February 2001

Grant Results

SUMMARY

Investigators from the University of Wisconsin-Madison Medical School examined changes in the availability of physicians in US urban areas from 1980 to 1997.

Key Findings
Key findings reported to the Robert Wood Johnson Foundation (RWJF) and the Council on Graduate Medical Education include the following:

  • The number of office-based primary care physicians grew from 1980 to 1997, and availability was higher in non-poverty areas.
  • The number of specialists and hospital-based physicians grew much faster in poverty areas during this period.
  • Physician availability is most strongly associated with the concentration of hospitals in an area.
  • No single policy aimed at altering the medical workforce showed a dramatic impact on physician availability.
  • In 1997, the availability of office-based primary care physicians in both high- and low-poverty areas was below levels considered adequate by a panel of 11 medical workforce experts polled by the principal investigator.

Funding
RWJF provided $197,040 in funding from June 1997 to April 1999 to support the project.

 See Grant Detail & Contact Information
 Back to the Table of Contents


THE PROJECT

Although much progress has been made in understanding the overall supply and demand for physicians in the workforce, the availability of physicians in inner-city areas is an issue that has been poorly understood and generally neglected by researchers. This neglect was especially significant during the 1990s as many states attempted to improve physician availability in urban areas.

In 1987, the project director and others published a study of trends in physician availability in 10 urban areas from 1963 to 1980. They found that the availability of office-based primary care (OBPC) physicians declined over the period, while the availability of office-based specialists and hospital-based physicians increased. The result was an approximate 29 percent net increase in the number of physicians per 100,000 people in the 10 cities studied. The loss of primary care physicians was greater, and the overall increase in physicians was smaller in poverty areas than in non-poverty areas.

This grant from RWJF supported a study of changes in the availability of physicians in US urban areas from 1980 to 1997. The study sought to update and extend the previous research and improve the knowledge base for physician workforce policy. (In two previous RWJF grants, see Grant Results on ID#s 019407 and 024109, the principal investigator and his colleagues explored trends in the ratio of primary care physicians to specialist physicians, the supply of international medical graduates, and minority representation in the physician workforce.)

Although the study described here was originally intended to analyze trends between 1980 and 1995, access to additional data permitted the investigators to extend their analysis to 1997. One particular focus of their analysis concerned the availability of physicians between 1990 and 1997 — a period when many states enacted policies aimed at improving the availability of physicians in underserved areas, particularly those in primary care practice.

Methodology

  • This study collected, combined, and analyzed data from eight sources, including the American Medical Association (AMA) Masterfile extracts of physician addresses, the American Osteopathic Association (AOA) physician addresses, the American Hospital Association (AHA) hospital addresses, data from the US Census on a population and tract-by-tract basis, and managed-care enrollment data. Using these data, the study examined the number of physicians practicing in each Census tract of 25 randomly chosen Metropolitan Statistical Areas (MSAs). Census tracts in which more than 20 percent of families lived below the poverty line in 1990 were designated as "poverty tracts." All others were designated as "non-poverty tracts."
  • Census tract boundaries as defined in the1980 and 1990 national censuses did not always match, and no conversion standard was available from the Census Bureau. The researchers constructed their own software to reconcile the differences.
  • Physician practice location was not always clear from the address data available. In some years, only about half of the physician addresses available were for the physicians' practice location, while the rest were for their home address. In other years it remained unknown, whether the available address was for the home or the practice location. Researchers computed the probability that an address in a given tract was a practice address, and applied these probabilities for all tracts to estimate the number of practice addresses in each tract.

 Back to the Table of Contents


FINDINGS

The principal investigator reported the following findings to RWJF:

  • Between 1980 and 1997, the number of OBPC physicians per 100,000 population increased at roughly the same rate in poverty tracts as it did in non-poverty tracts. In 1997 there were still more OBPC physicians in non-poverty tracts than in poverty tracts. However, the number of specialists and hospital-based physicians grew much faster in poverty tracts than in non-poverty tracts, so that overall physician availability in poverty tracts actually exceeded that of non-poverty tracts in 1995 and 1997.

