Association of short-term outcomes and long-term survival after lung cancer resection.

Abstract

Quality metrics that compare care across institutions typically are based on short-term outcomes. This study evaluated whether short-term quality metrics for non-small cell lung cancer (NSCLC) resection predict long-term survival. Centers in the National Cancer Database that performed ≥30 NSCLC resections between 2010 and 2019 were ranked based on major postoperative morbidity, defined as a weighted composite of 30-day mortality, unplanned readmissions, and hospital length of stay longer than 14 days. Centers were stratified by morbidity rates into quintiles, with the top quintile designated high quality. The impact of care at high-quality institutions on survival was assessed with Kaplan-Meier analysis and Cox proportional hazards modeling. The study included 198,115 patients from 928 centers. Compared to non-high-quality centers, high-quality centers had lower rates of 30-day mortality (0.8% [n = 362/47,321] vs 2.4% [n = 3614/150,794]: P < .001) and morbidity (median, 5% [interquartile range (IQR), 4.0%-5.9%] vs 10.8% [IQR, 8.7%-14.0%]; P < .001). Patients treated at high-quality centers had improved long-term survival compared to other patients in both univariable analysis (5-year survival, 71.5% [95% confidence interval (CI), 71.0%-71.9%] vs 62.6% [95% CI, 62.3%-62.8%]; P < .0001) and multivariable analysis (hazard ratio [HR], 0.75; 95% CI, 0.73-0.77; P < .001). Sensitivity analysis of stage IA patients treated with lobectomy and no induction therapy showed similar survival benefits as care at high-quality centers in both univariable analysis (5-year survival, 79% [95% CI, 78.3%-79.7%] vs 73.2% [95% CI, 72.8%-73.6%]; P < .001) and multivariable analysis (HR, 0.76; 95% CI, 0.73-0.78; P < .001). Patients who underwent lung cancer resection at institutions deemed high-quality based on short-term outcomes also had better long-term survival.