PREDICTING WHO DIES DEPENDS ON HOW SEVERITY IS MEASURED - IMPLICATIONS FOR EVALUATING PATIENT OUTCOMES

Objective: To determine whether assessments of illness severity, defin ed as risk for in-hospital death, varied across four severity measures . Design: Retrospective cohort study. Setting: 100 hospitals using the MedisGroups severity measure. Patients: 11 880 adults managed medical ly for acute myocardial infarction; 1574 in-hospital deaths (13.2%). M easurements: For each patient, probability of death was predicted four times, each time by using patient age and sex and one of four common severity measures: 1) admission MedisGroups scores for probability of death scores; 2) scores based on values for 17 physiologic variables a t time of admission; 3) Disease Staging's probability-of-mortality mod el; and 4) All Patient Refined Diagnosis Related Groups (APR-DRGs). Pa tients were ranked according to probability of death as predicted by e ach severity measure, and rankings were compared across measures. The presence or absence of each of six clinical findings considered to ind icate poor prognosis in patients with myocardial infarction (congestiv e heart failure, pulmonary edema, coma, low systolic blood pressure, l ow left ventricular ejection fraction, and high blood urea nitrogen le vel) was determined for patients ranked differently by different sever ity measures. Results: MedisGroups and the physiology score gave 94.7% of patients similar rankings. Disease Staging, MedisGroups, and the p hysiology score gave only 78% of patients similar rankings. MedisGroup s and APR-DRGs gave 80% of patients similar rankings. Patients whose i llnesses were more severe according to MedisGroups and the physiology score were more likely to have the six clinical findings than were pat ients whose illnesses were more severe according to Disease Staging an d APR-DRGs. Conclusions: Some pairs of severity measures assigned very different severity levels to more than 20% of patients. Evaluations o f patient outcomes need to be sensitive to the severity measures used for risk adjustment.