Technology assessment using the association between outcome measures and patterns of illness severity

The inclusion of a patient's illness experience as outcome in the assessment of health care technology has revealed methodological limitations such as the interpretation of multi-attribute scores and lack of knowledge about the association between illness and disease information. In an attempt to overcome these limitations, a cross-sectional study is performed to search for patterns of illness severity and investigate the association between illness measures and between illness patterns and disease factors. A sample of 211 patients with ulcerative colitis is studied using the sickness impact profile (SIP) and the rating form for inflammatory bowel disease patient concerns (RFIPC) as illness measures. SIP and RFIPC scores show low association, suggesting that they provide complementary information about the patient's illness status. Cluster analysis is performed using the two measures of illness separately to identify groups of patients with different degrees of severity of illness (clusters). The cluster description covers illness, disease and social and demographic variables. The RFIPC clusters show a general pattern of ascendant rank scores for the RFIPC items. SIP clusters differ, not only in the level of severity, but also in specific types of disability. The patients in the clusters with the highest degree of disability (reflected by SIP) show a non-linear relationship with patients' concerns (reflected by RFIPC) and disease factors.