The impact of the COVID-19 school closure on adolescents' use of mental healthcare services in Sweden

BackgroundSchool closures used to contain the COVID-19 pandemic may have negative impacts on students mental health but credible evidence is scarce. Sweden moved upper-secondary students to remote learning but, as the only country in the OECD, kept schools at lower levers open throughout the pandemic. MethodsUsing nationwide register data, we estimate the difference in the use of mental healthcare services between upper- and lower-secondary students during the pandemic, and relate this to the same group difference prior to the pandemic. For the main analysis, this difference-in-difference approach is applied to the period April-June 2020 when upper-secondary schools were fully closed. We also study the periods July-December 2020 when upper-secondary schools were largely open, and January-March 2021 when they were partially closed. We study the impact on all contacts with hospitals and specialist psychiatric care due to mental and behavioral disorder, as well as prescriptions for antidepressants, insomnia, and ADHD drugs. FindingsCompared with expected rates, the use of mental healthcare services among upper-secondary students fell by -3.71 [CI95 -5.52 to -1.91] cases per 1000 during April-June 2020, largely due to a reduction in depression and anxiety-related diagnoses (-1.49; CI95 [-2.36 to -0.63]) and prescriptions (-1.80; CI95 [-2.93 to -0.68]). This reduction in the use of mental healthcare services corresponds to a 4.36% CI95 [-6.47 to -2.25]) decrease compared to the level prior to the pandemic. The decrease compared to expected rates persists through July-December 2020 (-3.55%; CI95 [-5.38 to -1.71]) and January-March 2021 (-5.23%; CI95 [-7.24 to -3.21]). The reduction is stronger among students in the 2nd (-5.06%; CI95 [-8.02 to -2.09]) and 3rd (-4.86%; CI95 [-8.19 to -1.53]) year of upper-secondary school. The decrease is concentrated to students who was not in contact with mental healthcare services earlier in the academic year (-16.70%; CI95 [-22.20 to -11.20]). The relative reduction is largest for unplanned care (-13.88%; CI95 [-19.35 to -8.42]) and care at emergency units (-18.19%; CI95 [-26.44 to -9.92]). InterpretationClosing upper-secondary schools in Sweden reduced use of mental healthcare services. There is no indication of this being due to reduced accessibility. In a setting with no strict lockdown, moving to online teaching for a limited period did not worsen mental health among students in upper-secondary schools.