ABSTRACT
Objective
To investigate discrepancies in Medicaid enrollees’ hospital discharges reported in two data sources widely used in health services research: the CMS Hospital Cost Report Information System (HCRIS) and the T-MSIS Analytic Files (TAF).
Study Setting and Design
This is a descriptive study comparing inpatient discharges reported in the two data sets. We included inpatient admissions at general hospitals in 2020–2021.
Data Sources and Analytic Sample
We used HCRIS data covering reporting periods starting in 2020 and ending sometime in 2021 (this varied by hospital) and extracted the reported total and Health Maintenance Organization (HMO) funded Medicaid discharges and patient days. We used the 2020 and 2021 TAF inpatient files and included inpatient admissions within each hospital’s HCRIS reporting period, and calculated discharges for each hospital.
Principal Findings
There were 25 states where some hospitals had higher TAF discharge counts than HCRIS, and these same hospitals had inconsistent reporting of HMO-funded Medicaid discharges and patient days in HCRIS. This included California, New York, and Texas. There were 20 states with similar values reported in both HCRIS and TAF, and 9 of these were in states with < 5% of their enrolled Medicaid population in a comprehensive managed care plan.
Conclusions
The discrepancies between HCRIS and TAF data indicate that HCRIS may not reliably capture hospital discharge volumes for Medicaid patients, particularly those funded by managed care. These inconsistencies can misinform policy decisions and evaluations of hospital performance. Policymakers and researchers should exercise caution when using HCRIS data for Medicaid discharge counts and consider supplementing it with TAF or other sources.