A Growing Hole in the Safety Net: Physician Charity Care Declines Again

Tracking Report No. 13

March 2006

Peter J. Cunningham, Jessica H. May

Extending a trend that has persisted for at least a decade, the share of U.S. physicians who delivered charity care fell to 68 percent in 2004-05, down from 76 percent in 1996-97, based on a nationally representative study conducted by the Center for Studying Health System Change (HSC). This sustained erosion in physician charity care is particularly worrisome given the concurrent rise in the ranks of uninsured Americans, especially during the opening years of the 2000s. Reductions in charity care were evident across virtually all major specialties, practice configurations, income brackets, and geographic areas. Growing financial strain and evolving practice structures likely contribute to this ongoing decline in charity care among physicians.

Physician Charity Care Continues to Decline

For decades, physicians have served as a cornerstone of the nation's health care safety net by offering free or discounted services to uninsured patients, whether in their own offices or through volunteer work at free clinics. Yet mounting financial demands and time constraints, combined with ongoing shifts in the medical marketplace, appear to be driving down the percentage of doctors who offer such care. The share of physicians delivering any free or reduced-price care dropped to 68.2 percent in 2004-05 from 71.5 percent in 2000-01, according to HSC's nationally representative Community Tracking Study Physician Survey (see Data Source). This pattern extends back to at least 1996-97, when more than three out of four physicians (76.3%) reported providing some form of charity care (see Figure 1 and Table 1).

Even though the proportion of doctors delivering charity care has fallen, the absolute number of physicians engaged in charity care has stayed relatively constant. This is because the overall pool of practicing U.S. physicians grew from roughly 347,000 in 1996-97 to 397,000 in 2004-05. Moreover, among those doctors who did provide charity care, the volume of care delivered showed little meaningful change over this timeframe. Between 1996-97 and 2004-05, both the average hours devoted to charity care and the percentage of practice time allocated to it edged down slightly, but neither shift reached statistical significance.

Nonetheless, the supply of physician charity care has diminished in relation to the size of the uninsured population. Total charity care hours per 100 uninsured individuals fell from 7.7 hours in 1996-97 to 6.3 in 2004-05, representing an 18 percent reduction.(1) The bulk of this drop occurred after 2000-01, driven primarily by a sharp increase in the number of uninsured Americans, which rose from 39.6 million in 2000 to 45.5 million in 2004.

Table 1: Physician Charity Care, 1996-97 to 2004-05

* Change from 2000-01 is statistically significant at p <.05. # Change from 1996-97 is statistically significant at p <.05. Source: Community Tracking Study Physician Survey

Physician Specialty

The proportion of doctors offering charity care fell across every major specialty category (see Table 2), as well as across all geographic regions and in both metropolitan and rural settings (latter data not displayed). Surgical specialists remained the most likely among specialty physicians to deliver charity care, in large part because many are obligated to serve on call at hospitals and consequently have less ability to choose whether to treat uninsured patients. Pediatricians were the least inclined to provide charity care, which may reflect the reality that fewer children lack insurance coverage thanks to more expansive public eligibility criteria compared with those available to adults.

Table 2: Physicians Providing Charity Care, by Physician Characteristics

* Change from 2000-01 is statistically significant at p <.05. # Change from 1996-97 is statistically significant at p <.05. Source: Community Tracking Study Physician Survey

Physician Practice Characteristics

Charity care is receding across nearly all types of physician practices, and doctors are progressively gravitating toward practice settings where charity care is less commonly offered.

Charity care rates are highest among physicians working in solo or small group practices and among those who hold full or partial ownership of their practice, likely because these doctors exercise more control over which patients they treat and because there are fewer organizational hurdles for uninsured individuals seeking to see them. Approximately 80 percent of doctors in solo practices or small groups (10 physicians or fewer) delivered charity care in 2004-05, a figure that has not shifted significantly since 1996-97.

In contrast, physicians practicing in larger groups and institution-based settings (such as medical schools or hospitals) are considerably less inclined to provide charity care, and the rate among these doctors fell steeply between 1996-97 and 2004-05. Additionally, a shrinking number of physicians operate solo practices, while a growing number have moved into the practice environments where charity care is less prevalent. From 1996-97 to 2004-05, the share of physicians in solo or two-physician practices declined from 40 percent to 31 percent, whereas the percentage in large groups, hospitals, and medical schools rose from 21 percent to 26 percent (data not displayed).

