Psychiatric Emergencies Following the 2008 Economic Recession: An Ecological Examination of Population-Level Responses in Four US States.

Background: Research examining mental health outcomes following economic downturns finds both pro-cyclic and counter-cyclic associations. Pro-cyclic findings (i.e. economic downturns correspond with decline in illnesses) invoke increase in leisure time and risk-averse behavior as underlying drivers of reduction in harmful consumption during economic recessions. By contrast, counter-cyclic evidence (i.e. economic downturns correspond with increase in illnesses) suggests increase in mental illness with economic decline owing to heightened stress and loss of resources, particularly among certain age and socioeconomic groups.

Aim of the study: To examine the relation between monthly aggregate employment decline and psychiatric emergency department visits across 96 counties within 49 Metropolitan Statistical Areas in the United States.

Methods: For this study, data on all psychiatric outpatient Emergency Department (ED) visits for 4 US states (Arizona, California, New Jersey and New York) were retrieved from the State Emergency Department Database (SEDD) and aggregated by county-month, for the time period of 2006 to 2011. Exposure to recession was operationalized as population-level employment change in a Metropolitan Statistical Area (MSA). This information was obtained from MSA-level employment provided by the US Bureau of Labor Statistics. Brief exposure time lags of 0 to 3 months were specified to estimate proximate responses to MSA-level economic decline. Income level was approximated based on insurance status (private insurance= high-income, public insurance = low-income). Linear regression analysis was used to test whether monthly decline in aggregate employment in an MSA corresponds with (i) changes in population rates of psychiatric ED visits and (ii) whether the relation between the outcome and exposure varies by insurance status (private, public) and age group (children, age < 20 years; working-age adults, age 20 to 64 years; elderly adults, age > 64 years). Regression methods controlled for region, year and month fixed effects, and state-specific linear time trends.

Results: Linear regression results indicate that overall, psychiatric ED visits (per 100,000 population) decline with decline in monthly employment at exposure lag 0 (coefficient: 0.54, p < 0.001) and lag 2 (coefficient: 0.52, p < 0.001). Privately insured (high-income) groups also show a decline in psychiatric ED visits following decline in aggregate employment. Conversely, publicly insured children show an increase in psychiatric ED visit rates one month (i.e. lag 1) following employment decline (coefficient: -0.35, p value < 0.01). Exploratory analyses by disorder groups show that the population-level decline in psychiatric ED visits concentrates among visits for alcohol use disorders at 0, 1 and 2 month lags of employment decline.

Discussion: This study's findings provide evidence of pro- as well as counter-cyclic trends in psychiatric emergency visits following aggregate employment decline in an MSA. Whereas declines in psychiatric emergencies support a risk-averse response to economic recessions, these aggregate trends may mask countervailing trends among vulnerable groups. Limitations of this study include the absence of sex-specific analyses and lack of information on emergent or non-emergent nature of psychiatric ED visits.

Implications for health care provision and use: Psychiatric ED visits during recessions may vary by age and income groups.

Implications for health policies: Findings from this study may serve to develop targeted policies for low-income groups during macroeconomic downturns.

Implications for further research: Future research may examine trends in emergent versus non-emergent psychiatric ED visits following economic recessions.