Sensitivity of Medicare Data to Identify Oncologists

Area of Science:

• Health Services Research
• Oncology
• Data Science

Background:

• Health services researchers often use Center for Medicare and Medicaid Services (CMS) or American Medical Association (AMA) data to identify oncologists.
• The accuracy of CMS data for identifying oncologists has not been fully established.

Purpose of the Study:

• To assess the completeness of CMS data resources for identifying oncologists compared to the AMA Physician Masterfile.
• To determine the sensitivity of individual and combined CMS files in capturing different types of oncologists.

Main Methods:

• Identified oncologists from three CMS data sources: physician claims, National Plan and Provider Enumeration System Registry, and Medicare Data on Provider Practice and Specialty file.
• Linked CMS and AMA data using a unique physician identifier.
• Calculated sensitivity for identifying oncologists, radiation oncologists (ROs), surgical oncologists (SOs), and medical oncologists (MOs).

Main Results:

• Individual CMS files captured approximately 83% of oncologists from the AMA Masterfile; combined files captured 90.4%.
• Combined CMS data identified 98.2% of ROs, 89.3% of MOs, and 70.1% of SOs.
• CMS specialty data for ROs and SOs not identified as oncologists often corresponded to their AMA subspecialty (radiologists and surgeons, respectively).

Conclusions:

• Combined CMS data effectively identifies most radiation and medical oncologists but has limitations for surgical oncologists.
• The choice between AMA and CMS data depends on the research question and the specific oncologist subspecialty of interest.