Variability in Invasive Mediastinal Staging for Lung Cancer: A Multi-Center Regional Study
Lucas W Thornblade1, *Douglas E. Wood1, *Michael S. Mulligan1, Alexander S. Farivar2, *Michal Hubka3, Kimberly E. Costas4, *Bahirathan Krishnadasan5, Farhood Farjah1
1University of Washington, Seattle, WA;2Swedish Cancer Institute, Seattle, WA;3Virginia Mason Medical Center, Seattle, WA;4Providence Regional Medical Center, Everett, WA;5CHI Franciscan Health System, Tacoma, WA
OBJECTIVES: We sought to determine whether rates of invasive mediastinal staging for lung cancer varied across hospitals participating in a regional quality improvement and research collaborative.
METHODS: We performed a retrospective study (July 2011-December 2013) of resected lung cancer patients treated in the Puget Sound region of Washington State using data from the Surgical Clinical Outcomes and Assessment Program (a physician-led quality initiative that monitors performance by benchmarking processes of care). Invasive mediastinal staging was defined by mediastinoscopy (MED), endobronchial ultrasound-guided nodal-aspiration (EBUS), and/or esophageal ultrasound-guided nodal-aspiration (EUS) performed prior to and/or at the time of lung resection. In order to avoid inflated estimates of hospital-level variation, we used a mixed-effects model to account for the influence of chance (which can arise from relatively small sample sizes/surgical volumes at any one hospital). We also adjusted for hospital-level differences in the frequency of clinical stage IA disease. A likelihood ratio test was used to test whether rates of invasive staging varied across hospitals.
RESULTS: A total of 406 patients (mean age 68 years, 57% female, 69% clinical stage IA, 83% lobectomy, and 40% thoracotomy) underwent resection at 5 hospitals (4 community, 1 academic). Overall, invasive staging occurred in 66% of patients (95% confidence interval 61-71%). Mediastinoscopy was the most commonly performed invasive staging modality (MED only 85%; EBUS & MED 11%; EBUS only 3%; EBUS, EUS & MED <1%). Rates of invasive staging varied significantly (p<0.001) across hospitals after adjusting for random variation (i.e. chance) and clinical stage (Figure). Adjustment for clinical stage accounted for only 3% of the observed variability in rates of invasive staging across hospitals. Confidence interval inspection revealed that two hospitals performed invasive staging significantly more often than the population average, whereas two hospitals performed invasive staging significantly less often than the overall average (Figure).
CONCLUSIONS: We demonstrate—for the first time—that rates of invasive mediastinal staging vary significantly across hospitals. Variation not explained by chance or case-mix is often believed to represent poor quality care. However, because low levels of evidence support current practice guidelines, it is equally plausible that variation in care could represent clinical uncertainty over the appropriate indications for invasive mediastinal staging. A randomized trial comparing patient outcomes between varying staging strategies may clarify the indications for invasive mediastinal staging.
Back to 2017 Program