Predictors of an appropriate admission to an ICU after a major pulmonary resection.

Background: There are no recommendations about admission to an ICU after a major lung resection and there are considerable differences among institutions in this respect.

Objectives: To audit the practice of admission to an ICU after a major lung resection and evaluate factors predicting the need for intensive care.

Methods: Clinicalrecords of all patients who underwent major pulmonary resections in a 14-month period were reviewed retrospectively. The criteria for postoperative admission to the ICU were: (1) standard pneumonectomy if comorbidity index (CI) >0 and/or ASA score >1, and/or abnormal spirometry or arterial gas analysis; (2) extended pneumonectomy; (3) lobectomy if CI >or=4 and/or ASA >or=3; (4) lobectomy if FEV(1) <60% of predicted; (5) lobectomy if FEV(1) is between 60 and 80% and hypercapnia.

Results: Among the 49 patients postoperatively admitted to the surgical ward, only 1 needed late intensive care. Among the 55 patients admitted to the ICU, 25 did not require specific intensive care and were discharged 24 h postoperatively, whereas the remaining 30 patients required specific intensive care. Multivariate analysis identified ASA score, predictive postoperative DL(CO), and predictive postoperative product (PPP) as independent predictors of a need for admission to an ICU.

Conclusion: This empirical protocol was useful in identifying patients not likely to need admission to the ICU. ASA score, predictive postoperative DL(CO), and PPP are independent predictors of a need for admission to an ICU.