NY SPARCS Statewide Planning and Research Cooperative System - onetomapanalytics/Meta_Data GitHub Wiki

NY SPARCS - Statewide Planning and Research Cooperative System

General description

  1. Database primary purpose - Contribute to the goal of providing high-quality medical care at a reasonable cost to the inhabitants of the State by serving as an information source for organizations and agencies seeking to promote the efficient delivery of health care services.
  2. Overall data type - Health outcomes, economics
  3. Dataset type - Cross-sectional
  4. Data source - Claims
  5. Data level - Patient level
  6. Geographic location of the data collection sites - State of New York
  7. Sponsor, manager, or home institution - NYS Department of Health
  8. Date range - 2016 - 2020
  9. Geolocation data - Zip codes, address line 2, city, state, county code, SPARCS region code
  10. Dates - Cover period (CCYYMMDD), admission and discharge dates, hour and weekday, length of stay, leave of absence days
  11. Hospital identifiers - Permanent Facility Identifier (PFI), facility name, operating certificate number, and national provider ID (NPI)
  12. NPI - Facility’s National Provider ID
  13. Longitudinal tracking - Track patients within hospitals (Patient Control Number, assigned by the provider), patients across hospitals (Medical Record Number - MRN, assigned by the Medical Records Department), and providers (provider and physician state license number)
  14. Financial variables - Source of payment, insurance policy number, payer ID, covered and non-covered days, alternate care days, leave of absence days, expected reimbursement, and cover by Workers' Compensation (WC) and No-Fault Insurance (NF)
  15. Clinical areas of interest - All
  1. Other - information on Data Elements is available as (Appendices on their website)[https://www.health.ny.gov/statistics/sparcs/sysdoc/appendix.htm]

Applicable methods

  1. Association, such as multivariable regressions (1), multiple linear regression (2), multivariate logistic regression (3, 4)
  2. Machine learning (5)
  3. Propensity score (6)
  4. Sensitivity analysis (7, 8)
  5. Subgroup analysis (1)
  6. Univariate logistic regression (9)

High-impact designs

  • Establish the association of surgical and hospital volume and patient characteristics to develop a specific risk stratification model (9)

  • Investigate gender disparities in clinical productivity and case volume (10)

  • Describe comorbid conditions and patterns of rehospitalizations (7)

Data dictionary

To access the NY SPARCS Inpatient output data dictionary, click here.

To access the NY SPARCS Outpatient output data dictionary, click here

Variable categories

  1. Patient demographics [e.g., age, sex, race, ethnicity, residence indicator (e.g., homeless, foreign)]
  2. Hospital discharge records (i.e., inpatient discharges, ambulatory surgery visit, emergency department admission, and outpatient visits)
  3. Newborn segment (e.g., discharge status, mother's MRN, birth weight)
  4. Special program (i.e., entitlement to Medicaid benefits due to disability, family planning procedures, physically handicapped children's program, or the special funding project)
  5. Diagnosis (e.g., admission type, principal, and "other" diagnosis code)
  6. Procedure (e.g., principal and "other" procedure code, pre-op and post-op days)
  7. Charges (total charges, ancillary and accommodation total charges, non-covered charges)
  8. HIPAA segment (i.e., indication of AIDS/HIV and abortion in discharge records)

Linkage to other datasets

  • NY SPARCS can theoretically be linked to any dataset where the hospital name and location (i.e., city, zip code) are mentioned, although that linkage might require the use of NLP and manual checking.