CMS HCRIS Hospital Cost Report - onetomapanalytics/Meta_Data GitHub Wiki

CMS HCRIS Hospital Cost Report

General description

  1. Database primary purpose - Provide descriptive, financial, and statistical data to determine if Medicare over or underpaid the provider (facilities that care for Medicare patients and that receive Medicare payment) and to collect information for use in setting prospective payment rates (Wage Index, DSH adjustment, IME/GME, outliers)
  2. Overall data type - Hospital expenditures
  3. Dataset type - Longitudinal
  4. Data source - Claims
  5. Data level - Hospital level
  6. Geographic location of the data collection sites - United States
  7. Sponsor, manager, or home institution - Centers for Medicare & Medicaid Services (CMS) Healthcare Cost Report Information System (HCRIS)
  8. Date range - 2010 - 2020
  9. Geolocation data - Street, P. O. Box, city, state, zip code, and county
  10. Dates - MM/DD/YYYY for fiscal intermediary create and receipt date, fiscal year begin and end date, notice of program reimbursement date, certification date, and process date
  11. Hospital identifiers - Medicare provider number and National Provider Identifier (NPI) for hospitals and sub providers; and name for hospitals
  12. Longitudinal tracking - Track providers using the above-mentioned identifiers
  13. Financial variables - Capital-related cost (buildings and fixtures, movable equipment, and other), employee benefits, charges, cost-to-charge ratio
  14. Clinical areas of interest - all
  15. Variables that are uniquely present in this dataset - CMS Cost Reports is the only national data available for all types of providers (non-profit, for-profit, government). HCRIS collects data from the Hospital Cost Report (CMS-2552-96 and CMS2552-10), Skilled Nursing Facility Cost Report (CMS-2540-96 and 2540-10), Home Health Agency Cost Report(CMS-1728-94), Renal Facility Cost Report (CMS-265-94) and Hospice Cost Report (CMS-1984-99), Rural Health Clinic/Federally Qualified Health Center (222-92); and Community Mental Health Center (2088-92)
  16. Database caveats and limitations - HCRIS does not collect all the data from some cost reports (e.g., 2552-96, 2540-96, and 1728-94) cost reports, but just a subset of this data. Cost reports are not required and therefore not available for Federal Hospitals (i.e., Veterans Hospitals and Indian Health Services Hospitals), some children’s hospitals, and Emergency Hospital (hospitals outside of the US)

Applicable methods

  1. Association analysis, such as generalized linear models (GLM) (1, 2)
  2. Causation, such as difference-in-differences (3, 4)
  3. Descriptive analyses (5)
  4. Time to event (6)

High-impact designs

  1. Enrichment of the dataset by linking it with another dataset, such as Healthcare Cost and Utilization Project's (HCUP) State Emergency Department Databases (SEDD) (7)

Data dictionary

To access the data dictionary, click here

Variable categories

  1. Facility type characteristics (e.g., ownership status, type of facility, urban/rural, referral center, transplant center, teaching hospital/program, sole community hospital)
  2. Facility infrastructure (e.g., number of beds, number of interns and residents, employees on the payroll, nonpaid workers)
  3. Health outcomes (e.g., hours patients spend in inpatient/outpatient visits, admitted and not admitted observation bed days, ambulance trip, labor and delivery days)
  4. Hospital's elected payment methods (e.g., prospective, hold harmless)
  5. Hospital Wage Index information (e.g., total salaries, paid hours, average hourly Wage)
  6. Visit (e.g., skilled nursing, physical therapy, occupational therapy, speech therapy, medical social service, home health aide, total visits)
  7. Charges (e.g., skilled nursing; physical therapy; occupational therapy; speech therapy; medical social service; home health aide; medical supply; total SCHIP, Medicaid and uncompensated care charges; inpatient and outpatient charges)
  8. Cost (e.g., cost to charge ratio; total SCHIP, Medicaid, and uncompensated care costs; salaries; insurance; taxes; outside supplier service; inpatient and outpatient cost)
  9. Census data (e.g., number of patients receiving hospice care, average length of stay, census count)

Linkage to other datasets

  • Linkages can be established for any dataset that might have hospital identifiers (Medicare provider number, NPI, name of hospitals) or geolocation data (zip code)