Use Case 40: Patient demographics for provider utilization datasets

From Demand-Driven Open Data
(Redirected from Use Case 40)
Jump to: navigation, search

Want to be notified when this use case or any of its datasets change?
You’re signing up to be notified on changes to this use case and associated data sources
* indicates required


Use case summary


There's a need to better understand the data in provider utilization datasets by incorporating patient demographics. That's because Medicare patients do not all look alike.

  • A provider who sees a majority of patients with disabilities who are under the age of 65 should have their other data points interpreted completely differently.)
  • Similarly, a doctor whose average patient age is 95 should be evaluated differently then one whose patient age is 67.
  • Tranches of ages and gender should be considered for this report. To avoid the risk re-identification for these general cases, it would be fine to provide no patient counts under 50.


  • Value to industry/public:
    • This information enables industry analysts and media more effectively interpret the previously published provider utilization data. Utilization data has sometimes been used incorrectly to imply waste, fraud and abuse.
    • Consumers benefit from having more accurate information when choosing their healthcare provider.
    • Additionally, the potential misinterpretation of utilization data could have an adverse effect on the decisions healthcare providers make when treating patients.


  • DocGraph
  • ProPublica

Current data and limitation

  • Data source: Provider utilization dataset.
  • Limitations: Doesn't give sufficient insight into demographics to distinguish populations.


  • Fields:
    • Original fields: NPI, HCPCS (CPT), number of services, number of beneficiaries, average submitted charge
    • Beneficiary demographic fields: Age, sex, race, Medicare / Medicaid entitlement, chronic conditions, risk scores
  • Joins between datasets: Demographic fields should be consistent and joinable with other datasets for analytics
  • Update frequency: Same as provider utilization datasets
  • Lag time: Same as provider utilization datasets
  • History: Same as provider utilization datasets
  • Delivery mechanism: Same as provider utilization datasets


Datasets provided

An update has been made on 10/5/2015 to the relevant tables of CY 2012 and CY 2013 "Medicare Provider Utilization and Payment Data: Physician and Other Supplier" datasets that includes the items asked for.

  • Provider-level summary tables now include aggregated information on the following beneficiary attributes:
    1. Age:
      • "Average Age of Beneficiaries
      • "Number of Beneficiaries Age...": Less than 65, 65 to 74, 75 to 84, Greater than 84
    2. Sex:
      • "Number of Female Beneficiaries", "Number of Male Beneficiaries"
    3. Race:
      • "Number of Beneficiaries" who are... Black or African American, Asian Pacific Islander, Hispanic, American Indian/Alaska Native, Race Not Elsewhere Classified
    4. Medicare and Medicaid entitlement:
      • "Number of Beneficiaries With Medicare Only Entitlement"
      • "Number of Beneficiaries With Medicare & Medicaid Entitlement"
    5. 16 chronic conditions:
      • "Percent (%) of Beneficiaries Identified With..." Alzheimer’s Disease or Dementia, Asthma, Atrial Fibrillation, Cancer, Chronic Kidney Disease, Chronic Obstructive Pulmonary Disease, Depression, Diabetes, Heart Failure, Hyperlipidemia, Hypertension, Ischemic Heart Disease, Osteoporosis, Rheumatoid Arthritis / Osteoarthritis, Schizophrenia / Other Psychotic Disorders, Stroke
    6. Risk scores:
      • "Average HCC Risk Score of Beneficiaries"

  • File locations
CY 2012: "Medicare Physician and Other Supplier Aggregate table, CY2012, Microsoft Excel (.xlsx)"
CY 2013: "Medicare Physician and Other Supplier Aggregate table, CY2013, Microsoft Excel (.xlsx)"

Limitations and challenges

  • Users should keep in mind the limits of these datasets:
    • The data includes only care delivered to Medicare beneficiaries in the fee-for-service (Medicare Part B) program. So it may not represent a provider's entire patient population. (It doesn't capture patients who are uninsured or enrolled in Medicare Advantage, Medicaid coverage or private health insurance.)
    • For services provided in a facility setting (i.e., in outside a physician's office), typically only the payment to the provider is included, but not to the facility.
  • There's an associated "Lookup Tool" at that's populated from the same data. But as of 11/3/2015, it lacked the new demographic data.

Want to be notified when this use case or any of its datasets change?