Background

Skin disorders including psoriasis and eczema, also referred to as atopic dermatitis, are the fourth leading cause of disability worldwide [1]. These disorders have an impact on quality of life [2–6] and create a burden on healthcare resources [4–6]. The annual cost of eczema and psoriasis in the USA have been estimated at $5.3billion [7] and as much as $63.2billion [8] respectively and it has been reported that skin diseases globally contributed 1.8% (18th leading contributor) and 1.4% (22nd leading contributor) to disability-adjusted life years (DALYs) in 2013 and 2016 respectively; eczema and psoriasis contributing 0.4% and 0.2% respectively in the former report [1,9]. Skin disorders are more common in children than adults. Worldwide, the prevalence of eczema has been reported as 17–20% in children compared with 1–3% in adults [10,11]. Childhood prevalence varies greatly between countries [12,13], ranging from 0.2% in China to 24.6% in Columbia. Psoriasis is a less common condition in children but still affects around 0.5% [14]. Children who have a skin disorder often suffer from additional medical conditions. Studies have reported increased risk of asthma, allergic rhinitis, hay fever and autoimmune disorders among children with eczema [15–18] and a higher prevalence of obesity, diabetes, cardiovascular disorders, rheumatoid arthritis, and Crohn’s disease among children with psoriasis [14,19–21]. Eczema and psoriasis have also been linked with a number of psychiatric disorders and mental health problems including depression and anxiety [22–28]. In addition to depression, childhood eczema has been associated with comorbid attention deficit hyperactivity disorder (ADHD) [29–31] and suicidal ideation/behaviour [32–34]. Adults with skin disorders reportedly suffer from increased mortality [35,36] and, whilst studies investigating hospitalisation and mortality in children and adolescents are rare, increased psychiatric hospitalisation [37] and mortality [38] among children with psoriasis have been reported.

Given that these disorders not only impact on physical and mental health but also affect sleep and daily function [39], it is plausible that having a skin disorder may adversely affect educational attainment and employment. Increased absence from school and work [39] and perceived school difficulty among students [40] have been reported, however a recent systematic review of educational and employment outcomes among children with eczema identified only one eligible study [41]. The study reported no association between eczema and performance in a secondary school entrance exam undertaken at 11years of age but did not have information on subsequent school academic attainment or employment after leaving school [42]. A subsequent national study across Sweden found no associations between atopic dermatitis and lower cognitive function or lower academic attainment but did not investigate other educational or health outcomes. Further, it only studied men aged 17–20 between 1969 and 1976 who possibly had more severe atopic dermatitis, limiting generalizability [43]. This study addresses a significant gap in the literature by investigating school attendance, educational attainment, and employment, as well as hospitalisations and mortality, in a single unselected country-wide cohort of schoolchildren treated for a skin disorder compared to their peers.

Methods

Results

Between 2009 and 2013, 766,244 singleton children attended Scottish schools. Overall, 130,087 (17.0%) children were treated for a skin disorder (Table 1). Children treated for a skin disorder were more likely to be female and Asian, and less likely to live in deprived areas. They were more likely to have required assisted or operative delivery and had lower five-minute Apgar scores. Their mothers were older, less likely to have smoked during pregnancy and were more likely to have been nulliparous. Compared to their peers, children treated for a skin disorder were also more likely to be on medication for diabetes (0.5% versus 0.4% of peers, p = 0.006), asthma (11.8% versus 4.8% of peers, p<0.001), epilepsy (1.0% versus 0.6% of peers, p<0.001), ADHD (1.0% versus 0.9% of peers, p = 0.007) and depression (1.0% versus 0.6% of peers, p<0.001) (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Sociodemographic characteristics of schoolchildren by the presence of a treated Skin Disorder (SD).

https://doi.org/10.1371/journal.pone.0243383.t001

Analyses of school absences and exclusions were conducted on 702,210 children. Children treated for a skin disorder had more days absent from school than their peers after adjusting for sociodemographic and maternity factors (IRR 1.07, 95% CI 1.06–1.08, p<0.001). There was an interaction with age (p<0.001) whereby the association between skin disorder and absenteeism was stronger among children aged 11–14 years (IRR 1.09, 95% CI 1.07–1.11, p<0.001) compared to children older than 14 years (IRR 1.03, 95% CI 1.01–1.04, p<0.001) and younger than 11 years (IRR 1.07, 95% CI 1.06–1.08, p<0.001).