    The number of physicians per 100,000 persons by poverty tract, type of physician, and year, for 25 MSAs:

    Tract Type Physician Type 1980 1990 1995 1997
    Poverty All 151.38 157.08 211.50 216.37
    OBPC 41.04 41.22 51.65 51.44
    Non-Poverty All 157.22 183.77 199.73 194.87
    OBPC 48.74 55.92 60.00 59.48
    Total All 156.24 179.76 201.41 197.88
    OBPC 47.45 53.71 58.81 58.35
  • The proportion of all physicians who were located in poverty tracts declined between 1980 and 1997. For OBPC physicians, the percentage declined from 14.5 percent to 12.35 percent of all physicians, a decrease of 15 percent. Most of the decline occurred between 1980 and 1990.

    Percentage of all physicians who were located in tracts in which more than 20 percent of families lived below the poverty line in 1990, by type of physician and year, for 25 MSAs:

    Physician Type 1980 1990 1995 1997
    All 16.24% 13.14% 15.00% 15.31%
    OBPC 14.50% 11.54% 12.55% 12.35%
  • In 1997, the availability of OBPC physicians in both poverty and non-poverty tracts remained below adequate levels. In 1997, there were 51 physicians per 100,000 population in poverty tracts and 59 per 100,000 population in non-poverty tracts. In a recent survey by the principal investigator of 11 physician workforce experts, the consensus placed the minimal level of physicians required as between 43 and 53 per 100,000 population (depending on the size and location of the community) and the adequate level at 59.5 to 75.7 per 100,000.
  • The growth in physician availability is most strongly associated with the concentration of hospitals. Hospital-based physicians and office-based specialists increased more rapidly than OBPC physicians, relative to population. The availability of physicians, even OBPC physicians, grew fastest in tracts with hospitals in them during the period from 1990 to 1997.
  • States that implemented many of the more effective workforce policies saw greater improvement in the availability of OBPC physicians between 1990 and 1997 than did states that took few or no effective policy steps. However, no single medical workforce policy showed a dramatic impact on physician availability. Education-based-incentives and financial and administrative support were among the more effective state workforce policies. Policies that relaxed licensing, liability, or charity allowances were seen as less effective.

Limitations

The researchers offered the following caveat:

  • The impact of policy actions on the growth in OBPC physician availability is difficult to gauge and should be approached with caution. Other variables, such as a state's general level of affluence or revenues, might explain the state's ability to enact workforce legislation and its attractiveness as a location for physicians' practices.

Communications

The principal investigators presented an informal update of study results to a workgroup at the 1999 national meeting in Washington, DC, of COGME. They also reported their findings in a chapter of a compendium published by COGME. (See the Bibliography for details.)

The principal investigators are currently preparing manuscripts for publication that will focus on physician availability in urban areas. In addition, they will develop thematic maps to illustrate patterns and trends in the data, with the intention of preparing poster presentations for health policy or health services research conferences.

 Back to the Table of Contents


AFTER THE GRANT

At the request of COGME, the principal investigators submitted a written report in January 2000 that provided estimates of the number of physicians needed to adequately supply underserved areas. The report also estimated the number of participants that will be required during the next few years in the National Health Service Corps — a federal program designed to recruit primary care physicians to practice in underserved area.

 Back to the Table of Contents


GRANT DETAILS & CONTACT INFORMATION

Project

Study of Urban Physician Supply Trends

Grantee

University of Wisconsin-Madison Medical School (Madison,  WI)

  • Amount: $ 197,040
    Dates: June 1997 to April 1999
    ID#:  030968

Contact

David A. Kindig, M.D., Ph.D.
(608) 263-6294
dakindig@facstaff.wisc.edu

 Back to the Table of Contents


BIBLIOGRAPHY

(Current as of date of this report; as provided by grantee organization; not verified by RWJF; items not available from RWJF.)

Book Chapters

Libby D and Kindig D. "Estimates of Physicians Needed to Adequately Supply Underserved Americans Until Universal Coverage." In Update on the Physician Workforce. Resource Paper Compendium. Washington, D.C.: US Department of Health and Human Services, Health Resources and Services Administration, August 2000.

Presentations and Testimony

David Kindig and Donald L. Libby, "Study Results," to a workgroup at the national meeting of the Council on Graduate Medical Education (COGME), April 19, 1999, Washington, D.C.

 Back to the Table of Contents


Report prepared by: Bill Berlin
Reviewed by: Robert Narus
Reviewed by: Richard Camer
Program Officer: Susan Hassmiller