216. Institutional Factors Associated with Non-Adherence to the Thoracic Surgery Outcomes Research Network (ThORN) Quality Measures in Treatment of Stage I/II Non-Small Cell Lung Cancer
Rhami Khorfan1, Julia Coughlin2, *David T. Cooke3, Robert Meguid4, Leah Bakhus5, *Thomas K. Varghese6, Joseph Phillips7, Karl Bilimoria1, David Odell1
1Northwestern University Feinberg School of Medicine, Chicago, IL; 2Rush University Medical Center, Chicago, IL; 3University of California, Davis Medical Center, Sacramento, CA; 4University of Colorado, Aurora, CO; 5Stanford University, Stanford, CA;6University of Utah, Salt Lake City, UT;7Dartmouth Hitchcock Medical Center, Lebanon, NH
Invited Discussant: *Haiquan Chen
Objective: Previous studies have identified variation in the delivery of quality lung cancer care. However, existing quality metrics fall short of fully capturing the complexity of cancer care. To address this, the Thoracic Surgery Outcomes Research Network (ThORN) developed non-small cell lung cancer (NSCLC) quality measures through an expert consensus process. Our goal was to evaluate baseline national patient- and hospital-level adherence to these measures, and identify factors associated with non-adherence.
Methods: Patients with stage I or II NSCLC between 2010 and 2015 were identified in the National Cancer Database. Patient-level and hospital-level adherence to 7 newly defined lung cancer quality measures was calculated. Hospital-level pass/fail threshold was set at 85% adherence. Hospital factors influencing non-adherence were identified using logistic regression, adjusting for patient characteristic such as age, sex, race, insurance status, and co-morbidities.
Results: There were 270,658 patients from 1,325 hospitals that met inclusion criteria during the study period. Patient-level and hospital-level adherence rates are reported in Table 1. Lymph node sampling during resection was not performed adequately in 27.8% (367) of hospitals. An anatomic resection was not consistently performed for T1b or greater tumors in 32.4% (429) of hospitals. 80.3% (1,064) of hospitals failed to appropriately recommend adjuvant chemotherapy for pathologic stage II NSCLC. 39.0% (517) of hospitals administering primary radiation treatment did not have pathologic confirmation of the diagnosis prior to definitive local therapy. Higher annual lung cancer treatment volume, academic institution, and Pacific/Mountain region were all associated with an increased risk of non-adherence to adjuvant chemotherapy and radiation therapy measures. Conversely, lower volume and treatment at non-academic institution were associated with increased likelihood of inadequate lymph node sampling and non-anatomic resection.
Conclusions: Significant gaps in the delivery of quality lung cancer care to patients with early-stage NSCLC persist. The highest failure rates were in the appropriate administration of adjuvant chemotherapy, and in confirmation of pathologic diagnosis prior to initiation of radiation therapy. Adherence rates for appropriate lymph node sampling and anatomic resections for T1b tumors were also below threshold. Interestingly, higher volume and academic hospitals had higher adherence rates for the measures mainly related to surgical care (lymph node sampling and anatomic resection) but lower adherence rates for the measures related to coordination of care (referrals for adjuvant chemotherapy and biopsy prior to radiation therapy). This finding requires further investigation, but suggests targets for quality improvement efforts may vary by type of institution.