We evaluated the relationship between hospital non-small cell lung cancer (NSCLC) surgery volume and patient outcomes in Victoria.
Victorians with a primary diagnosis of NSCLC between 2008 and 2014 were identified in the Victorian Cancer Registry (n=15,827), 3,177 (21%) of whom had surgery. Primary outcome was death within 90 days of surgery and secondary outcomes were overall survival, use of postoperative ventilation, ≥24hours spent in ICU and length of stay >17days. Hospital volume quartiles were defined as Q1: 1-17, Q2: 18-30, Q3: 31-53 and Q4: 54+ surgeries per year.
56% (1,767/3,177) underwent lobectomy, 39% (1,241/3,177) sub-lobar resection and 5% (169/3,177) pneumonectomy. There was a weak linear relationship between lower hospital volume and greater 90-day mortality in patients undergoing sub-lobar resections, but not lobectomy, after adjustment (p-trend 0.1). The relationship was statistically significant when metastatic disease was removed as a covariate from the model, suggesting patient selection played a significant role (p-trend 0.03). Hospitals in the lower quartiles resected more patients with metastatic disease (Q1. 12% vs. Q4. 6% p-trend <0.001). Patients operated in lower volume centres had more admissions in ICU ≥24hours (Q1. 12% vs. Q4. 48%, p-trend <0.001). Median overall survival was 5.7 years, 4.8 years and 5.8 years for lobectomy, sub-lobar resection and pneumonectomy, respectively. Survival was higher compared to NSW and overseas data despite similar resection rates.
Case selection appears to play a significant role in poorer outcomes after sub-lobar surgical resection in lower volume centres. There are a greater proportion of patients with metastatic disease undergoing surgery in lower volume centres thus affecting patient outcomes.