Abstract
BACKGROUND: Guidelines recommend mediastinal sampling first for patients with mediastinal lymphadenopathy with suspected lung cancer. The objective of this study was to describe practice patterns and outcomes of diagnostic strategies in patients with lung cancer. METHODS: This study included a retrospective cohort of 15,914 patients with lung cancer with T1-3N1-3M0 disease diagnosed from 2004 to 2013 in the National Cancer Institute's Surveillance, Epidemiology, and End Results or Texas Cancer Registry Medicare-linked databases. Patients who had mediastinal sampling as their first invasive test were classified as guideline consistent; all others were guideline inconsistent. Propensity matching was used to compare the number of tests performed, and multivariable logistic regression was used to compare the incidence of complications. RESULTS: Guideline-consistent care increased from 23% to 34% of patients from 2004 to 2013 (P < .001). Use of endobronchial ultrasound-guided transbronchial needle aspiration increased from 0.1% to 25% of all patients (P < .001), and mediastinal sampling increased from 54% to 64% (P < .0001). Guideline-consistent care was associated with fewer thoracotomies (38% vs 71%; P < .001) and CT scan-guided biopsies (10% vs 75%; P < .001) than guideline-inconsistent care but more transbronchial needle aspirations (59% vs 12%; P < .001). Guideline-consistent care was associated with fewer pneumothoraxes (5.1% vs 22%; P < .001), chest tubes (0.9% vs 4.4%; P < .001), hemorrhages (3.5% vs 5.8%; P < .001), and respiratory failure events (2.7% vs 3.7%; P = .047) than guideline-inconsistent care. Bronchoscopic mediastinal sampling was associated with fewer complications than surgical mediastinal sampling. CONCLUSIONS: Guideline-consistent care with mediastinal sampling first was associated with fewer tests and complications. Quality gaps decreased with the introduction of endobronchial ultrasound-guided transbronchial needle aspiration but persist. Gaps include failure to sample the mediastinum first, failure to sample the mediastinum at all, and overuse of thoracotomy.