Risk factors for delirium in hospitalized elderly

• Published in: JAMA : the journal of the American Medical Association Vol. 267; no. 6; p. 827
• Main Authors: Schor, J D, Levkoff, S E, Lipsitz, L A, Reilly, C H, Cleary, P D, Rowe, J W, Evans, D A
• Format: Journal Article
• Language: English
• Published: United States 12.02.1992
• Subjects: Age Factors; Aged; Aged, 80 and over; Boston; Delirium - chemically induced; Delirium - etiology; Female; Hospitalization; Humans; Incidence; Male; Multivariate Analysis; Odds Ratio; Prospective Studies; Risk Factors; Boston
• ISSN: 0098-7484
• DOI: 10.1001/jama.1992.03480060073033

To determine risk factors for delirium in elderly hospitalized patients. Cohort analytic study. Using a reliable and valid instrument for detection of delirium, we prospectively followed up a cohort of elderly patients admitted to an acute care hospital. Using standardized criteria, we collected risk factor data from patient medical records. General medical and surgical wards of a tertiary-care hospital. Patients (n = 325) were 65 years of age or older, from either a geographically defined community or a long-term-care institution. We studied those patients (n = 291) not delirious on first evaluation. Fifty-seven patients or their families refused participation. Incidence of delirium and risk factors calculated as adjusted odds ratios (ORs). Delirium developed in 91 patients. By stepwise logistic regression, the independent risk factors for in-hospital delirium included prior cognitive impairment (OR, 8.97; 95% confidence interval [CI], 3.99 to 20.14), age over 80 years (OR, 5.22; 95% CI, 2.60 to 10.46), fracture on admission (OR, 6.57; 95% CI, 2.23 to 19.33), symptomatic infection (OR, 2.96; 95% CI, 1.42 to 6.15), and male sex (OR, 2.40; 95% CI, 1.19 to 4.84). Among medication groups, only neuroleptic use (OR, 4.48; 95% CI, 1.82 to 10.45) and narcotic use (OR, 2.54; 95% CI, 1.24 to 5.18) were independently associated with delirium. Anticholinergic use was not associated with delirium. Delirium in hospitalized patients is most closely associated with factors already present on admission such as prior cognitive impairment, advanced age, and fracture. In the hospital, use of neuroleptics and narcotics and the presence of infection are less strongly associated with this syndrome.