Clustering patients according to health perceptions -: Relationships to psychosocial characteristics and medication nonadherence

Background: Little is known about how patients rate their health perceptions. Our objectives were to identify systematic multivariate patterns of perceptions using cluster analysis, and to investigate associations among the clusters, psychosocial characteristics and medication nonadherence. Methods: Demographic, clinical and psychosocial data on 200 patients with inflammatory bowel disease (IBD) were collected prior to the index office visit and health perceptions were collected afterwards. Cluster analysis using a k-means method was used to identify subgroups of patients based on their responses to the Patient-physician Discordance Scales (PPDS), an instrument that assesses perceptions of health status and of the clinical visit. Results: We identified five different patient groups: a healthy, not distressed, good communication, low expectation for medication/testing group; a healthy, relatively distressed, good communication, high expectation for medication, low expectation for testing group; a symptomatic, distressed, good communication, high expectation for medication/testing group; a healthy, not distressed, good communication, high expectation for medication/testing group; and a relatively healthy, relatively distressed, poor communication, low expectation for medication/testing group. After adjustment for age, sex, language, form of IBD, and disease activity, statistically significant between-clusters differences were found in psychological distress, social support satisfaction and medication nonadherence. Conclusions: Distinct patterns of patients' health perceptions correlated with psychological health and adherence to treatment. This categorization may be used to help identify patients at higher risks for ineffective communication and nonadherence to medication. (C) 2004 Elsevier Inc. All rights reserved.