Predictive Validity of a Questionnaire That Identifies Older Persons at Risk for Hospital Admission

Publisher Website
Google Scholar
Abstract
To determine the predictive validity of a mailed questionnaire designed to measure older adults' risk of repeated hospitalization. Prospective cohort study. Ramsey County, Minnesota. Medicaid recipients aged 70 and older who completed the questionnaire. Responses were used to calculate the subjects' probability of repeated admission (Pra) to hospitals within 4 years. Subjects were classified as low-risk (Pra < 0.5) or high risk (Pra > or = 0.5). One year later, Medicaid claims data were analyzed to determine the subjects' actual use of hospitals. One-fifth of the respondents (20.6%) were classified as high-risk at baseline. During the following year, the high-risk subjects used hospitals at approximately twice the rate of the low-risk subjects (4.5 vs 2.4 days/person-year, P = .009). The instrument, which was previously found to be valid in a national sample of Medicare beneficiaries, appears to be valid also in a local sample of Medicaid beneficiaries. Older adults at risk of heavy hospital use can be identified prospectively through their responses to this brief, mailed, self-administered questionnaire. The instrument may be useful in targeting older persons for interventions designed to prevent the need for hospital care.