Development and validation of a prognostic index for 2‐year mortality in Chinese older residents living in nursing homes

Abstract

There is no mortality prediction index for Chinese nursing home older residents. The objective of this study was to derive and validate a 2-year mortality prognostic index for them. We carried out a prospective cohort study on 1120 older residents from 12 nursing homes of Hong Kong. We obtained potential predictors of mortality and carried out updated functional assessment. Each risk factor associated independently with 2-year mortality in a derivation cohort was assigned a score based on the odds ratio, and risk scores were calculated for each participant by adding the points of risk factors present. Similar analysis was carried out on the validation cohort. Independent predictors of mortality included: aged 86-90 years (3 points); aged ≥ 91 years (4 points); Charlson comorbidity index ≥ 4 (6 points); Barthel Index 5-60 (5 points); Barthel Index 0 (10 points); number of hospitalizations in the preceding year (Adbefore) 1 (4 points); Adbefore 2 (5 points) and Adbefore ≥ 3 (6 points). In the derivation cohort, 2-year mortality was 10.8% in the low-risk group (≤ 4 points) and 59.9% in the high-risk group (≥ 14 points). In the validation cohort, 2-year mortality was 11.8% in the low-risk group and 60.4% in the high-risk group. The receiver-operator characteristic curve area was 0.761 for the derivation cohort and 0.742 for the validation cohort. Our prognostic index had satisfactory discrimination and calibration in an independent sample of Chinese nursing home older residents. It can be used to identify older residents with a high risk for poor outcomes, who need a different level of care.