Treated skin disorders were associated with lower rates of school exclusion after adjusting for sociodemographic and maternity confounders (IRR 0.91, 95% CI 0.85–0.97, p = 0.004). However, on subgroup analyses, this association only reached statistical significance among boys (IRR 0.91, 95% CI 0.84–0.98, p = 0.015), and children aged >14 years (IRR 0.86, 95% CI 0.78–0.94, p = 0.001).

Children treated for a skin disorder were more likely to have a record of special educational need after adjusting for sociodemographic and maternity factors (OR 1.19, 95% CI 1.17–1.21, p<0.001). There was an interaction with age (p<0.001) whereby the association was stronger among children older than 14 years of age (OR 1.28, 95% CI 1.23–1.32, p<0.001) compared to children aged 11–14 years (OR 1.23, 95% CI 1.19–1.27, p<0.001) and children younger than 11 years (OR 1.13, 95% CI 1.10–1.16, p<0.001).

Among the 139,205 children who had undertaken exams, there was no significant association between treatment for a skin disorder and academic attainment after adjustment for sociodemographic and maternity factors (OR 1.01, 95% CI 0.96–1.05, p = 0.735). Among the 217,924 children who left school during the study period, 3,917 (25.5%) of those treated for a skin disorder left school before 16 years of age compared with 58,864 (29.1%) of their peers (p<0.001). Having a skin disorder was associated with lower risk of unemployment overall after adjusting for sociodemographic and maternity factors (OR 0.90, 95% CI 0.85–0.96, p<0.001). However, on subgroup analyses, this association was only statistically significant among boys (IRR 0.86, 95% CI 0.79–0.94, p = 0.001), and children in the most deprived quintile (IRR 0.83, 95% CI 0.74–0.92, p = 0.001).

Linkage to hospital records provided 2.84 million person years of follow-up. The mean follow-up duration was 3.7 years. Of the 766,244 children, 153,002 were admitted to hospital at least once and 294,864 hospital admissions occurred in total. In the Cox proportional hazards models, children treated for a skin disorder experienced increased risk of hospitalisation for any cause (HR 1.29, 95% CI 1.27–1.30, p<0.001). However, the proportionality assumption was not met (p<0.001). On running a Poisson piecewise regression model, children treated for a skin disorder were more likely to be hospitalised over all five study years and across all ages. However, the incidence rate ratio was highest in the first recorded year of treatment (IRR 1.38, 95% CI 1.35–1.41, p<0.001) and fell over the remaining four years. It also peaked among children aged 11–12 years (IRR 1.40, 95% CI 1.35–1.46, p<0.001) on stratifying the analyses by age at admission (<6, 7–8, 9–10, 11–12, 13–14, 15–16, >17 years). There was an interaction with sex (p<0.001) whereby the association between a treated skin disorder and hospital admission was present in both boys and girls but stronger in the former. Fig 1A and 1B show fully adjusted incidence rate ratios, stratified by sex, for all-cause hospitalisation for each year of follow-up and by age at admission derived from Poisson piecewise regression models.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Association between treatment for a skin disorder and all-cause hospitalisation by sex (boys = solid square; girls = hollow diamond).

a. By time from commencement of medication. b. By age at admission. Adjusted for age, sex, deprivation quintile, ethnic group, maternal age, maternal smoking, parity, mode of delivery, gestation at delivery, sex- gestation-specific birthweight centile and 5-minute Apgar score. SIMD–Scottish Index of Multiple Deprivation.

https://doi.org/10.1371/journal.pone.0243383.g001

There were 491 deaths over the follow up period; 87 among children treated for a skin disorder and 404 among their peers. Children treated for a skin disorder were more likely to die from any cause over follow-up after adjusting for sociodemographic and maternity factors (HR 1.50, 95% CI 1.18–1.90, p<0.001). Among all children who died, the three most commonly recorded causes of death were external causes (37.5%), neoplasms (17.8%) and diseases of the nervous system (10.4%). External causes accounted for a smaller proportion of deaths among children treated for skin disorders than their peers (20.7% versus 41.1%) as did neoplasms (13.8% versus 17.8%). In contrast, children treated for a skin disorder were more likely to have their cause of death attributed to diseases of the nervous system (14.9% versus 9.4%).

