Measuring disability with parsimony: evidence for the utility of a single item.

Abstract

To provide evidence for the utility of a single item measure designed to quantify disability in the past 30 days. Australian data from studies comprising a community-dwelling elderly sample (N = 328) and a sample of chronic osteoarthritis (OA) patients (N = 119) are reported. Degree of disability was classified as 0-30 days, 0 vs. 1 or more days, and 0-9 vs. 10 or more days. Associations between disability and a range of demographic, health-related, and psychological variables were assessed. Participants with OA reported a significantly higher level of disability than participants from the community sample regardless of how disability was classified. Modest levels of association were noted between the number of disability days (0-30) and both health and psychological indices in both samples. The three alternative classifications of disability made little difference to the significance of associations. On balance, the comparison of 0-9 vs. 10 or more days appeared the more predictive classification of disability. The performance of the single item measure of disability was generally very satisfactory. Future investigations into the clinical application of the item across a range of patient groups are encouraged.