Abstract
RATIONALE: Up to 80% of patients with lung cancer have comorbid chronic obstructive pulmonary disease (COPD). Many of them are poor candidates for stage-specific lung cancer treatment due to diminished lung function and poor functional status, and many forego treatment. The negative effect of COPD may be moderated by pulmonologist-guided management. OBJECTIVES: This study examined the association between pulmonologist management and the probability of receiving the recommended stage-specific treatment modality and overall survival among patients with non-small cell lung cancer (NSCLC) with preexisting COPD. METHODS: Early- and advanced-stage NSCLC cases diagnosed between 2002 and 2005 with a prior COPD diagnosis (3-24 months before NSCLC diagnosis) were identified in Surveillance, Epidemiology, and End Results tumor registry data linked to Medicare claims. Study outcomes included receipt of recommended stage-specific treatment (surgical resection for early-stage NSCLC and chemotherapy for advanced-stage NSCLC [advNSCLC]) and overall survival. Pulmonologist management was considered present if one or more Evaluation and Management visit claims with pulmonologist specialty were observed within 6 months after NSCLC diagnosis. Stage-specific multivariate logistic regression tested association between pulmonologist management and treatment received. Cox proportional hazard models examined the independent association between pulmonologist care and mortality. Two-stage residual inclusion instrumental variable (2SRI-IV) analyses tested and adjusted for potential confounding based on unobserved factors or measurement error. MEASUREMENTS AND MAIN RESULTS: The cohorts included 5,488 patients with early-stage NSCLC and 6,426 patients with advNSCLC disease with preexisting COPD. Pulmonologist management was recorded for 54.9% of patients with early stage NSCLC and 35.7% of patients with advNSCLC. Of those patients with pulmonologist involvement, 58.5% of patients with early NSCLC received surgical resection, and 43.6% of patients with advNSCLC received chemotherapy. Pulmonologist management post NSCLC diagnosis was associated with increased surgical resection rates (odds ratio, 1.26; 95% confidence interval, 1.11-1.45) for early NSCLC and increased chemotherapy rates (odds ratio, 1.88; 95% confidence interval, 1.67-2.10) for advNSCLC. Pulmonologist management was also associated with reduced mortality risk for patients with early-stage NSCLC but not AdvNSCLC. CONCLUSIONS: Pulmonologist management had a positive association with rates of stage-specific treatment in both groups and overall survival in early-stage NSCLC. These results provide preliminary support for the recently published guidelines emphasizing the role of pulmonologists in lung cancer management.