Propensity Score Analysis Comparing Videothoracoscopic Lobectomy With Thoracotomy: A French Nationwide Study

Background. Video-assisted thoracoscopic surgery (VATS) lobectomy has recently become the recommended approach for stage I non-small cell lung cancer. However, these guidelines are not based on any large randomized control trial. Our study used propensity scores and a sensitivity analysis to compare VATS lobectomy with open thoracotomy. Methods. From 2005 to 2012, 24,811 patients (95.1%) were operated on by open thoracotomy and 1,278 (4.9%) by VATS. The end points were 30-day postoperative death, postoperative complications, hospital stay, overall survival, and disease-free survival. Two propensity scores analyses were performed: matching and inverse probability of treatment weighting, and one sensitivity analysis to unmask potential hidden bias. A subgroup analysis was performed to compare ``high-risk'' with ``low-risk'' patients. Results are reported by odds ratios or hazard ratios and their 95% confidence intervals. Results. Postoperative death was not significantly reduced by VATS whatever the analysis. Concerning postoperative complications, VATS significantly decreased the occurrence of atelectasis and pneumopathy with both analysis methods, but there were no differences in the occurrence of other postoperative complications. VATS did not provide a benefit for high-risk patients. The VATS approach decreased the hospital length of stay from 2.4 days (95% confidence interval, 1.7 to 3 days) to 4.68 days (95% confidence interval, 8.5 to 0.9 days). Overall survival and disease free survival were not influenced by the surgical approach. The sensitivity analysis showed potential biases. Conclusions. The results must be interpreted carefully because of the differences observed according to the propensity scores method used. A multicenter randomized controlled trial is necessary to limit the biases. (C) 2016 by The Society of Thoracic Surgeons