Feasibility Study of Linkage of Cancer Registry to Multiple Healthcare Payers This agreement will fund a study to explore the feasibility of linking Atlanta SEER and GA state registry data with GA state hospital discharge data, and with health claims data from GA Medicaid, GA Kaiser (an affiliate of the NCI CRN) and with claims maintained by MedStat for the GA Stage Health Benefits Plan, which includes such major payer organizations as Blue Cross and United Health Care. Linking SEER and other cancer registry data to non-Medicare health insurance based claims data on the use and cost of health care procedures and services would greatly enhance the capability to conduct cancer related health services research. The linked Surveillance, Epidemiology, and End Results (SEER) - Medicare database has been an invaluable resource for the conduct of cancer-related health services research. However, this resource has several well known limitations. In particular, the non-disabled population under age 65 is not included in the data and data for individuals over the age of 65 but enrolled in Medicare HMO programs (currently over 20% of Medicare enrollees) are not informative. Because Medicaid and private health insurance programs do not have the same unified national structure and administration as Medicare, there are multiple challenges to replicating the SEER-Medicare link with these other payers. A promising start to addressing this question would be to fund a feasibility study at the state level, since Medicaid and private health insurance plans are typically administered at this level. This agreement would fund such a feasibility study for the state of Georgia. Georgia is an opportune site for this study because the Atlanta SEER program and the Georgia state registry already link to Medicare and because of the existence of a relatively well developed Comprehensive Cancer Control Plan (CCCP) in Georgia. Officials and researchers working with the Georgia CCCP have already formulated a preliminary proposal to explore the feasibility of linking Atlanta SEER and Georgia state registry data with Georgia state hospital discharge data, and with health claims data from Georgia Medicaid, Georgia Kaiser (an affiliate of the NCI Cancer Research Network) and with claims maintained by MedStat for the Georgia Stage Health Benefits Plan (similar to the federal health benefits plan), which includes such major payer organizations as Blue Cross and United Health Care. Linking SEER and other cancer registry data to non-Medicare health insurance based claims data on the use and cost of health care procedures and services would greatly enhance the capability to conduct cancer related health services research. Accomplishing such linkages is most likely to be done at the state level. However the feasibility of implementing such linkages and the quality and utility of the resulting data is unknown at this time. A feasibility study conducted at the state level would make it possible for NCI to assess whether additional major investments are justified in this area. This following tasks will be completed during the period of performance for this project: (1) Data for all incident cases of breast and colorectal cancer in Georgia over the 1999-2005 period from the Georgia Comprehensive Cancer Registry (GA Registry) will be linked (in a bilateral manner) with administrative, claims, and/or clinical data drawn from each of the following data sources: ? Medicare files; ? Medicaid files; ? State Health Benefit Plan (State Plan) files covering all employees of the state of Georgia (and who are or were enrolled in a variety of private indemnity and managed care plans offered by several major private entities, including Blue Cross and United Health Care); ? Kaiser Permanente Georgia (Kaiser GA) clinical and administrative data for all enrollees across their various managed care offerings; ? Medical record and chart review data (for quality measures requiring clinical detail not available in administrative files);and ? Georgia state hospital discharge data (to augment GA Registry data with additional variables, such as insurance enrollment status, prior to their collection by the registry), as well as other secondary data sets (such as the Area Resource File) that can support interpretative analyses. (2) Each bilateral linked data set will be subject to a rigorous data quality examination, focusing on the accuracy and completeness of the linkage at the patient level, and the degree of concordance between the component data sets in terms of breast and colorectal cancer incidence and other variables relevant to quality-of-care assessment (e.g., dates for initiation of adjuvant therapy). There will be a focus also on completeness of claims data in recording treatment and the adequacy of post-diagnosis follow-up time in claims data. (3) The extent to which each bilateral linked dataset can be used successfully to compute adherence rates for a selected group of cancer quality measures will be investigated. The measures, focused on cancer diagnosis and treatment, will be drawn from those recommended recently by the National Quality Forum, the Institute of Medicine, and recent cancer quality-of-care literature. (4) An initial version of a consolidated Georgia cancer data set that draws jointly from the administrative and clinical variables in the bilateral linked data sets will be produced in a way that supports tracking of the breast and colorectal cancer quality measures adopted for this project. This version of a consolidated data set will lay the groundwork for a statewide Georgia Cancer Quality and Outcomes Research Data System and the test the capability of this initial consolidated data set for quality-of-care assessment in breast and colorectal cancers will be tested. (5) The technical requirements and costs of developing and implementing the software needed for maintenance and sustained application of the consolidated linked Georgia cancer data set will be assessed. (6) Policies will be developed to ensure patient confidentiality and compliance with all federal and state rules regarding protection of human subjects, while also assessing how to minimize the impact of these important requirements on the development and maintenance of a state cancer data system.
