HSC Tracking Report No. 6 -- Results from the Community Tracking Study
December 2002 -- by Peter J. Cunningham
Between 1997 and 2001, the share of physicians providing any charity care declined from 76.3 percent to 71.5 percent, according to a study by the Center for Studying Health System Change (HSC). The proportion of physicians receiving income from Medicaid also edged downward, from 87.1 percent in 1997 to 85.4 percent in 2001. While the modest decrease in Medicaid participation did not appear to have had adverse effects on access to physicians among Medicaid beneficiaries, the larger drop in charity care provision was consistent with growing evidence of worsening access for uninsured individuals. Budget pressures facing state governments raised the prospect of Medicaid payment freezes or reductions, which could further erode physician willingness to serve low-income populations.
Many low-income Americans -- particularly those enrolled in Medicaid or lacking any insurance -- have long faced difficulty finding physicians willing to accept them. Low Medicaid reimbursement rates have persistently acted as a barrier to physician participation, and constraints on payments from managed care and other sources have progressively reduced the ability of physicians to cross-subsidize free care for uninsured patients. These financial disincentives came at a time when Medicaid enrollment was rising, the ranks of the uninsured were growing, and the need for physician services among these populations was increasing.
Despite these headwinds, the majority of physicians continued to participate in Medicaid and to provide at least some charity care as of 2001. The proportion deriving revenue from Medicaid fell only slightly, from 87.1 percent to 85.4 percent over the four-year period. The decline in charity care was more pronounced, dropping from 76.3 percent to 71.5 percent of physicians. These figures are drawn from the HSC Community Tracking Study Physician and Household Surveys, both nationally representative telephone surveys conducted over rolling 12-month periods in 1996-97, 1998-99, and 2000-01.
Most physicians who treated Medicaid patients or provided charity care saw relatively few of them. Among physicians with any Medicaid revenue, more than half derived 10 percent or less of their total practice revenue from the program, and only about one-quarter (25.6 percent) received more than 20 percent of revenue from Medicaid. Between 1997 and 2001, the percentage of low-volume Medicaid providers decreased (from 56.6 percent to 53.1 percent), while the share of high-volume providers increased (from 22.2 percent to 25.6 percent). This shift indicates a growing concentration of Medicaid patients in fewer practices and suggests that the slightly smaller number of practices serving Medicaid were each seeing a greater volume of patients.
For charity care, the trend moved in the opposite direction. Among physicians providing any uncompensated care, 70.2 percent spent less than 5 percent of their total practice time on it, while 29.8 percent devoted 5 percent or more of their time to charity care. Unlike Medicaid, where high-volume providers increased, the proportion of high-volume charity care providers declined between 1997 and 2001 (from 33.5 percent to 29.8 percent), and the share of low-volume providers grew (from 66.5 percent to 70.2 percent). Not only were fewer physicians providing charity care, but those who continued were spending less time doing it.
Beyond overall participation rates, physicians also limited the number of new Medicaid and uninsured patients they would accept into their practices. In 2001, approximately one-fifth of physicians were not accepting any new Medicaid patients, and 16 percent were not accepting any new uninsured patients. These figures compare with less than 5 percent of physicians who were declining all new Medicare or privately insured patients. About half of physicians accepted all new Medicaid patients in 2001, and less than half (44 percent) accepted all new uninsured patients, compared with 71.1 percent who accepted all new Medicare patients and 68.2 percent who accepted all new privately insured patients.
The proportion of physicians accepting no new Medicaid patients rose from 19.4 percent in 1997 to 20.9 percent in 2001. More physicians were also closing their practices to new privately insured and Medicare patients over the same period, indicating that broader capacity pressures -- from both public and private payers -- were contributing to the increases in closed practices, beyond Medicaid-specific factors alone. Among physicians who accepted most or all new privately insured patients, the share refusing all new Medicaid patients held steady at roughly 16 percent, suggesting that cost pressures and rising patient demand across the board may have been driving the general growth in closed practices.
Physicians who were heavily involved in managed care were increasingly closing their practices to new Medicaid patients. Among physicians who received more than 75 percent of their practice revenue from managed care, the proportion with practices closed to new Medicaid patients grew from 20.7 percent in 1997 to 27.3 percent in 2001. Among physicians in group- or staff-model HMOs, the share not accepting new Medicaid patients rose from 15.1 percent to 21.7 percent over the same period. By contrast, physicians who derived less income from managed care did not change their Medicaid acceptance rates.
