COVID-19-related school closures and patterns of hospital admissions with stress-related presentations in secondary school-aged adolescents: weekly time series
This study examines health service indicators of stress-related presentations (relating to pain, mental illness, psychosomatic symptoms and self-harm) in adolescents of secondary school age, using Hospital Episode Statistics data for England. We examined weekly time series data for three academic years spanning the time before (2018–2019) and during the COVID-19 pandemic (2019–2020 and 2020–2021), including the first lockdown when schools were closed to the majority of pupils. For all secondary school children, weekly stress presentations dropped following school closures. However, patterns of elevated stress during school terms re-established after reopening, with girls aged 11–15 showing an overall increase compared with pre-pandemic rates.
Significant declines in the mental health and well-being of children and adolescents have been reported during the coronavirus (COVID-19) pandemic, with the closure of schools and subsequent disruption to learning and assessment, social contacts, routines and services accessed through schools playing a pivotal role.Reference Viner, Russell and Saulle1
Prior to the pandemic, hospital admissions relating to manifestations of stress in secondary school-aged adolescents were positively correlated with term times.Reference Blackburn, Ajetunmobi, Mc Grath-Lone, Hardelid, Shafran and Gilbert2 Stress-related presentations are those relating to pain, mental health or psychosomatic symptoms. They may indicate emerging or relapsing mental illnessReference Jensen, Goldman and Offord3 and the burden on schools and in health is significant, affecting 7.9% of girls and 4.1% of boys between the ages of 11 and 17 years, and accounting for over 30% of all emergency admissions for this age group.Reference Blackburn, Ajetunmobi, Mc Grath-Lone, Hardelid, Shafran and Gilbert2 Rates of stress-related presentations were highest in girls aged 14–15 years and consistently higher in term times than holiday periods for children of secondary school age in 2014–2015 to 2017–2018.Reference Blackburn, Ajetunmobi, Mc Grath-Lone, Hardelid, Shafran and Gilbert2
We aimed to investigate the impact of the pandemic and lockdown (including school closures) on rates of emergency hospital admissions with a stress-related presentation. Our study examined secondary school-aged children to better characterise the relationships between schools and stress.
We used Hospital Episode Statistics (HES) data for all adolescents aged 11 to 17 who were admitted to a National Health Service (NHS) hospital in England in the academic years (1 September–31 August) of 2018–2019, 2019–2020 and 2020–2021. These years were selected to cover pre-pandemic and pandemic periods. We excluded records for adolescents of unknown gender or who were not resident in England.
Measures of stress
HES data on hospital admissions capture data on up to 20 distinct diagnoses coded using ICD-10. Stress-related presentations were defined as emergency hospital admissions with a primary diagnosis relating to mental health, pain without a medical cause or potentially psychosomatic symptoms (e.g. fatigue), or where a diagnostic code indicating self-harm was recorded in any diagnostic position. Codes are outlined in supplementary Tables 1 and 2 (available at https://doi.org/10.1192/bjp.2022.113) and were grouped into mutually exclusive categories relating to mental health/behavioural or pain/somatic presentations.Reference Blackburn, Ajetunmobi, Mc Grath-Lone, Hardelid, Shafran and Gilbert2 Admissions were excluded if they related to maternity care or where the adolescent died prior to discharge, because coding may differ relative to those discharged alive.
We estimated weekly rates of admissions with a stress-related presentation (per 100 000 girls or boys) using Office for National Statistics mid-year population estimates for England stratified into 1-year age bands. Weekly rates of stress-related presentations were plotted and visualised against school term times, which were defined from a sample of school timetables published online by local authorities in England. Incident rate ratios (IRRs) for term- versus holiday-time occurrence of stress-related presentations were estimated using separate negative binomial regression models that were age- and gender-specific and adjusted for academic year (fitted as a categorical variable).
Approvals for the use of HES data were obtained from NHS Digital through the standard approval process. This study is exempt from NHS Research Ethics Committee approval and participant consent was not applicable because we analysed an existing data-set of de-identified data collected as part of routine care.
We identified a total of 662 000 emergency admissions for adolescents aged 11–17 years. Of these, 21 818 were excluded because they related to: exact duplicates (n = 260), adolescents not residing in England (n = 4049) or with unrecorded gender (n = 196), maternity care (n = 7873) or adolescents who died in hospital (n = 498). A further 8942 admissions with abdominal pain were excluded because a likely medical cause was recorded. We analysed the remaining 210 441 admissions with a stress-related presentation (32% of the 662 000 emergency admissions identified).
Figure 1 shows the weekly rate of emergency admissions in the three academic years examined. The rate of stress-related presentations was highest in girls, particularly those aged 14–15 years (supplementary Fig. 1). During the first lockdown there was a rapid decline in stress-related presentations – including the loss of term-time peaks – that slightly preceded the closure of schools on 20 March 2020. This was followed by a rebound in rates of stress-related presentations and resumption of term-time peaks to historic levels by July 2020. The small number of holiday periods during the pandemic (Easter and summer half-term 2020) when schools were closed limits scope for formal evaluation of changes in term versus holiday presentations.
