The overall goal of this project is to understand the underlying factors that influence a respondent's decision to identify a disease as a major source of health status problems. To do so, this project investigates the diverse factors that contribute to causal attribution decisions made by individuals with arthritis. In order to understand this phenomenon, four specific aims will be addressed. First, using a computerized diagnosis database from a central Massachusetts Health Maintenance Organization, the Fallen Community Health Plan, a sample of 1200 individuals with OA and 500 with RA will be surveyed. Sampling will be stratified within each disease category in order to obtain comparable members of subjects with no comorbidities and subjects with at least one comorbid condition. Second, data on health status, causal attribution of health status problems, socio- demographics, and social and psychological factors will be collected by mailed questionnaires from members of each stratum. Third, physicians will be asked to rate the severity of OA or RA for each respondent. Fourth, associations between arthritis and explanatory factors will be analyzed. Analysis will consist of univariate descriptors of study participants along the array of variables under study. Respondent, nonrespondent comparisons will be made. Bivariate correlations of attribution with various disease, socio-demographic, social and psychological variables will then be made. This will be followed by recursive partitioning of all of the explanatory variables using Classification and Regression Tree (CART) analysis, as interaction between explanatory variables is expected. The results of this recursive partitioning will provide insight into the complexity and the strengths of the associations between significant explanatory variables and attribution decisions concerning lower extremity problems, upper extremity problems and pain, within disease groups and within specific sample sub groups. This project is significant in that it will address an important issue in the area of arthritis health status assessment, - the analysis of factors that may influence causal attribution decisions - precisely at a time when the use of attribution responses is being advocated by researchers. To improve the relative sensitivity and precision of health status measures in identifying arthritis-specific problems, attribution questions have been added to these measures. The results of this research will allow for a more accurate interpretation of health status outcome assessments when causal attribution data is used to adjust health status scores.
Showing the most recent 10 out of 178 publications
