Predictors of outcome in cardiac surgical patients with prolonged intensive care stay

Abstract

Objective: To determine the predictors of outcome in cardiac surgical patients with prolonged ICU stay.

Design: Inception cohort with retrospective chart review.

Setting: Adult cardiovascular ICU.

Patients: All patients admitted after cardiac surgery who stayed in ICU for at least 14 consecutive days.

Interventions: Collection of data, including preoperative demographics, comorbidity, routine laboratory testing, surgical procedure, duration of cardiopulmonary bypass and aortic cross-clamping, postoperative requirement for transfusion and intra-aortic balloon counterpulsation, and postoperative indexes of organ dysfunction 14 and 28 days after surgery. An organ failure score (OFS) was calculated for days 1, 14, and 28.

Outcome measures: Hospital mortality.

Results: One hundred forty-one of 324 (43.5%) ICU admissions lasting at least 14 days resulted in hospital mortality. Seventy-four of 166 (45%) ICU admissions lasting at least 28 days resulted in hospital mortality. Preoperative demographics, morbidity, and indexes of organ failure in the first 24 h after surgery were not predictive of hospital mortality. Indexes of organ failure predictive of hospital death at 14 days included requirement for epinephrine infusion, diminished Glasgow coma scale, requirement for dialysis, greater value of BUN, lower value of creatinine, greater value of bilirubin, greater value of arterial PCO2, lower platelet count, and lower value of serum albumin. After a 28-day stay in ICU, the indexes of organ failure predictive of hospital mortality included requirement for dopamine or norepinephrine infusions, diminished Glasgow coma score, greater value of bilirubin, greater value of arterial PCO2, lower value of serum albumin, and advanced age. The area under the receiver operating characteristic curve for the OFS on day 1 was 0.55+/-0.04 (p=0.12), on day 14 it was 0.75+/-0.03 (p<0.0001), and on day 28 it was 0.76+/-0.04 (p<0.0001).

Conclusion: Preoperative health status and early organ failure were not predictive of late hospital mortality. The pattern of late organ failure associated with hospital mortality changed with time.