Adjusting the main analyses for presence of comorbid conditions, both individually and then together, did not alter the findings (Table 2). After adjusting for all of the comorbid conditions, children treated for a skin disorder still had increased risk of absenteeism (IRR 1.05, 95% CI 1.04–1.05, p<0.001), special educational need (OR 1.14, 95% CI 1.12–1.16, p<0.001), hospitalisation (HR 1.22, 95% CI 1.22–1.24, p<0.001), and mortality (HR 1.31, 95% CI 1.03–1.67, p = 0.027), but decreased risk of school exclusion (IRR 0.91, 95% CI 0.86–0.97, p = 0.006) and unemployment (OR 0.89, 95% CI 0.84–0.95, p<0.001). There remained no association with academic attainment (OR 0.99, 95% CI 0.95–1.04, p = 0.762). Confirmation of these results will require further research, identifying patients by their diagnosis and recording the clinical severity of their condition.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Educational and health outcomes additionally adjusted for comorbid chronic conditions.

https://doi.org/10.1371/journal.pone.0243383.t002

Discussion

This study is unique in identifying a large cohort of children from across Scotland who have received topical treatment for a skin problem. The absence of primary care data or a national database for outpatient diagnoses is a limitation and complicates the process of patient identification and deduction of type and severity of disorder. Nevertheless, in identifying patients by community prescriptions of topical treatment, this study found that, compared to peers, children treated for a skin disorder were more likely to be hospitalised, had more absences from school, were more likely to also be on medication for asthma, diabetes, epilepsy, depression and ADHD, and were more likely to be recorded as having special educational need. However, there was no evidence that this resulted in a longer-term adverse impact on attainment, which concurs with previous studies on atopic dermatitis [42,43], and employment. These latter findings may provide reassurance to affected children and their parents.

Whilst the findings relating to educational and employment outcomes were reassuring, we did demonstrate that, even after adjusting for comorbid conditions, children treated for a skin disorder were at increased risk of all-cause mortality. It is possible however, that patients with terminal disease, disease that results in immobility, or disease treatment that can result in skin side effects, may be prescribed emollients and steroids in the community for rashes that do not correspond to the innate, chronic skin disease that we would want to study. In any observational study, residual confounding and reverse causation are possible. Therefore, these findings require corroboration in other studies and if replicated merit further investigation. In particular, our findings that children with a skin disorder had less risk of exclusion, albeit only statistically significant among boys and those older than 14years of age, and less risk of unemployment, albeit only statistically significant among boys and those in the most deprived quintile, were unexpected and may be due to residual confounders. Therefore, these results should be corroborated and investigated further in future studies.

Previous studies investigating health outcomes of children with skin disorders are limited in number. Whilst adults with skin disorders reportedly suffer from increased mortality [35,36], studies reporting hospitalisation and mortality among children and adolescents are rare [37,38]. However, our finding of increased risk of death does concur with a previous study reporting greater premature mortality among children with psoriasis [38]. Our findings of increased special educational need and treatment for asthma, diabetes, epilepsy, depression and ADHD concur partially with previous studies which have reported greater risk of asthma [15,16,18] and ADHD [29–31] among children with eczema, increased incidence of diabetes [14,19,21] among children with psoriasis, and more psychiatric disorders and mental health problems among both including specifically depression and anxiety [22–28]. We have previously reported that children treated for diabetes, asthma, epilepsy, ADHD, and depression have poorer educational and health outcomes, including hospitalisation and mortality, compared to peers, and therefore adjusted for these conditions in our analyses [45–50]. However wider comorbidities exist which we could not adjust for. For example, childhood eczema has also been associated with comorbid allergic rhinitis, hay fever and autoimmune disorders [15–17] and even suicidal ideation/behaviour [32–34] whilst psoriasis has been associated with obesity, cardiovascular disorders, rheumatoid arthritis, and Crohn’s disease [14,19–21] as well as more psychiatric hospitalisations [37]. Therefore, in addition to the direct physical and mental effects associated with skin disorders, it is possible that the greater hospitalisation and mortality observed may be partly attributable to the increased physical and mental comorbidities associated with both eczema and psoriasis over and above those adjusted for in our analyses.

