Will My Patient Fall?

Published in: The Journal of the American Medical Association, v. 297, no. 1, Jan. 3, 2007, p. 77-86

Posted on RAND.org on January 01, 2007

by David A. Ganz, Yeran Bao, Paul G. Shekelle, Laurence Rubenstein

Read More

Access further information on this document at jama.ama-assn.org

This article was published outside of RAND. The full text of the article can be found at the link above.

CONTEXT: Effective multifactorial interventions reduce the frequent falling rate of older patients by 30% to 40%. However, clinical consensus suggests reserving these interventions for high-risk patients. Limiting fall prevention programs to high-risk patients implies that clinicians must recognize features that predict future falls. OBJECTIVE: To identify the prognostic value of risk factors for future falls among older patients. DATA SOURCES AND STUDY SELECTION: Search of MEDLINE (1966-September 2004), CINAHL (1982-September 2004), and authors' own files to identify prospective cohort studies of risk factors for falls that performed a multivariate analysis of such factors. DATA EXTRACTION: Two reviewers independently determined inclusion of articles and assessed study quality. Disagreements were resolved by consensus. Included studies were those identifying the prognostic value of risk factors for future falls among community-dwelling persons 65 years and older. Clinically identifiable risk factors were identified across 6 domains: orthostatic hypotension, visual impairment, impairment of gait or balance, medication use, limitations in basic or instrumental activities of daily living, and cognitive impairment. DATA SYNTHESIS: Eighteen studies met inclusion criteria and provided a multivariate analysis including at least 1 of the risk factor domains. The estimated pretest probability of falling at least once in any given year for individuals 65 years and older was 27% (95% confidence interval, 19%-36%). Patients who have fallen in the past year are more likely to fall again [likelihood ratio range, 2.3-2.8]. The most consistent predictors of future falls are clinically detected abnormalities of gait or balance (likelihood ratio range, 1.7-2.4). Visual impairment, medication variables, decreased activities of daily living, and impaired cognition did not consistently predict falls across studies. Orthostatic hypotension did not predict falls after controlling for other factors. CONCLUSIONS: Screening for risk of falling during the clinical examination begins with determining if the patient has fallen in the past year. For patients who have not previously fallen, screening consists of an assessment of gait and balance. Patients who have fallen or who have a gait or balance problem are at higher risk of future falls.

Research conducted by

This report is part of the RAND Corporation external publication series. Many RAND studies are published in peer-reviewed scholarly journals, as chapters in commercial books, or as documents published by other organizations.

The RAND Corporation is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.