Across the 3 years investigated, mean weekly term-time rates of stress-related presentations were higher than holiday-time rates for girls of all ages and boys aged 11–16 years (supplementary Figs 1 and 2). Compared with 2018–2019, incident rate ratios for weekly rates of stress-related presentations were 17–22% higher in 2020–2021 for girls aged 11–15, but were unchanged for boys and for girls aged 16–17 years (supplementary Fig. 3). A total of 43% (90 392) of stress presentations were classified as mental health/behavioural, with the remaining 57% (120 049) pain/somatic. The proportion of stress presentations classified as mental health/behavioural increased from 40% in 2018–2019 to 46% in 2020–2021 (chi P < 0.001).
Our main findings are that following the start of the pandemic, rates of hospital admission with stress-related presentations have increased substantially for girls aged 11–15 years, reflecting higher rates in both term-time and holiday periods. In boys, stress-related admission rates were much lower than for girls and remain at pre-pandemic levels. The observed increase for girls corroborates reports of declining levels of adolescent well-being exacerbated by insufficient support, as illustrated by a 94% increase in referrals to child and adolescent mental health services (CAMHS) in 2021 relative to 2019.Reference Viner, Russell and Saulle1,5
We found that term-time rates of stress-related admissions were higher than rates during holiday periods for girls and boys aged 11–16 years. This is similar to pre-pandemic data for EnglandReference Blackburn, Ajetunmobi, Mc Grath-Lone, Hardelid, Shafran and Gilbert2 and Ireland, where patterns of CAMHS referrals reduced during holiday times.Reference McNicholas, Kelleher and Hedderman6 Qualitative research is needed to determine whether the term-time peaks reflect schools as drivers of stress (e.g. through academic pressures or bullying) or whether schools increase identification of distressed pupils and facilitate help-seeking.
We also identified a rapid drop in stress-related presentations that preceded the closure of schools on 20 March 2020.Reference Leahy4 This is similar to that reported more widely, including a 50–65% decline in contacts for paediatric psychiatric and self-harm presentations.Reference Viner, Russell and Saulle1 This sharp decline most likely reflects reduced healthcare attendance (e.g. owing to fear of nosocomial COVID infection), rather than reduced need.
As our data on hospital admissions capture only the tip of the iceberg of stress manifestations, the true burden will be substantially greater and warrants ongoing action by researchers, schools and clinicians. Stress-related presentations accounted for almost one-third of all emergency hospital admissions in 11- to 17-year-olds. This indicates a real need for earlier intervention in other settings as well as interdisciplinary support at admission – potentially through mental health liaison services – to address the biopsychosocial factors that are likely contributing to the need for emergency care. Further research should use linked clinical data to examine referral patterns and service use associated with these stress presentationsReference Perera, Broadbent and Callard7 and determine individuals’ suitability for evaluations of mental health interventions, including those delivered through schools.
We gratefully acknowledge all young people whose de-identified data are used in this research.
R.M.B. conceived of the study, analysed the data and wrote the first draft of the manuscript. All authors made substantial contributions to the interpretation of the analysis and provided final approval of the version for publication. All authors are accountable for the accuracy and integrity of the work.
This research benefits from and contributes to the National Institute for Health Research (NIHR) Children and Families Policy Research Unit, but was not commissioned by the NIHR Policy Research Programme. Research at UCL Great Ormond Street Institute of Child Health is supported by the NIHR Great Ormond Street Hospital Biomedical Research Centre. R.M.B. is supported by a UK Research and Innovation Fellowship funded by grant MR/S003797/1 from the Medical Research Council. J.D. received support from a NIHR Clinician Scientist Fellowship (CS-2018-18-ST2-014) and Psychiatry Research Trust Peggy Pollak Research Fellowship in Developmental Psychiatry.
Declaration of interest
Viner, R, Russell, S, Saulle, R, Croker H, Stansfield C, Packer J, et al. School closures during social lockdown and mental health, health behaviors, and well-being among children and adolescents during the first COVID-19 wave: a systematic review. JAMA Pediatr 2022; 176: 400–9.CrossRefGoogle ScholarPubMed Blackburn, R, Ajetunmobi, O, Mc Grath-Lone, L, Hardelid, P, Shafran, R, Gilbert, R, et al. Hospital admissions for stress-related presentations among school-aged adolescents during term time versus holidays in England: weekly time series and retrospective cross-sectional analysis. BJPsych Open 2021; 7(6): e215.CrossRefGoogle ScholarPubMed Jensen, PS, Goldman, E, Offord, D, Costello EJ, Friedman R, Huff B, et al. Overlooked and underserved: “action signs” for identifying children with unmet mental health needs. Pediatrics 2011; 128(5): 970–9.CrossRefGoogle ScholarPubMed McNicholas, F, Kelleher, I, Hedderman, E, Lynch F, Healy E, Thornton T, et al. Referral patterns for specialist child and adolescent mental health services in the Republic of Ireland during the COVID-19 pandemic compared with 2019 and 2018. BJPsych Open 2021; 7(3): e91.CrossRefGoogle ScholarPubMed Perera, G, Broadbent, M, Callard, F, Chang CK, Downs J, Dutta R, et al. Cohort profile of the South London and Maudsley NHS foundation trust biomedical research centre (SLaM BRC) case register: current status and recent enhancement of an electronic mental health record-derived data resource. BMJ Open 2016; 6(3): e008721.CrossRefGoogle ScholarPubMed