Physicians with heavy managed care involvement were also increasingly closing their practices to new Medicare patients. Among those receiving 76 percent or more of their revenue from managed care, the percentage declining new Medicare patients rose from 12.7 percent to 15.7 percent between 1997 and 2001. The percentage refusing new privately insured patients held steady at about 4 percent. This pattern of selectively closing to Medicaid and Medicare while continuing to accept private-pay patients may be linked to the exit of many commercial health plans from Medicaid and Medicare managed care, as well as low payment rates from those programs. The departure of these plans may have left some provider networks no longer contracted to serve Medicaid or Medicare enrollees.
Earlier research had established that physicians in states with higher Medicaid reimbursement rates are more likely to accept Medicaid patients. This pattern was partially confirmed in the 12 nationally representative communities where HSC conducted site visits. In 2001, Northern New Jersey had the highest proportion of practices closed to Medicaid (38.3 percent) and one of the lowest Medicaid reimbursement rates for office visits ($16). By contrast, Cleveland, Boston, and Little Rock had higher Medicaid fees and smaller percentages of practices closed to new Medicaid patients.
However, not all communities followed this pattern. Phoenix, for example, had a relatively high proportion of practices closed to Medicaid despite having some of the highest Medicaid fees ($45 for an office visit). A booming population in the Phoenix area may have fueled increased demand for physician services, leading to severe capacity constraints among medical providers. Physicians there may have responded by limiting Medicaid patients in favor of more profitable privately insured patients. Other community-level factors such as the prevalence of Medicaid managed care, administrative rules, and broader market dynamics also contributed to differences in Medicaid acceptance rates.
Significant changes in Medicaid acceptance also occurred within communities over the study period. In Cleveland, the percentage of physicians with closed practices decreased significantly, possibly reflecting a recent 25 percent increase in Ohio s Medicaid office visit reimbursement. In Seattle, by contrast, practices closed to Medicaid increased significantly, as aggressive cost containment efforts in Washington state led several health plans to withdraw from Medicaid managed care after physicians refused to participate at the low payment levels offered.
The decline in physician Medicaid participation was modest through 2001, and evidence of reduced access to physicians among Medicaid beneficiaries was limited. The proportion of Medicaid enrollees with a usual source of care decreased slightly (from 92.9 percent to 90.6 percent), mirroring an overall national decline. Actual utilization of physician services by Medicaid enrollees did not change during this period. The more substantial reductions in charity care, however, aligned with evidence of declining access among uninsured individuals. The percentage of uninsured persons with a usual source of care fell from 68.6 percent to 64.2 percent, and the proportion seeing a physician dropped from 51.5 percent to 46.6 percent between 1997 and 2001.
Continued financial pressures on physicians raised the prospect of further erosion in their willingness to serve Medicaid patients and the uninsured. States facing serious budget deficits were considering reducing or freezing Medicaid physician reimbursement to control program costs. Rising health care costs and reductions in other provider payments could further constrain physicians ability to cross-subsidize uncompensated care. Access to physicians was just one dimension of a broader set of cost-containment pressures that also threatened access to hospitals and prescription drugs for Medicaid and uninsured populations. Ongoing monitoring of physician involvement with Medicaid and charity care would be essential for assessing access trends.
This Tracking Report draws on findings from the HSC Community Tracking Study Physician and Household Surveys conducted in 1996-97, 1998-99, and 2000-01. Both are nationally representative telephone surveys. Each survey was conducted over a 12-month period. The Physician Survey covers nonfederal patient care physicians who spend at least 20 hours a week in direct patient care. Each round of the survey includes approximately 12,500 physicians, with response rates ranging from 59 percent to 65 percent. The Household Survey covers the civilian, noninstitutionalized population, with data supplemented by in-person interviews of households without telephones. Each round includes approximately 60,000 people, and response rates range from 59 to 65 percent.
For additional HSC research on physician participation in Medicaid, visit the
This research was published in December 2002 examining physician participation in Medicaid and charity care. For current resources: