Objective: To establish a scoring system for predicting the occurrence of postoperative acute respiratory failure (ARF) after general thoracic surgery.
Methods: Eighty-five cases of post-operative ARF and 300 postoperative cases without ARF, treated from July 1981 to December 1998, were retrospectively analyzed by means of Logistic regression. The main risk factors were determined, and scores were allocated to the corresponding interval respectively, so as to establish the scoring system.
Results: Ten items were included in the system: lung function (existence of symptom of shortness of break, forced expiratory volume in one second (FEV1)), arterial blood gas index (arterial oxygen pressure (PaO2)), nutritional status, cigarette-smoking index, impact of operation (blood loss, influence on the lung), complicating diseases (asthma, myasthenia gravis), and age. The sensitivity and specificity of this system was 67.05 percent and 90.33 percent respectively.
Conclusion: This newly established scoring system for predicting postoperative ARF has the advantage of convenience and objectivity.