Author of

• 11946

  +

  P2.24 - Poster Session 2 - Supportive Care (ID 157)

  • Event: WCLC 2013
  • Type: Poster Session
  • Track: Supportive Care
  • Presentations: 1
  • Moderators:
  • Coordinates: 10/29/2013, 09:30 - 16:30, Exhibit Hall, Ground Level

  • 12007

    +

    P2.24-061 - Profiling Hospital Performance: Adherence to Guidelines for Perioperative Lung Cancer Imaging (ID 3063)

    09:30 - 09:30  |  Author(s): S. Zeliadt
    • Abstract

    Background
    Ranking hospitals, publishing reports of high and low performing facilities, and altering reimbursement for outlier sites is becoming common in many areas of healthcare. We explore challenges of profiling hospitals on appropriate use of advanced imaging in lung cancer care.

    Methods
    We identified 17,325 patients with newly diagnosed lung cancer over a 4 year period from the Veterans Health Administration’s Central Cancer Registry, of whom 5,424 (31%) were Stage IIB, IIIA, or IIIB and treated in one of 84 VA Medical Centers. Imaging was assessed 180 days pre and post diagnosis per National Comprehensive Cancer Network guidelines and American College of Radiology Appropriateness Criteria for Imaging. Risk-adjusted hospital rankings were estimated using a hierarchical logistic modelling approach.

    Results
    Recommended imaging was performed in 70% of patients overall for brain imaging and 51% of patients for positron emission tomography (PET). Overutilization, with combined bone scintigraphy and PET (BS/PET), occurred in 20% of patients. Widespread variation in non-adherence to guidelines was observed, with distinct geographic patterns. The New England region of the U.S. had imaging rates that were 27 percentage points higher than the Great Plains area, which had the lowest use of imaging. Receipt of recommended imaging ranged from a low of 61% of patients in one facility, to 94% in the best performing facility. Models calculating traditional observed-to-expected ratios identified 6 (7%) underperforming facilities, while hierarchical logistic models recommended by Centers for Medicare and Medicaid for hospital profiling, identified 14 (17%) underperforming facilities.Figure 1.

    Conclusion
    A significant proportion of patients do not receive recommended imaging, while other patients appear to undergo excessive imaging for lung cancer. These observations are hallmarks of poor quality. Hospital ranking and assessment of performance is sensitive to statistical approach. Lung cancer care providers should engage in developing uniform approaches to reporting and monitoring quality of care, and work to identify opportunities to improve efficiency and reduce waste.