In a similar vein, practice owners are more likely to deliver charity care than non-owners (who tend to be concentrated in large practices and institutional environments). Although charity care declined from 1996-97 to 2004-05 among both owners and non-owners, the proportion of physicians who hold full or partial ownership has been steadily declining, from roughly 62 percent in 1996-97 to 54 percent in 2004-05.

Physician Income

The delivery of charity care has declined among physicians at every income level. Doctors with the highest incomes continue to report the greatest rates of charity care: 75.6 percent of physicians with practice incomes exceeding $250,000 provided charity care in 2004-05, compared with 66.4 percent of those earning less than $120,000. However, physician incomes have fallen in inflation-adjusted terms in recent years, for both primary care doctors and specialists, largely due to constraints imposed by public and private reimbursement rates.(2) Shrinking practice incomes likely make it increasingly difficult for physicians to subsidize the cost of providing free or reduced-price care.

Location of Charity Care

The vast majority of physicians who delivered charity care (more than 70%) indicated that they typically did so within their own practice (see Table 3). Roughly 14 percent of physicians reported that their charity care occurred primarily while on call in a hospital emergency department, about 6 percent in a different practice or clinic (such as volunteering at a free care clinic), and 8.4 percent at some other unspecified location.

Among major specialty categories, pediatricians and family practice physicians were the most likely to deliver charity care at their primary practice. Medical and surgical specialists were the most inclined to provide charity care through their on-call duties at hospitals, though most specialists nonetheless identified their own practice as the usual venue for charity care.

Table 3: Typical Location of Charity Care, 2004-05

Source: Community Tracking Study Physician Survey

Declines in Access to Medical Care

The long-running reduction in physician charity care may have played a role in diminishing access to medical services for uninsured individuals. In 2003, 63.1 percent of the uninsured reported having a regular source of medical care, a decline from 68.6 percent in 1996-97 (see Table 4). Furthermore, the share of uninsured individuals who had visited a physician in the past year dropped from 51.6 percent to 46.1 percent over the same period.

Emergency department visits as a share of all ambulatory care visits rose for the uninsured between 1996-97 and 2000-01, and the proportion of uninsured people with unmet medical needs also increased. However, the downward trajectory in access stabilized between 2000-01 and 2003, and actually showed improvement with respect to unmet medical needs. Ongoing decreases in physician charity care during the early 2000s may have been partially counterbalanced by strengthening of the safety net in certain communities, particularly through the growth of federally funded community health centers.(3)

Table 4: Trends in Access to Medical Care Among the Uninsured

* Change from 2000-01 is statistically significant at p <.05. # Change from 1996-97 is statistically significant at p <.05. Source: Community Tracking Study Household Survey

Implications

Historically, physicians and hospitals have underwritten charity care for uninsured patients in part by charging higher fees to other patients, especially those covered by private insurance. But both government and private payers have increasingly constrained reimbursement levels for medical providers, generating financial pressures that may be curbing physicians' capacity to offer charity care.

Hospital uncompensated care levels have remained relatively steady in recent years, buoyed by private insurance payments to hospitals that have risen somewhat faster than their costs.(4) Yet declining practice incomes among physicians indicate that they have been less effective than hospitals in negotiating fees with private insurers, while also contending with continued limits from public payers. Time pressures on physicians have also intensified, partly because of a pronounced surge in demand for physician services during the late 1990s.(5) Consequently, the decline in charity care between 1996-97 and 2004-05 may reflect a growing number of physicians who feel they can no longer afford or find the time to deliver charity care.

The shift toward larger practice arrangements and reduced ownership stakes is also playing a role in the decline of charity care, though this transition may itself be partly a response to the financial and time pressures confronting physicians. Larger practices in which doctors serve as employees can help insulate them from these pressures, offer greater bargaining power when negotiating fees with private payers, and ease the burden of on-call responsibilities. On the other hand, larger practices may also create more institutional barriers for uninsured patients, and employed physicians typically have less discretion over which patients they see.

