Performance of two measures of general health-related quality of life, the EQ-5D and the RAND-36 among critically ill patients.

Abstract

To compare two health-related quality of life measures, the preference-based EQ-5D with five questions and the profile-based RAND-36 with 36 questions, in previous critically ill patients. Prospective observational study. A ten-bed medical-surgical intensive care unit (ICU) in a tertiary care university hospital. Of the 2,709 critically ill patients, treated during the years 1995-2000, the 1,099 patients of the 1,443 still alive who returned both mailed measures were included in the study. None. The EQ-5D and the RAND-36 correlated well (P <.001). Ceiling effect was more obvious with the EQ-5D; the values of the RAND-36 varied usually from 0 to 100 in all the three levels of the corresponding EQ-5D question, and the weakest statistically significant differences were between the EQ levels 2 and 3. In particular, the RAND-36 proved to differentiate better the levels of mobility, self-care, and poor outcome. The EQ-5D and the RAND-36 correlated well, but when more precisely stated information is needed, especially regarding mobility, self-care, or low quality of life levels of previous critically ill patients, the profile-based RAND-36 may discriminate better.