Effective new imaging technologies for detecting cancer and determining treatment have great potential to improve the health of the U.S. population. However, research is needed to investigate whether these technologies actually improve clinical care and health outcomes under realistic circumstances when implemented in the U.S. health care system. Ongoing research within the Center for CER in Cancer imaging is:
- Addressing the health care utilization impact of potentially significant incidental imaging findings associated with the use of spiral CT for lung cancer screening (Project 1)
- Characterizing the diffusion of Positron Emission Tomography (PET) and its effect on care patterns and assessing the degree to which the physician-reported impact of PET on planned care is reflected in actual care patterns (Project 2)
These projects will have important implications for the design of future registries to support CER and will enhance cancer CER capacity for advanced imaging. An accompanying Center goal is the development of novel methodologies and the innovative use of data resources so as to optimize the design of future imaging studies.
Project 1: Characterizing Significant Incidental Findings Identified with Advanced Imaging
This project entails assessment of the health care utilization impact of potentially significant incidental imaging findings associated with the use of spiral CT for lung cancer screening. Although the primary endpoint of the National Lung Screening Trial is lung cancer mortality, an important secondary endpoint is downstream medical care utilization, much of which may be attributable to the detection of incidental findings on low-dose helical computed tomography (LDCT). In this project, we will use Medicare claims data to estimate the rates and costs of medical utilization related to the detection of incidental findings in the Medicare-eligible subset of the ACRIN-NLST (>27% of participants). In addition, we will use the Medicare claims data to compare these rates and costs to those obtained from the NLST medical abstraction process triggered by self-reporting of medical utilization.
Project 2: CER Studies of PET
CER studies of PET are focused on the population of US Medicare beneficiaries with cancer who are ages 65 and older and on participants in the National Oncologic PET Registry (NOPR). Thus, two distinct cohorts are studied:
- Participants in the National Oncologic PET Registry (NOPR)
- Beneficiaries with claims for six different cancer types for which Medicare expanded payment in 2001
NOPR was established in 2006 under the Center for Medicare and Medicaid Services (CMS) coverage with evidence development (CED) designation for PET use in individuals with specific cancers. From 2006 to 2008, NOPR collected data on more than 100,000 individuals with cancer. On the basis of NOPR-reported data, in April 2009 CMS expanded coverage of PET to include a single imaging study to guide initial therapy clinical management for most beneficiaries with solid tumors. It is notable that the NOPR evidence was based solely on physician-reported changes in management plans rather than on observed management changes. Little is known regarding the concordance between the NOPR physician-reported care plans and actual subsequent care. Likewise, little is known about how the diffusion of PET has affected use of other imaging modalities. This project addresses these knowledge gaps. For NOPR participants, we will compare intended post-PET care plans with observed resource utilization through analysis of CMS administrative claims data. Our findings will have important implications for the design of future registries.
More generally, we will characterize PET use and diffusion patterns in the Medicare population. In July 2001, PET became a covered service for Medicare beneficiaries when used for diagnosis, staging and restaging of non-small cell lung, esophageal, colorectal and head and neck cancers, lymphoma, and melanoma. We will be assessing 1) temporal trends of PET use relative to that of computed tomography, magnetic resonance imaging, or bone scintigraphy and seek to determine whether PET was or was not temporally clustered with other imaging (i.e., whether it was additive or replacement) and 2) assess regional health care market characteristics and socio-demographic profiles of Medicare beneficiaries undergoing imaging to explain differences in the rate of adoption of PET.