Background: Surgical resection with lobectomy is recommended for T1/T2 N0 small cell lung cancer (SCLC) patients after negative mediastinal staging. We sought to characterize variation in surgical therapy for early SCLC and determine the effect of a hospital's practice patterns on patient survival.
Methods: The National Cancer Database was examined from 2004 to 2013. Risk- and reliability-adjusted hierarchical logistic regression was used to estimate the adjusted odds of resection by hospital. Hospitals were then grouped into quartiles by observed-to-expected rates of surgical treatment. Patient, tumor, and hospital characteristics were compared across quartiles. Kaplan-Meier plots and Cox proportional hazard models were built to compare patient survival as a function of a hospital's tendency to use surgical intervention.
Results: We identified 5,079 patients with T1/T2 N0 SCLC in 317 hospitals, and 1,260 underwent resection. Analysis after adjusting for demographic, comorbidity, and tumor factors showed patients treated at hospitals in the highest quartile of surgical use were 17 times more likely to undergo surgical resection than those in the lowest quartile (44.8% vs 7.6%; odds ratio, 16.7l; 95% confidence interval, 12.59 to 22.18). Hospitals in the highest quartile were more likely to be academic centers (48% vs 21%), more likely to perform lobectomy (28.3% vs 5.0%), and treated more mixed-histology tumors (11.1% vs 4.5%). Survival was significantly longer for patients treated at hospitals most likely to use surgical therapy (median, 25.3 vs. 18.8 months; p < 0.0001). Hazard ratio differences in mortality persisted in multivariate Cox models (hazard ratio, 0.80; 95% confidence interval, 0.72 to 0.89; p < 0.0001).
Conclusions: Large variation exists in the use of surgical therapy for early SCLC in the United States, which may represent a significant quality improvement opportunity for patients with early SCLC.
Copyright © 2017 The Society of Thoracic Surgeons. Published by Elsevier Inc. All rights reserved.