To search for a particular data definition or source, you can also use the Knowledge Base
Workforce Dataset used in the Med School Mapper:
HealthLandscape and the Med School Mapper employ the American Medical Association (AMA) Physician Masterfile to track the graduate footprints of U.S. Medical Schools. The AMA Masterfile is a workforce dataset commonly used in studies of the U.S. physician workforce. Despite its limitations (described below), the AMA Masterfile is a relatively comprehensive resource for understanding physician workforce at the national, state, and local level.
The AMA Masterfile contains information on physician demographics, professional activities (patient care vs. non-patient care), specialties (self-reported), medical education, and certifications and licensures[i]. The AMA retrieves this information from a variety of sources, including physician surveys, surveys of medical schools and residency programs, the Liaison Committee on Medical Education (LCME), Accreditation Council for Graduate Medical Education (ACGME), United States Medical Licensing Exam (USMLE) and state licensing boards for all physicians practicing in the US, including physicians who are not members of the AMA, osteopaths, and international medical graduates practicing in the US.
The AMA Masterfile is not without its limitations and there are alternative sources of information available for physician workforce analysis. For example, the American Osteopathic Association (AOA) has an Osteopathic Physician Masterfile, collecting much of the same information as the AMA Masterfile.[ii] Additionally, the National Provider Identifier (NPI) Number Registry provides demographic information on any physician with an NPI number, which is required for providers who bill Medicare, Medicaid, and most private insurance companies. State licensure boards also collect information on all physicians practicing in their state. As mentioned above, the Med School Mapper currently only utilizes data from the AMA Masterfile.
Breadth: The AMA Masterfile’s goal is to capture every physician in the US, including those who are not members of the AMA, osteopaths, and international medical graduates.
Depth: In contrast to the NPI Registry and other census-style datasets that provide only demographic information, the AMA Masterfile also provides information on physicians’ practices and educational background.
Accuracy: While there has been some debate regarding the accuracy of the AMA Masterfile, previous work by the Robert Graham Center has shown very little duplication of records or invalid values, with a maximum of 22 duplicate entries out of 892,978 in the May 2001Masterfile[iii]. No one actively tends the NPI for the purposes of purging out of date information or providers that have ceased to practice. State licensure data for all 50 states would be exceedingly difficult to collect and maintain on an annual basis, risks overcounting and duplication for physicians that serve in multiple states, and doesn’t capture information for the purposes of workforce study.
Poor response to physician practice changes: Compared with other sources of information, the AMA Masterfile is slow to record changes in practice setting, deaths, or retirements.[iv] A recent study found that in comparison to direct physician survey and medical license analysis, the sensitivity of the AMA Masterfile regarding departure from clinical activity was 9.0%, with a positive predictive value of 52.9%[v]. One possible explanation for this discrepancy is the AMA’s reliance on periodic physician surveys to ascertain physician practice, as it may be years between surveys for a particular physician. When compared to the US Census Bureau Current Population Survey (CPS), the AMA Masterfile, in general, has a larger number of older physicians and a smaller number of younger physicians in the workforce, again indicating that the AMA Masterfile is insensitive with respects to changes in the workforce.[vi]
Self-reporting of specialty: The AMA Masterfile surveys allow physicians to self-report their primary specialty and a secondary specialty they practice. This may overestimate the number of generalist physicians by incorporating specialists who dedicate some of their practice to providing primary care.[vii] According to some calculations, the AMA Masterfile may overestimate the number of primary care physicians by 25%[viii].
These limitations are important to keep in mind, particularly when using the AMA Masterfile to develop physician workforce supply projections. While the Masterfile may not represent a real-time look at the national physician workforce and may overestimate the number of primary care physicians, it is a comprehensive, robust dataset that aids in analyses of the workforce at the national, state, and local level.
[i] AMA, “Description of AMA Physician Masterfile Data Elements,” Accessed at http://www.ama-assn.org/ama1/pub/upload/mm/eProfiles/mm/mfile_elements.pdf on April 21, 2010.
[ii] AOA, “Privacy Statement” Accessed at http://www.do-online.org/iLearn/index.cfm?pageid=ment_privacy on April 21, 2010
[iii] Robert Graham Center, “The Physician Workforce of the United States: A Family Medicine Perspective,” Accessed online at http://www.graham-center.org/PreBuilt/physician_workforce.pdf on April 21, 2010.
[iv] Rittenhouse DR et al, “No Exit: An Evaluation of Measures of Physician Attrition,” Health Services Research, 39:5:1571-88.
[vi] Staiger, DO et al. “Comparison of Physician Workforce Estimates and Supply Projections,” JAMA 2009;302(15):1674-80.
[vii] Grumbach, K et al. “The Challenge of Defining and Counting Generalist Physicians: An Analysis of Physician Masterfile Data.” Am J Public Health 1995;85:1402-7
[viii] Grumbach, K et al. “The Challenge of Defining and Counting Generalist Physicians: An Analysis of Physician Masterfile Data.” Am J Public Health 1995;85:1402-7
Health Professional Shortage Areas - (Data obtained monthly from the HRSA Geospatial Data Warehouse using Adobe GoLive software) This layer displays primary medical care Health Professional Shortage Areas (HPSAs), which may be designated as having a shortage of primary medical care, may be urban or rural areas, population groups, or medical or other public facilities. For more information, see http://bhpr.hrsa.gov/shortage/.
Medically Underserved Areas/Populations - (Data obtained monthly from the HRSA Geospatial Data Warehouse using Adobe GoLive software) This layer displays Medically Underserved Areas/Populations (MUA/Ps), which may be a whole county or a group of contiguous counties, a group of county or civil divisions, a group of urban census tracts in which residents have a shortage of personal health services, or Exceptional/Governor designated. Medically Underserved Populations (MUPs) may include groups of persons who face economic, cultural or linguistic barriers to health care. For more information, see http://bhpr.hrsa.gov/shortage/.
111th Congressional Boundaries - Boundaries of every congressional district and its seat-holder in the 111th U.S. Congress.
State Upper Senate
State Lower House