Diseases/function/self-health in risk indices for elders
Funded Grant
Overview
Affiliation
View All
Overview
description
DESCRIPTION (provided by applicant): Prognostic (case-mix/risk adjustment) indices classify subjects into varying levels of outcome risk by combining information about multiple risk factors. Such indices have numerous uses in clinical medicine, policy evaluation, and health services/epidemiologic research. Our goal is to develop and validate prognostic indices for mortality and functional decline in older subjects using health information obtained from subject report. Our study will produce prognostic indices that will be of use to clinicians, investigators, and policy makers, enhance our understanding of the role different health measures can play in these indices, and compare different approaches to developing indices. The specific aims of our proposal are to (1) develop and validate a prognostic index using information about comorbidity and functional status that differentiates between subjects at high and low risk of 4-year mortality; (2) develop and validate a prognostic index using data about comorbidity and functional status that differentiates between subjects at high and low risk of 4-year decline in Activity of Daily Living (ADL) function; (3) examine the extent to which global self-rated health improves the ability of these indices to predict mortality and functional decline; and (4) examine whether alternative approaches of developing the indices in aim 1 and aim 2 (with an emphasis on tree based methods) improves the accuracy of the indices. We will conduct this study using baseline data from 20,451 subjects (age >50) enrolled in the 1998 wave of the Health and Retirement Study (HRS) and 4-year outcome data from the 2002 wave. HRS is a population-based study of older Americans with extensive data on baseline risk factors. We will develop our indices in 13,489 subjects in one geographic region, and validate the index 6962 subjects in another region. The results of our study will be useful to clinicians who need prognostic information to counsel patients or identify high risk patients for intervention, policy makers who need indices to compare outcomes across providers or systems of care, and investigators who need pragmatic and accurate methods of assessing baseline risk in observation studies.
