SAPS II predicts 30 day mortality in gram negative bacteremia

Introduction: Several studies have attempted to apply clinical and laboratory prognostic factors to predict the outcome of gram negative bacteremia (GNB). There have been no published studies using the Simplified Acute Physiology Score II (SAPS H) specifically in patients with GNB. Design: A retrospective chart review was performed on all patients with GNB in a public county hospital from July 1, 1994 to December 31, 1995. Results: 363 cases of GNB were identified, and 302 had charts available for review. The average age was 49, and 52% were male. Thirty day mortality was 27%. Forty percent were nosocomial. E. coli was the most common isolate (37%), and the genitourinary system was the most frequently identified source (32%). Seventy-seven percent were receiving appropriate antibiotics (abx) 72 hours after the blood culture (BC) was drawn, and the mean length of time from drawing the BC to receipt of the first dose of appropriate abx was 21 hours. SAPS II was calculated for the day of the first positive BC for GNB. The mean SAPS II score in survivors was 27; it was 40 in those who died more than 30 days after the BC was drawn and 48 in those who died within 30 days of the BC. All patients with a SAPS II <20 survived (59/59), and SAPS II ≥ 70 was associated with 100% 30 day mortality (11/11). For each 10 point increment in SAPS II between 20 and 70, there was a stepwise increase in mortality(10%-75%). Using multivariate analysis, higher SAPS II, nosocomial acquisition, and never receiving appropriate abx were independent predictors of 30 day mortality. Conclusions: GNB is still associated with high mortality, and higher SAPS II, nosocomial acquisition, and never receiving appropriate abx therapy are each independently associated with an increased risk of mortality.