Predictors of Immediate and 6‐Month Outcomes in Hospitalized Elderly Patients

Prakash Narain, Laurence Z. Rubenstein, G. Darryl Wieland, Brad Rosbrook, L. Sandra Strome, Fern M. Pietruszka, John E. Morley

Journal of the American Geriatrics Society · 1988 · 335 citations · 21 references

Concepts

TL;DR

The study prospectively examined hospitalized elderly VA patients to identify admission factors predicting in‑hospital and 6‑month mortality, nursing home admission, length of stay, and rehospitalization. The study included 396 patients aged 70 or older admitted to acute‑care medical wards over one year. Decreased functional status, mental status, and admitting diagnosis predicted 6‑month mortality, while decreased functional status, living location, and mental status predicted nursing home admission, and functional status was a stronger predictor of length of stay, mortality, and nursing home placement than diagnosis, indicating its utility for discharge planning and geriatric program targeting.

Abstract

This article presents results of a prospective multivariate study of hospitalized elderly patients at an acute‐care Veterans Administration (VA) hospital to identify factors on hospital admission predictive of several short‐ and long‐term outcomes: in‐hospital and 6‐month mortality, immediate and delayed nursing home admission, length of hospital stay, and 6‐month rehospitalization. All patients aged 70 years and over admitted to acute‐care beds on the medical service wards during a 1‐year period were included in the study (N = 396). Factors most predictive of 6‐month mortality (using logistic regression) were decreased functional status, admitting diagnosis, and de creased mental status. Factors most predictive of nursing home admission were decreased functional status, living location, and decreased mental status. Functional status was a stronger predictor of length of stay, mortality, and nursing home placement than was principal admitting diagnosis — of relevance to the current emphasis on diagnosis‐related groups (DRGs). These data may be helpful in improving discharge planning, in resource allocation, and in targeting patients for different specialized geriatric programs.

References

21