Hospital Intelligence Brief
What Is in the SPARCS De-Identified Inpatient File, and What Is Left Out
The public inpatient file keeps enough detail for serious analysis and removes what could identify a patient. Knowing the exact fields and exclusions prevents wrong conclusions.
Short answer
The public SPARCS inpatient discharge file includes facility, county, patient age group, gender, race, ethnicity, three-digit ZIP, length of stay, APR DRG and severity, payment typology, charges and costs. It leaves out names, exact dates, exact ages, individual secondary diagnoses and procedures, and revenue codes.[1]
What is the SPARCS de-identified inpatient file?
The annual inpatient discharge file is the best-known SPARCS public use dataset. Each year’s file, published on the Department of Health’s open data portal, contains discharge-level detail on patient characteristics, diagnoses, treatments, services and charges, with data protected under HIPAA removed.[1] It is detailed enough that people expect it to answer almost any question about an inpatient stay. It was also built to protect patients, so some details are coarsened or absent on purpose.
The Department describes this file as one of its public use files, available free of charge to policymakers, researchers and the public.[2] This article draws on the metadata for the 2024 file. Reading the field list and notes before an analysis begins is the fastest way to see what is possible.
Which fields does the file contain?
The metadata for the 2024 inpatient file lists fields in five groups. For place, it includes health service area, hospital county, operating certificate number, permanent facility identifier and facility name, and patient ZIP code. For the patient, it includes age group, gender, race and ethnicity. For the stay, it includes length of stay, type of admission, patient disposition, discharge year, an emergency department indicator and birth weight.[1]
Clinical groupings are a major strength. The file carries CCSR diagnosis and procedure categories and descriptions, the APR DRG, the APR major diagnostic category, APR severity of illness, risk of mortality, and a medical or surgical description. For payment, it provides up to three payment typology fields, plus total charges and total costs.[1]
Together these fields support meaningful comparisons: how many discharges a hospital had for a given APR DRG, how severity varies between facilities, how long patients stayed, and how payer mix differs between counties.
How is the data de-identified?
The notes describe the protections in some detail. The file does not contain protected health information as defined under HIPAA, and elements considered identifiable are redacted. Dates have the day and month removed. Age is given only in groups: 0 to 17, 18 to 29, 30 to 49, 50 to 69, and 70 or older. Patient ZIP codes are limited to the first three digits, and are blank where the population is under 20,000, where the cell size in the patient classification strata is under 10, or where the record is an abortion. Out-of-state ZIP codes appear as “OOS”.[1]
| Element | How it appears in the public file |
|---|---|
| Dates | Day and month removed; discharge year retained |
| Age | Five groups: 0–17, 18–29, 30–49, 50–69, 70 or older |
| Patient ZIP code | First three digits only; blank in small populations, small cells and abortions; “OOS” for out of state |
| Abortion discharges | Facility identifier, name, health service area, county and operating certificate redacted |
| Secondary diagnoses, procedures, revenue codes | Excluded |
The notes also say that, for years beginning with 2011, the variable showing the expected payer changed from source of payment to payment typology, effective December 2015.[1] That affects any comparison of payer mix over time.
What can and cannot be done with it?
These rules mean a user can compare facilities, services, severity levels and payer mix, estimate market share by county or service, and study length of stay. A user cannot follow a patient across stays, build exact age cohorts, study the day of the week or the season of admission, or analyze secondary diagnoses and individual procedures. The inference follows from the field list and exclusions, not from a separate statement by the Department.
The notes do not explain how total costs are derived from charges. Anyone treating the two fields as interchangeable should read the data dictionary first, because a charge is a facility’s billed amount and a cost is an estimate, and the two can differ substantially.[1]
Geography needs care. A three-digit ZIP prefix covers an area larger than a town and does not align neatly with a county, so a market analysis based on it will be approximate. The county fields are a more reliable alternative for county-level work.
What if the public file is not enough?
Some questions need excluded detail: exact dates for time-to-treatment studies, secondary diagnoses for comorbidity analysis, or a patient identifier to follow readmissions. For those, the Department offers limited and identifiable data through an application process, with fees, review and conditions.[2] That process is described in its own article in this series.
Before applying, check whether the public file can answer a modified version of the question. A request for restricted data takes time and money, and a well-designed analysis of the public file is often sufficient for planning and market work.
Frequently asked questions
What fields are in the SPARCS inpatient discharge file?
Facility and place fields, patient age group, gender, race, ethnicity, length of stay, admission type, disposition, APR DRG and severity, CCSR categories, payment typology, charges and costs.[1]
Does the SPARCS public file include exact ages or dates?
No. Age is given in five groups and dates have the day and month removed.[1]
Can I find secondary diagnoses in the public SPARCS file?
No. Individual secondary diagnoses and procedures, and revenue codes, are excluded.[1]
How precise is patient location in the file?
Patient ZIP codes are limited to the first three digits and are blank in small populations or small cells.[1]
Where can I get more detailed SPARCS data?
Through the Department’s application process for limited or identifiable data, which carries fees, review and conditions.[2]
What to do next
- Match your analysis plan to the field list before downloading.
- Note the redactions that affect your question.
- Test on a small subset first.
- Document exclusions and the file year in any result you share.
Explore Your Hospital’s Opportunities
Two Pillars can help hospital leadership investigate market, financial, operational and strategic questions using relevant organizational data, public information and a structured assessment process. A good first step is a SPARCS review that confirms your obligations, contacts and calendar, and sets up a simple routine your team can keep.
Sources
- New York State Department of Health, Hospital Inpatient Discharges (SPARCS De-Identified): 2024. Accessed October 2026.
- New York State Department of Health, SPARCS Data Access. Accessed October 2026.
Sources checked October 2026. This guide is general information for healthcare leaders and is not legal, financial, clinical or compliance advice. Figures and rules are cited to the sources below and may have changed since they were accessed, so confirm anything you rely on against the current source. Two Pillars & Co is independent of the organizations cited. See our editorial policy.