Predictors of Outcome in Modern Surgery for Lung Abscess

Background Surgery for lung abscess is a challenging task. Timing and indications for surgery are not well established. Identification of predictors of outcome could help to clarify the role of surgery. Methods Patients who underwent major thoracic surgery for infectious lung abscess were identified at six centers for general thoracic surgery in Germany, Spain, the United Kingdom, and the United States. Study period was 2000 to 2016. Results There were 91 patients. Pulmonary sepsis (48), pleural empyema (43), persistent air leakage (25), acute renal failure (12), and respiratory failure with mechanical ventilation (25) were already preoperatively present. The mean Charlson index of comorbidity was 3.0 (median: 2.0; interquartile range: 3). Procedures were segmentectomy (18), lobectomy (58), and pneumonectomy (15). The 30-day mortality following surgery was 13/91.Preoperative sepsis (odds ratio [OR]: 13.69; 95% confidence interval [CI]: 1.86-610.53; p < 0.01), preoperative persistent air leak (OR: 13.46, 95% CI: 3.00-85.37, p < 0.01), respiratory failure (OR: 5.60; 95% CI: 1.41-24.84; p < 0.01), acute renal failure (OR: 6.15 ; 95% CI: 1.24-29.56 ; p = 0.01), and Charlson index of comorbidity ≥ 3 (OR: 7.19 ; 95% CI: 1.43-71.21 ; p < 0.01) are associated with higher mortality, whereas age > 70 years (p = 0.46) and the extent of pulmonary resection (segmentectomy, lobectomy, pneumonectomy) have no significant influence on mortality. Patients with fatal outcome have significantly higher Charlson index of comorbidity (p < 0.01). Conclusions Delayed referral for surgery is common. Significant predictors for fatal outcome are pulmonary sepsis, septic complications (air leak, pleural empyema), septic organ failure (respiratory, acute renal failure), and preexisting comorbidity (Charlson index of comorbidity ≥ 3). The extent of surgical resection shows no significant influence.