Current issues of ACP Journal Club are published in Annals of Internal Medicine


Depression at admission predicted mortality in elderly nursing-home patients

ACP J Club. 1991 Jul-Aug;115:23. doi:10.7326/ACPJC-1991-115-1-023

Source Citation

Rovner BW, German PS, Brant LJ, et al. Depression and mortality in nursing homes. JAMA. 1991 Feb 27;265:993-6.



To examine the relation between depression at nursing home admission and mortality in elderly patients.


Cohort study for 1 year.


8 Baltimore-area chain nursing homes, drawing patients from the surrounding, mainly white, middle-income communities, from 1987 to 1988.


562 consecutive, newly admitted patients who had not resided in nursing homes in the 6 months before their admission to the study facilities were eligible. 454 (81%) were enrolled and 14 patients (3%) were lost to follow-up. 61% of patients were between 65 and 84 years of age and 35% were older, 94% were white, and 77% were women.

Assessment of prognostic factors

Psychiatric examination included a semistructured clinical examination (Modified Present State Examination) and the Mini-Mental State Examination. Agreement on psychiatric diagnosis (according to American Psychiatric Association criteria) made by 3 psychiatrists on 8 patients was 89% above chance agreement (P < 0.001). Information on each patient's activities of daily living was obtained from independent interviews with patients (if possible), nursing staff, and family members. Demographic information and nursing home physicians' diagnoses at admission were obtained from records. Medical severity was estimated from the presence of immobility, incontinence, or pressure sores; hospitalization during the year; and the number of medical diagnoses at admission.

Main outcome measure

All-cause mortality.

Main results

67% of patients were demented at admission. 57 patients (13%) had depressive disorder, 82 patients (18%) had depressive symptoms, and 315 patients (69%) had no depression. Patients with depressive disorder and depressive symptoms were less often demented (P < 0.05), had higher cognitive function (P < 0.001), and were more capable of doing activities of daily living (P < 0.05). The recognition rate of depressive disorder was 14% and 65% for physicians and nurses, respectively; for depressive symptoms, it was 14% and 54%, respectively.

At 1 year, 141 patients (31%) died. 47% of those with depressive disorder died compared with 24% of those with depressive symptoms, and with 30% of those without depression (P < 0.05). In proportional hazards analysis, mortality was most strongly predicted by increased disability in activities of daily living (relative risk [RR] = 1.52, 95% CI 1.03 to 2.25, P < 0.001), male sex (RR = 1.88, CI 1.31 to 2.71, P < 0.001), depressive disorder (RR = 1.59, CI 1.02 to 2.51, P < 0.05), and hospitalization during the study year (RR = 1.44, CI 1.01 to 2.06, P < 0.05).


Elderly patients with a psychiatrist's diagnosis of depressive disorder at admission to a nursing home were more likely than nondepressed patients to die within a year.

Source of funding: Not stated.

Address for article reprint: Dr. B.W. Rovner, The Johns Hopkins Hospital, Osler 320, 600 North Wolfe Street, Baltimore, MD 21205, USA.


Depression is a treatable and potentially reversible condition. It produces morbidity and affects prognosis by compromising the person's motivation to overcome or adapt to the functional limitations imposed by illness. Rovner and colleagues showed the high prevalence of depression that exists among nursing-home patients and its significant, independent, adverse effect on survival. Depression is also common among hospitalized older patients and is associated with greater in-hospital mortality and a greater use of health care by those who did not die (1). The study by Rovner and colleagues also documents the disturbingly common failure of physicians to recognize depression. In one study of elderly outpatients, the likelihood of a geriatric fellow's clinical impression agreeing with a standardized depression screen was no better than chance (2). The results of the present study underscore the need for primary care physicians to systematically screen their elderly patients for depression, especially in high-prevalence settings such as nursing homes.

Whether depressed nursing-home patients will benefit from antidepressant therapy is unclear. The authors found no difference in survival among patients with depressive disorder who received or did not receive antidepressants, but the numbers were small and the types of medication used were not given. Because various comparatively safe and effective treatment modalities exist, physicians should not be reluctant to treat depression in nursing-home patients, including "situational depression," when the situation is permanent.

Calvin Hirsch, MD
University of California Davis Medical CenterSacramento, California, USA


1. Koenig HG, Shelp F, Goli V, Cohen HJ, Blazer DG. Survival and health care utilization in elderly medical inpatients with major depression. J Am Geriatr Soc. 1989;37:599-606.

2. Winograd CH, Gallagher D, Norris JT, Soderhom-Diafatte V. Depression: prevalence and use of screening instruments in geriatric outpatients [Abstract]. J Am Geriatr Soc. 1988;36:656.