The upshot is that uninsured individuals likely must depend even more heavily on formal safety net providers, including community health centers, other free clinics, and public hospitals -- or they simply receive less medical care. This also implies a transfer of the financial burden for treating the uninsured away from physicians and toward taxpayers, through both direct and indirect subsidies to major safety net providers. The recent expansion of federally supported community health centers may have helped arrest the decline in access to care among uninsured Americans that was observed during the late 1990s, even though overall federal safety net expenditures -- while growing -- have not kept pace with the rising number of uninsured people.(6) Without measures to halt and even reverse the increase in the uninsured population, safety net resources will likely become even more strained under the weight of rising demand.

Notes

1. The overall number of charity care hours per 100 uninsured people was calculated as the ratio of the total weighted number of charity care hours provided (derived from the Community Tracking Study Physician Survey) to the total number of uninsured, then multiplied by 100. The total number of uninsured is based on the Current Population Survey (Fronstin, Paul, Sources of Health Insurance and Characteristics of the Uninsured: Analysis of the March 2005 Current Population Survey, Issue Brief No. 287, Employee Benefit Research Institute, November 2005).

2. Reed, Marie C., and Paul B. Ginsburg, Behind the Times: Physician Income, 1995-1999, Data Bulletin No. 24, Center for Studying Health System Change, Washington, D.C. (March 2003). A forthcoming publication from the Center for Studying Health System Change documents a comparable trend of declining physician income in real terms across specialties for the period 1995-2003.

3. Hoadley, John F., Laurie E. Felland and Andrea B. Staiti, Federal Aid Strengthens Health Care Safety Net: The Strong Get Stronger, Issue Brief No. 80, Center for Studying Health System Change, Washington, D.C. (April 2004).

4. Dobson, Allen, Joan DaVanzo and Namrata Sen, "The Cost-Shift 'Hydraulic:' Foundation, History, and Implications," Health Affairs, Vol. 25, No. 1 (January/February 2006).

5. National Center for Health Statistics, Health, United States 2005 (Table 88), Hyattsville, Md. (2005).

6. Hadley, Jack, et al., Federal Spending on the Health Care Safety Net from 2001-2004: Has Spending Kept Pace with the Growth in the Uninsured? The Urban Institute (November 2005).

Data Source

This Tracking Report draws primarily on findings from the HSC Community Tracking Study Physician Survey, a nationally representative telephone survey of physicians engaged in direct patient care across the continental United States, administered in 1996-97, 1998-99 (results omitted for clarity of presentation), 2000-01, and 2004-05. The physician sample was drawn from the American Medical Association and the American Osteopathic Association master files and encompassed active, nonfederal, office- and hospital-based physicians who devoted at least 20 hours per week to direct patient care. Residents and fellows were excluded. The 1996-97, 1998-99, and 2000-01 surveys each gathered data on approximately 12,000 physicians, while the 2004-05 survey collected responses from more than 6,600 physicians. Response rates ranged from 52 percent to 65 percent.

Additionally, this Tracking Report incorporates findings from the HSC Community Tracking Study Household Survey, a nationally representative telephone survey of the civilian, noninstitutionalized population. Data were supplemented by in-person interviews with households that lacked telephones to ensure adequate representation. The 1996-97 and 2000-01 surveys each captured information on roughly 60,000 individuals, while the 2003 survey included responses from approximately 46,600 persons. Response rates ranged from 57 percent to 60 percent. More detailed information on survey methodology can be found at www.hschange.org.

Sources and Further Reading

KFF: The Uninsured Population — Kaiser Family Foundation data on the growing uninsured population that underlies the rising demand for physician charity care.

AHRQ: Health Care Safety Net — Federal research on safety net providers and their role in delivering care to uninsured and underserved populations.

Health Affairs: The Cost-Shift Payment Hydraulic — Peer-reviewed analysis of cost-shifting dynamics between insured and uninsured patients, cited in this study of declining charity care.

U.S. Census Bureau: Health Insurance Coverage Data — Federal statistics on uninsured rates used to measure the ratio of charity care supply to the size of the uninsured population.

CDC: Health, United States — Annual Report — National Center for Health Statistics annual report cited in this study for data on uncompensated care and safety net trends.