Risk factor analysis for fast track protocol failure.

Background: The introduction of fast-track treatment procedures following cardiac surgery has significantly shortened hospitalisation times in intensive care units (ICU). Readmission to intensive care units is generally considered a negative quality criterion. The aim of this retrospective study is to statistically analyse risk factors and predictors for re-admission to the ICU after a fast-track patient management program.

Methods: 229 operated patients (67 ± 11 years, 75% male, BMI 27 ± 3, 6/2010-5/2011) with use of extracorporeal circulation (70 ± 31 min aortic crossclamping, CABG 62%) were selected for a preoperative fast-track procedure (transfer on the day of surgery to an intermediate care (IMC) unit, stable circulatory conditions, extubated). A uni- and multivariate analysis were performed to identify independent predictors for re-admission to the ICU.

Results: Over the 11-month study period, 36% of all preoperatively declared fast-track patients could not be transferred to an IMC unit on the day of surgery (n = 77) or had to be readmitted to the ICU after the first postoperative day (n = 4). Readmission or ICU stay signifies a dramatic worsening of the patient outcome (mortality 0/10%, mean hospital stay 10.3 ± 2.5/16.5 ± 16.3, mean transfusion rate 1.4 ± 1,7/5.3 ± 9.1). Predicators for failure of the fast-track procedure are a preoperative ASA class > 3, NYHA class > III and an operation time >267 min ± 74. The significant risk factors for a major postoperative event (= low cardiac output and/or mortality and/or renal failure and/or re-thoracotomy and/or septic shock and/or wound healing disturbances and/or stroke) are a poor EF (OR 2.7 CI 95% 0.98-7.6) and the described ICU readmission (OR 0.14 CI95% 0.05-0.36).

Conclusion: Re-admission to the ICU or failure to transfer patients to the IMC is associated with a high loss of patient outcome. The ASA > 3, NYHA class > 3 and operation time >267 minutes are independent predictors of fast track protocol failure.