Screening for depression in patients in long-term care facilities: A randomized controlled trial of physician response

OBJECTIVES: To determine the effect of a screening protocol using the Geriatric Depression Scale (GDS) on the frequency of primary care physicians' decisions to prescribe drug therapy or refer long-term care patients with possible depression to mental health care. DESIGN: Case-finding phase, followed by a randomized controlled trial of the effect of a physician-targeted intervention on antidepressant prescribing or referral to mental health services. SETTING: Twenty-two nonacademic long-term care facilities. PARTICIPANTS: One hundred three of 1,602 patients aged 65 and older who met criteria for cognitive function and untreated symptoms of depression. INTERVENTION: The 77 physicians of these patients were randomized as clusters into an early notification (experimental) or a delayed notification (control) group. MEASUREMENTS: Frequency of physician response (mental health consult or antidepressant therapy) at 4 and 8 weeks from notification, physician follow-up, and factors associated with physician response. RESULTS: Frequency of physician response in the early group (25%) was greater than in the delayed group (2%) (P <.005) 4 weeks from baseline. Physician response rate when the groups were combined was 36% (95% confidence interval (CI) = 26%-46%) 8 weeks from notification. Overall, there was evidence of physician action after letters of notification in 69% (95% CI = 60%-78%) of cases. Univariate logistic regression suggested that physicians' decisions were primarily associated with physician-related characteristics. CONCLUSIONS: Screening of long-term care patients for depression can increase the frequency of treatment or referral by primary care physicians.