Prevalence and correlates of clinically significant depressive symptoms in an urban hospital emergency department.

Abstract

In hospital settings, depression is an underdetected, undertreated, but prevalent and interfering illness that is associated with significant disability, morbidity, and mortality. A general hospital emergency department (ED) setting may be well suited to identify individuals with clinically significant depressive symptoms, facilitating their referral and treatment. Cross-sectional data of adult ED patients in a general hospital enrolling in a human immunodeficiency virus (HIV) screening study between February 2007 and March 2008 were analyzed. Data included demographic factors, the Center for Epidemiologic Studies Depression Scale (CES-D), alcohol and substance use history, sexual risk taking, and brief medical history. The primary outcome was a dichotomous measure of self-reported clinically significant depressive symptoms. Patients who scored >/= 16 on the CES-D were considered to screen positive for depressive symptoms. Of the 3,262 patients enrolled in the screening trial, 2,588 (79%) completed the survey between February 2007 and March 2008. Among these, 1,945 (75%) completed the psychosocial assessment battery; 596 (31%) survey completers screened positive for clinically significant depressive symptoms. In a multivariable model, female sex (RR = 1.36; 95% CI, 1.16-1.57), being unemployed (RR = 1.61; 95% CI, 1.32-1.93), and lower annual income (RR from 1.73 to 2.24) were associated with increased rates of clinically significant depressive symptoms (CES-D score >/= 16). Clinically significant depressive symptoms were more often present in patients who screened positive for alcohol dependence (RR = 1.48; 95% CI, 1.19-1.78), individuals reporting current smoking (RR = 1.39; 95% CI, 1.17-1.62), those with a prior psychiatric disorder diagnosis (RR = 2.20; 95% CI, 1.80-2.57) or history of hypertension (RR = 1.47; 95% CI, 1.18-1.79), and those who reported ever having sex with an HIV-infected partner (RR = 1.58; 95% CI, 1.08-2.09). This study identified a high frequency of undiagnosed clinically significant depressive symptoms among ED patients and highlights several new demographic and clinical correlates of such symptoms. Screening for and diagnosis of clinical depression in the ED setting introduces important opportunities for acute care providers and for the health care system as a whole.