinant factors among Iranian elementary school children. Since the results of this study have been estimated based on scientific and statistical procedures and were comprehensively compared with the results of other national and international studies, they can be reasonably used for policymaking.Many investigated studies during the current systematic review and meta-analysis did not report any significance level (p value) of the associations. Therefore, we could not perform p value meta-analysis for the correlations between pediculosis and other relevant factors. Remarkable heterogeneity between primary studies was another limitation of this study. Therefore, we applied a random effect model to combine the primary results in this meta-analysis.Our study revealed that there is a high prevalence of head lice infestation among Iranian primary school children, particularly girls. We also observed that parents' educational level and profession, shared use of personal hygiene items, hair length, family size, having a bathroom within the house, and frequency of bathing per week, can be considered as determinant factors of infestation among these students.