Challenges in Merging Medicaid and Medicare Databases to Obtain Healthcare Costs for Dual-Eligible Beneficiaries

Abstract

Dual-eligible Medicaid-Medicare beneficiaries represent a group of people who are in the lowest income bracket in the US, have numerous co-morbidities and place a heavy financial burden on the US healthcare system. As cost-effectiveness analyses are used to inform national policy decisions and to determine the value of implemented chronic disease control programmes, it is imperative that complete and valid determination of healthcare utilization and costs can be obtained from existing state and federal databases. Differences and inconsistencies between the Medicaid and Medicare databases have presented significant challenges when extracting accurate data for dual-eligible beneficiaries. To describe the challenges inherent in merging Medicaid and Medicare claims databases and to present a protocol that would allow successful linkage between these two disparate databases. Healthcare claims and costs were extracted from both Medicaid and Medicare databases for King County, Seattle, WA, USA. Three Medicaid files were linked to eight Medicare files for unique dual-eligible beneficiaries with type 2 diabetes mellitus. Although major differences were identified in how variables and claims were defined in each database, our method enabled us to link these two different databases to compile a complete and accurate assessment of healthcare use and costs for dual-eligible beneficiaries with a costly chronic condition. For example, of the 1759 dual-eligible beneficiaries with diabetes, the average cost of healthcare was USD 15,981 per capita, with an average of 76 claims per person per year. The resulting merged database provides a virtually complete documentation of both utilization and costs of medical care for a population who receives coverage from two different programmes. By identifying differences and implementing our linkage protocol, the merged database serves as a foundation for a broad array of analyses on healthcare use and costs for effectiveness research.

Validation of Medicaid Claims-based Diagnosis of Myocardial Infarction Using an HIV Clinical Cohort

“…Individuals qualify based on age, disability, income and financial resources. 22 The Medicaid data contains health care service reimbursement information including doctor visits, hospital care, outpatient visits, treatments, emergency use, prescription medications, and diagnoses, procedures and provider information. These data also include reimbursement information for Medicaid patients also enrolled in Medicare (dually eligible beneficiaries).…”

A natural language processing (NLP) program effectively extracts key pathologic findings from radical prostatectomy reports

“…However, these data include hundreds of variables with different coding schemes, if coded at all, which may create challenges in performing statistical analyses [8]. Furthermore, issues arise when attempting to merge data from multiple databases that use different definitions and labels for variables, such as Medicaid and Medicare data [9]. Thus, without proper search and extraction tools available, much of this rich information remains untouched.…”

Integrating databases for research on health and performance in small animals and horses in the Nordic countries

“…It should also be borne in mind that outside the Nordic countries personal-identity information is generally less available and its use more sensitive. A recent example of challenges encountered when merging human databases from the US has been provided [28]. …”