Comprehensive geriatric assessment: Toward understanding its efficacy

L. Z. Rubenstein, A. L. Siu, D. Wieland

Research output: Contribution to journalReview articlepeer-review

91 Scopus citations


Comprehensive geriatric assessment (CGA) offers health care professionals a technique for multidimensional diagnosis of frail elderly people to plan medical, psychosocial, and rehabilitative care. In the present paper, we provide a brief history of geriatric assessment, a description of the varied organization of geriatric assessment programs (GAPs), and a review of published effectiveness studies of programs worldwide performing comprehensive geriatric assessment. Program diversity has complicated drawing conclusions about the efficacy of CGA from a literature reporting generally positive, but not uniformly significant, results. We suggest that sample size limitations explain much of the variability in findings. Using the techniques of meta-analysis, we evaluate the effect of GAPs on mortality when all controlled trials are considered cumulatively. Meta-analysis of six-month mortality demonstrates a statistically significant 36% reduction of mortality for inpatient CGA programs (odds ratio = 0.64; 95% confidence interval = 0.50 to 0.83), and a 32% mortality reduction for all CGA programs (odds ratio = 0.68; 95% confidence interval = 0.57 to 0.80). Further use of meta-analytic techniques can be employed to clarify the effect of GAPs on other important outcomes (e.g., reduced hospital and nursing home use, improved functional status), and to identify program characteristics best promoting these benefits.

Original languageEnglish
Pages (from-to)87-98
Number of pages12
JournalAging clinical and experimental research
Issue number2
StatePublished - Dec 1989
Externally publishedYes


  • Geriatric medicine
  • functional assessment
  • meta-analysis
  • outcome studies
  • rehabilitation


Dive into the research topics of 'Comprehensive geriatric assessment: Toward understanding its efficacy'. Together they form a unique fingerprint.

Cite this