Email updates

Keep up to date with the latest news and content from BMC Geriatrics and BioMed Central.

Open Access Highly Accessed Research article

A standard procedure for creating a frailty index

Samuel D Searle1, Arnold Mitnitski123, Evelyne A Gahbauer4, Thomas M Gill4 and Kenneth Rockwood125*

Author Affiliations

1 Geriatric Medicine Research Unit, Dalhousie University & Capital District Health Authority, Halifax, Canada

2 Department of Medicine, Dalhousie University, Halifax, Canada

3 Department of Mathematics & Statistics, Dalhousie University, Halifax, Canada

4 Department of Internal Medicine, Yale University School of Medicine, New Haven, CT 06504, USA

5 Division of Geriatric Medicine, Dalhousie University, Halifax, Canada

For all author emails, please log on.

BMC Geriatrics 2008, 8:24  doi:10.1186/1471-2318-8-24

Published: 30 September 2008

Abstract

Background

Frailty can be measured in relation to the accumulation of deficits using a frailty index. A frailty index can be developed from most ageing databases. Our objective is to systematically describe a standard procedure for constructing a frailty index.

Methods

This is a secondary analysis of the Yale Precipitating Events Project cohort study, based in New Haven CT. Non-disabled people aged 70 years or older (n = 754) were enrolled and re-contacted every 18 months. The database includes variables on function, cognition, co-morbidity, health attitudes and practices and physical performance measures. Data came from the baseline cohort and those available at the first 18-month follow-up assessment.

Results

Procedures for selecting health variables as candidate deficits were applied to yield 40 deficits. Recoding procedures were applied for categorical, ordinal and interval variables such that they could be mapped to the interval 0–1, where 0 = absence of a deficit, and 1= full expression of the deficit. These individual deficit scores were combined in an index, where 0= no deficit present, and 1= all 40 deficits present. The values of the index were well fit by a gamma distribution. Between the baseline and follow-up cohorts, the age-related slope of deficit accumulation increased from 0.020 (95% confidence interval, 0.014–0.026) to 0.026 (0.020–0.032). The 99% limit to deficit accumulation was 0.6 in the baseline cohort and 0.7 in the follow-up cohort. Multivariate Cox analysis showed the frailty index, age and sex to be significant predictors of mortality.

Conclusion

A systematic process for creating a frailty index, which relates deficit accumulation to the individual risk of death, showed reproducible properties in the Yale Precipitating Events Project cohort study. This method of quantifying frailty can aid our understanding of frailty-related health characteristics in older adults.