Previous studies investigating educational outcomes of children with skin disorders are rarer still and face methodological limitations. Indeed, a systematic review of educational and employment outcomes among children with eczema identified only one eligible study [41]. General school difficulty among affected children has been reported previously [40] and reports of increased absenteeism [39] yet unaffected attainment [42,43] concur with our own findings. However, the latter studies on attainment were limited whereby one only investigated attainment on entry to school at age 11 for children with eczema, rather than final attainment on leaving school, and the other only investigated attainment among boys with severe atopic dermatitis limiting generalisability. We have previously demonstrated that children treated for diabetes, asthma, epilepsy ADHD, and depression have greater risk of absenteeism and special educational need compared to peers however our new findings pertaining to skin disorders remained after adjusting for presence of those conditions [45–50]. In addition to the direct effects of having a skin disorder, our findings of increased absenteeism and special educational need may therefore similarly be partly attributable to the wider increase in physical and mental comorbidities experienced by affected children which we could not adjust for as highlighted already.

Our study addressed a significant gap in the literature by investigating several health and educational outcomes. Ours was a large, non-selective study including children attending schools across the whole of Scotland. Because the sampling frame was mainstream and special schools, rather than hospital clinics, ascertainment of children treated for a skin disorder was not restricted to the most severe skin disorders and we were able to adjust for a wide range of potential confounders: sociodemographic, maternity and comorbid conditions. The large cohort provided sufficient power to test for statistical interactions and undertake sub-group analyses where appropriate, and we were able to analyse a wide range of educational, employment and health outcomes within the same cohort of children.

Our study only included children attending local authority maintained schools; however, in Scotland, less than 5% of children attend private schools. According to the 2011 Scottish Census, 11% of Scottish residents aged 5–19 years were born outside of Scotland; consistent with the 12% of Scottish children we could not link to Scottish maternity records. We used existing, administrative databases established for other purposes. However, they undergo regular quality assurance checks. Linkage of education and health records relied on probabilistic matching which has been validated to be 99% accurate for singletons [44].

Our findings require corroboration in other studies and if replicated merit further investigation. Whilst acknowledging these uncertainties we conclude that, based on the observed poorer health outcomes, the introduction of treatment plans is welcome as an attempt to provide children and adolescents who have skin disorders with more personalised healthcare. Given that these children regularly experience comorbid physical and mental health conditions, a more joined up approach to care is required. Indeed, in light of the common psychological burden of these conditions, a recent study highlighted unmet psychological and health care needs of adolescents transitioning from paediatric to adult services and a requirement for more dedicated dermatology clinics in the UK with embedded psychological support capable of providing developmentally appropriate healthcare and psychosocial support for this population [54]. Having observed poorer attendance and greater special educational need among affected children we conclude that early identification and support should be a priority and that interventions should focus on reducing the risk of school absenteeism. In order to reduce school absenteeism or mitigate its effects, children treated for skin disorders should receive integrated care from a multidisciplinary team covering physicians, teachers, parents, and where mental health comorbidities exist, educational psychologists and social services as appropriate. Their management should extend beyond healthcare to a programme of school-based interventions.

Conclusion

Children prescribed topical treatment for a skin disorder have an increased risk of hospitalisation and mortality, miss more days of school, have greater special educational need than peers, and are more likely to be treated for other chronic conditions including diabetes, asthma, epilepsy, ADHD and depression. However, there was no evidence of an adverse impact on longer-term educational outcomes in terms of poorer exam grades whilst at school and unemployment after leaving school. Whilst our findings relating to educational and employment outcomes are reassuring, the association with increased risk of mortality is alarming and merits further investigation. The introduction of treatment plans is welcome as an attempt to provide children who have skin disorders with more personalised and joined-up healthcare. School interventions should include measures aimed at reducing absenteeism and providing support in the form of integrated care from a multidisciplinary team.

Supporting information

Acknowledgments

The authors would like to acknowledge the support of the electronic Data Research and Innovation Services (eDRIS) within Public Health Scotland for their involvement in obtaining approvals, provisioning, and linking data and the use of the secure analytical platform within the National Safe Haven.