The role of the surgical volume for clinical outcomes in VATS lobectomy for lung cancer: a national large database multicenter analysis

Improving the quality of lung cancer care at a cost that can be sustained is a hotly debated issue. High-risk, low-volume procedures (such as lung resections) are believed to improve significantly when centralised in high-volume centres. However, limited evidence exists to support volume requirements in lung cancer surgery. On the other hand, there was no evidence that the number of lung resections affected either the short-term perioperative results or the long-term cost. Using data from an extensive nationwide registry, this study investigated the correlations between surgical volumes and selected perioperative outcomes. A retrospective analysis of a prospectively filled national registry that follows stringent quality assurance and security procedures was conducted to ensure data accuracy and security. Patients who underwent VATS lobectomy from 2014 to 2019 at the participating centres were included. Selected perioperative outcomes were reported. Total direct hospital cost is measured at discharge for hospitalisations with a primary diagnosis of lung cancer, hospital stay costs, and postoperative length of hospital stay after lobectomy. After the propensity score matched, centres were divided into three groups according to the surgical volume of the unit where VATS lobectomies were performed (high-volume centre: > 500 lobectomies; medium-volume centre: 200-500 lobectomies; low-volume centre: < 200 lobectomies). 11,347 patients were included and matched (low-volume center = 2890; medium-volume center = 3147; high-volume center = 2907). The mean operative time density plot (Fig. 1A) showed no statistically significant difference (p = 0.67). In contrast, the density plot of the harvested lymph nodes (Fig. 1B) showed significantly higher values in the high-volume centres (p = 0.045), albeit without being clinically significant. The adjusted rates of any and significant complications were higher in the low-volume centre (p = 0.034) without significantly affecting the length of hospital stay (p = 0.57). VATS lobectomies for lung cancer in higher-volume centres seem associated with a statistically significantly higher number of harvested lymph nodes and lower perioperative complications, yet without any significant impact in terms of costs and resource consumption. These findings may advise the investigation of the learning curve effect in a complete economic evaluation of VATS lobectomy in lung cancer. Fig. 1 The mean operative time density plot showed no statistically significant difference (p = 0.67).