Acute mental health service use following onset of the COVID-19 pandemic in Ontario, Canada: a trend analysis

Abstract

Background: The extent to which heightened distress during the COVID-19 pandemic translated to increases in severe mental health outcomes is unknown. We examined trends in psychiatric presentations to acute care settings in the first 12 months after onset of the pandemic.

Methods: This was a trends analysis of administrative population data in Ontario, Canada. We examined rates of hospitalizations and emergency department visits for mental health diagnoses overall and stratified by sex, age and diagnostic grouping (e.g., mood disorders, anxiety disorders, psychotic disorders), as well as visits for intentional self-injury for people aged 10 to 105 years, from January 2019 to March 2021. We used Joinpoint regression to identify significant inflection points after the onset of the pandemic in March 2020.

Results: Among the 12 968 100 people included in our analysis, rates of mental health-related hospitalizations and emergency department visits declined immediately after the onset of the pandemic (peak overall decline of 30% [hospitalizations] and 37% [emergency department visits] compared to April 2019) and returned to near prepandemic levels by March 2021. Compared to April 2019, visits for intentional self-injury declined by 33% and remained below prepandemic levels until March 2021. We observed the largest declines in service use among adolescents aged 14 to 17 years (55% decline in hospitalizations, 58% decline in emergency department visits) and 10 to 13 years (56% decline in self-injury), and for those with substance-related disorders (33% decline in emergency department visits) and anxiety disorders (61% decline in hospitalizations).

Interpretation: Contrary to expectations, the abrupt decline in acute mental health service use immediately after the onset of the pandemic and the return to near prepandemic levels that we observed suggest that changes and stressors in the first 12 months of the pandemic did not translate to increased service use. Continued surveillance of acute mental health service use is warranted.

Figure 1:

Emergency department visits and hospitalizations related to mental health and substance use disorders per 1000 population aged 10 to 105 years. Joinpoint regression analysis identified a statistically significant change in slope at July 2020 (*p ≤ 0.05) among overall and male visits to emergency departments. We observed a decrease in hospitalizations related to mental health and substance use disorders for both sexes in April 2020. Joinpoint regression analysis did not yield any significant changes in slope. Note: ED = emergency department.

Figure 2:

Emergency department visits for intentional self-injury, as well as admission rates and ICU or death rates per 10 000 population aged 10 to 105 years. We observed a decrease in emergency department visits for intentional self-injury in April 2020 among those who were admitted following self-injury and among those who were transferred to the ICU or died in the emergency department. Joinpoint regression analysis identified a statistically significant change in slope in July 2020 (*p ≤ 0.05) among those who were transferred to the ICU or died in the emergency department. Note: ED = emergency department, ICU = intensive care unit.

Figure 3:

(A) Emergency department visits and (B) hospitalizations related to mental health and substance use disorders per 1000 population aged 10 to 105 years, by age. For emergency department visits, joinpoint regression analysis identified statistically significant changes in slope in April 2020 (*p ≤ 0.05) for those aged 10 to 17 years and in July 2020 for those aged 18 to 21 years and those aged 25 to 105 years. For hospitalizations, we observed a decrease among all age groups in April 2020. Joinpoint regression analysis did not yield any significant changes in slope.

Figure 4:

(A) Emergency department visits and (B) hospitalizations related to mental health and substance use disorders per 1000 population aged 10 to 105 years, by diagnosis. Joinpoint regression analysis identified statistically significant changes in slope in July 2020 (*p ≤ 0.05) in emergency department visits for anxiety, schizophrenia spectrum and other psychotic disorders, and trauma and stressor-related disorders. We observed a decrease in hospitalizations across all diagnostic categories in April 2020, but joinpoint regression analysis did not yield any significant changes in slope.