Predicting outcome from intensive care for patients with rheumatologic diseases.

Abstract

To evaluate 2 prognostic scoring systems in patients with an underlying rheumatologic diagnosis admitted to an intensive care unit (ICU). A retrospective case series review, carried out at a medical ICU in a military referral hospital. All adult ICU admissions with a known rheumatologic diagnosis were evaluated during 28 consecutive months. There were 48 ICU admissions available for review in 36 patients (1.33 ICU admissions/patient) during the study period. All patients were assigned an APACHE II and TISS score based on the first 24 h of ICU admission. Eleven ICU admissions resulted in patient death (22.9%) and the remaining 37 admissions (77.1%) in patient survival and hospital discharge. Overall patient mortality was 30.6% for the study population. The APACHE II and TISS scores were each significantly different for survivor and nonsurvivor subgroups (APACHE II p less than 0.0001; TISS p less than 0.0001). In this group of patients evaluated at a single institution both the APACHE II and TISS scoring systems allowed subgroup separation between survivors and nonsurvivors of ICU admission. However, these scoring methods demonstrated limitations in terms of outcome prediction when applied to the individual patient.