The effect of surgeon volume on procedure selection in non-small cell lung cancer surgeries.

OBJECTIVES: Procedure selection by the surgeon can greatly affect patients' operative and long-term survival. This selection potentially reflects comfort with technically challenging surgeries. This study aims to examine surgeon choices for non-small cell lung cancer and whether surgeon volume predicts the type of procedure chosen, controlling for patient demographics, comorbidity, year of surgery, and institutional factors. METHODS: Data were abstracted from an Ontario population-based linked database from 2004 to 2011. Patient demographics, comorbidities, year of surgery, and institutional and surgical factors were evaluated. Three-level, random-effect, multilevel regression analyses were performed to examine factors influencing operative selection. RESULTS: Over the interval, 8070 patients (50.4% were male) underwent surgical resection, including pneumonectomy (n = 842), lobectomy (n = 6212), and wedge resection (n = 1002). Resections were performed by 124 unique physicians in 45 institutions. The proportion of patients undergoing pneumonectomy decreased from 14.8% in 2004 to 7.6% in 2011. Multilevel regression analysis showed physician volume, age, year of procedure, gender, and comorbidities were predictive of performing a pneumonectomy. By adjusting for these variables, the results indicated that for each 10-unit increase in physician volume, the relative risk of performing a pneumonectomy decreased by 9.1% (95% confidence interval, 8.2-10.0, P = .04). CONCLUSIONS: Although patient and temporal factors influence the type of resection a patient receives for non-small cell lung cancer, surgeon volume also is a strong predictor. This study may be limited by minimal stage data, but the suggestion that a surgeon's total procedural volume for non-small cell lung cancer significantly influences procedure selection has implications on how we deliver care to this patient population.

Pneumonectomy: the burden of death after discharge and predictors of surgical mortality.

BACKGROUND: Pneumonectomy has the highest mortality rate among resections for lung cancer, with limited literature differentiating predictors of postpneumonectomy in-hospital mortality (IHM) from early postdischarge mortality (PDM). This study aims to examine the burden of death over time and to identify potential predictive factors, including patient comorbidities and hospital and surgeon volumes. METHODS: Data were abstracted from an Ontario population-based linked database from 2005 to 2011. Proportional mortality and cumulative survival attributable to IHM and 90-day PDM is reported. Logistic and Cox regression analyses examined the role of potential factors related to death. Odds ratios (ORs) and hazard ratios (HRs) and 95% confidence intervals (CIs) were reported. RESULTS: Of 505 patients who underwent pneumonectomy, the median length of stay was 6 days (1-30 days). IHM was 4.4% (2.9%-6.5%), and 90-day PDM was an additional 6.4% (4.6%-9.0%). Logistic regression showed that congestive heart failure (CHF) (OR, 23.5; range, 4.0-136.0), cerebrovascular disease (OR, 12.5; range, 1.2-128.0), renal disease (OR, 8.8; range, 1.3-60.5), and previous myocardial infarction (MI) (OR, 5.4; range, 1.5-20.0) were predictive of IHM, whereas age (HR, 1.4; range, 1.1-1.7) per year and CHF (HR, 18.0; range, 4.0-79.0) were predictive of PDM. All other factors were not significant. CONCLUSIONS: PDM represents a distinct and underrecognized burden of postoperative death. More than half of postpneumonectomy mortality occurred after discharge, and the rate remained unchanged over the study period. Patient factors play a major role in both IHM and PDM, whereas institutional and physician volume do not influence outcome, suggesting the importance of patient selection and the need for continued evaluation of mortality.