The effect of the COVID-19 lockdown on mental health care use in South Africa: an interrupted time-series analysis

Abstract

Aims: The coronavirus disease 2019 (COVID-19) pandemic and ensuing restrictions have negatively affected the mental health and well-being of the general population, and there is increasing evidence suggesting that lockdowns have led to a disruption of health services. In March 2020, South Africa introduced a lockdown in response to the COVID-19 pandemic, entailing the suspension of all non-essential activities and a complete ban of tobacco and alcohol sales. We studied the effect of the lockdown on mental health care utilisation rates in private-sector care in South Africa.

Methods: We conducted an interrupted time-series analysis using insurance claims from 1 January 2017 to 1 June 2020 of beneficiaries 18 years or older from a large private sector medical insurance scheme. We calculated weekly outpatient consultation and hospital admission rates for organic mental disorders, substance use disorders, serious mental disorders, depression, anxiety, other mental disorders, any mental disorder and alcohol withdrawal syndrome. We calculated adjusted odds ratios (OR) for the effect of the lockdown on weekly outpatient consultation and hospital admission rates and the weekly change in rates during the lockdown until 1 June 2020.

Results: 710 367 persons were followed up for a median of 153 weeks. Hospital admission rates (OR 0.38; 95% confidence interval (CI) 0.33-0.44) and outpatient consultation rates (OR 0.74; 95% CI 0.63-0.87) for any mental disorder decreased substantially after the introduction of the lockdown and did not recover to pre-lockdown levels by 1 June 2020. Health care utilisation rates for alcohol withdrawal syndrome doubled after the introduction of the lockdown, but the statistical uncertainty around the estimates was large (OR 2.24; 95% CI 0.69-7.24).

Conclusions: Mental health care utilisation rates for inpatient and outpatient services decreased substantially after the introduction of the lockdown. Hospital admissions and outpatient consultations for alcohol withdrawal syndrome increased after the introduction of the lockdown, but statistical uncertainty precludes strong conclusions about a potential unintended effect of the alcohol sales ban. Governments should integrate strategies for ensuring access and continuity of essential mental health services during lockdowns in pandemic preparedness planning.

Keywords: Alcohol abuse; health service research; psychiatric hospital; psychiatric services.

Fig. 1.

Interrupted time-series analysis for changes in hospital admissions during the lockdown. Solid grey lines represent percentages of the study population admitted for the condition in each week between 1 January 2020, and 1 June 2020. Solid red lines depict the estimated average percentage admitted per week with 95% confidence intervals (grey shaded areas). Area between the dashed grey and dashed red line: data not used to account for anticipatory behaviour. Dashed red line: first Monday during lockdown level 5 (30 March 2020). Dashed black line: beginning of lockdown level 4 (30 April 2020). Lockdown: odds ratios (OR) for the immediate effect of the lockdown on admission rates. Weekly change: OR for the weekly change in the odds of hospital admission during the lockdown (week 14–22 in 2020). 95% confidence intervals for ORs in parentheses.

Fig. 2.

Interrupted time-series analysis for changes in outpatient consultations during the lockdown. Solid grey lines represent percentages of the study population consulting outpatient care for the condition in each week between 1 January 2020, and 1 June 2020. Red lines depict the estimated average percentage consulting outpatient care per week with 95% CIs (grey shaded areas). Area between the dashed grey and dashed red line: data not used to account for anticipatory behaviour. Dashed red line: first Monday during lockdown (30 March 2020). Dashed black line: beginning of lockdown level 4 (30 April 2020). Lockdown: odds ratios (OR) for the immediate effect of the lockdown on outpatient consultation rates. Weekly change: OR for the weekly change in the odds of outpatient consultation rates during the lockdown (week 14–22 in 2020). 95% confidence intervals for ORs in parentheses.

Fig. 3.

Interrupted time-series analysis for changes in mental health care use during the lockdown. Solid grey lines represent percentages of the study population admitted to a hospital or consulting outpatient care for condition in each week between 1 January 2020, and 1 June 2020. Red lines depict the estimated average percentage admitted or consulting outpatient care for the condition in a week with 95% CIs (grey shaded areas). Area between the dashed grey and dashed red line: data not used to account for anticipatory behaviour. Dashed red line: first Monday during lockdown (30 March 2020). Dashed black line: beginning of lockdown level 4 (30 April 2020). Lockdown: odds ratios (OR) for the immediate effect of the lockdown on hospital admission and outpatient consultation rates. Weekly change: OR for the weekly change in the odds of hospital admission and outpatient consultation rates during the lockdown (week 14–22 in 2020). 95% confidence intervals for ORs in parentheses.