The CMS 2010 PSBS PUF is an aggregated file in which each record summarizes information for a particular physician (individual or organization) identified by the NPI. However, the CMS 2010 PSPS PUF contains multiple records per NPI as a single rendering physician/supplier can potentially bill in multiple states. As a result, a NPI summary record exists for each unique combination of NPI, state of service, and hospital referral region (HRR). There are different categories of summary measures in the CMS 2010 PSBS PUF including beneficiary demographics and utilization measures. Summarized beneficiary measures include beneficiary demographics (e.g., breakdown of beneficiary counts by sex, age, race) and health status information (e.g., breakdown of beneficiary counts by selected chronic conditions). Summarized utilization measures include the number of physician/supplier events defined using Berenson-Eggers Type of Service (BETOS) Codes. The CMS 2010 PSBS PUF contains 135 variables.
Content of the PUF
The most important aspects of the CMS 2010 PSBS PUF are as follows:
- It contains information from 100% of the 2010 Medicare FFS beneficiary population and various levels of physician/supplier characteristics by Provider ID, Provider State of Service, and Provider Hospital Referral Region.
- It contains detailed breakdown of beneficiaries served by each physician/supplier including sex, age category, race, dual eligibility status, and chronic conditions.
- Every beneficiary count variable in the PUF contains at least 11 beneficiaries. The value is suppressed if there are less than 11 beneficiaries for a particular break down (e.g., race).
- The fact that the PUF does not provide beneficiary-level information reduces concern about the privacy of the beneficiaries. Nevertheless, the PUF is tested rigorously to ensure that it can be released to the public without compromising beneficiaries’ privacy.
In addition to the General documentation file, there are two documentation files for the IPBS PUF:
- The data dictionary and codebook file contains information about each variable on the file and its values.
- A data users document which gives the file layout for the downloadable CSV file, as well as SAS program code for creating a SAS dataset with variable formats.