PROXY-PERCEIVED PRIOR HEALTH-STATUS AND HOSPITAL OUTCOME AMONG THE CRITICALLY ILL - IS THERE ANY RELATIONSHIP

Objectives: To measure the health status of critically ill patients pr ior to hospital admission and to study the relationship between prior health status (PHS) and hospital mortality. Design: 523 patients admit ted to the intensive care department from October 1994 to June 1995 we re in eluded consecutively in the study. Health status 3 months prior to admission was assessed retrospectively by proxies using the EuroQol 5D (EQ-5D) and the Karnofsky Performance Status Scale (KF). Patients were classified into four admission categories: trauma injury, schedul ed surgery, unscheduled surgery and other medical conditions. Setting: Department of Intensive Medicine, University Hospital of Bellvitge, B arcelona, Spain. Patients: 84 trauma injury patients, 239 scheduled su rgery patients, 57 unscheduled surgery patients and 143 patients with other medical conditions. Interventions: The descriptive system and vi sual analogue scale (VAS) of the EQ-SD and the K.F Measurements anti m ain results: Using proxy responses we found that trauma injury patient s had the best PHS and scheduled surgery patients the worst. There wer e statistically significant differences in mean VAS scores and all EQ- 5D dimensions, except self-care, when trauma injury patients or schedu led surgery patients were compared with the other admission categories . No significant differences were found on these variables between uns cheduled surgery patients and medical patients. We found no statistica lly significant differences in PHS health status between patients who died and those who survived, either within each admission category or in the sample as a whole. Conclusions: The PHS of critically ill patie nts varied according to admission category. Given the instruments used and population studied, there was no association between PHS and hosp ital outcome.