The Uninsured Getting Care:

Originally published by the Center for Studying Health System Change (HSC). HSC was a nonpartisan policy research organization funded principally by the Robert Wood Johnson Foundation.

The Uninsured Getting Care: Where You Live Matters

Issue Brief No. 15 | September 1998 | Peter J. Cunningham, Peter Kemper

A substantial number of Americans -- 41 million people at the time of this research -- lacked health insurance, reflecting a 16 percent jump in the uninsured population since 1990. Research consistently demonstrated that people without coverage faced far greater obstacles than the insured when trying to obtain necessary medical services. This Issue Brief examines findings from HSC's Community Tracking Study, which revealed that the capacity of uninsured individuals to secure needed care differed markedly from one community to another. The study further showed that the personal demographics of the uninsured accounted for very little of this geographic disparity. These results represented an important early step toward helping policymakers grasp how the structure of local communities and their safety net systems shaped the ability of medically indigent populations to receive care.

Wide Community Variations

Health care delivery has always been a fundamentally local affair, so some degree of geographic variation in access for the uninsured was expected. What caught researchers off guard was the sheer magnitude of these disparities. For uninsured Americans, residence turned out to be a powerful determinant of how hard it was to obtain medical attention.

Across the nation, roughly 31 percent of uninsured individuals reported that they either failed to receive medical care they needed in the preceding year or had to delay getting it. That figure was more than double the corresponding rate among the privately insured. At the community level, the spread was dramatic: access difficulties exceeded 40 percent in Lansing and Cleveland -- the highest-rate communities -- while falling below 20 percent in Orange County, which had the lowest rate.

Only about 15 percent of the cross-community variation in obtaining needed care could be attributed to differences in the characteristics of uninsured individuals themselves -- factors like health status, age, sex, household income, and race or ethnicity. Much of the remaining gap likely stemmed from attributes of the communities that shaped access for the uninsured, though the study did not directly evaluate these. Identifying such community-level factors proved far more challenging than measuring personal characteristics, but probable influences included the number and capacity of safety net providers available to treat the medically indigent, state programs subsidizing charity and uncompensated care, and the degree of managed care penetration.

Non-health-system factors could also play a role, including residents' perceptions of their own ability to access medical services and the overall wealth and population size of a given area. Take the greater Newark area and Orange County as cases in point: uninsured people in both communities reported comparatively low rates of difficulty obtaining care. Both regions were relatively affluent (despite having concentrated pockets of poverty) and boasted an above-average supply of physicians, suggesting deeper health care resources. This may partly explain why Newark physicians donated more charity care -- averaging about 10 hours per month -- than doctors in any other study site, according to HSC's Physician Survey.

Both Newark and Orange County sit within the two largest metropolitan areas in the country -- New York and Los Angeles. Larger metro areas may broaden the options available to uninsured residents by giving them access to safety net facilities in adjacent jurisdictions. Some medically indigent patients in Orange County, for instance, traveled to the more extensive public health facilities in neighboring Los Angeles County, potentially improving their access to services.

While self-reported difficulty getting medical care was the most direct access measure available in the HSC survey, it reflected patients' own perceptions rather than a clinical evaluation of unmet need. HSC separately examined less direct access indicators, such as the share of uninsured without a regular source of care, the volume of ambulatory visits, and the distance from home or work to the nearest physician's office. These measures also varied considerably across communities, but their geographic patterns did not closely match the patterns seen in self-reported access difficulties. HSC researchers concluded that substantially more investigation was needed to untangle how safety net differences, local market dynamics, and unique community characteristics influenced access for uninsured populations.

Comparisons With The Insured Population

In line with prior research, the HSC study confirmed that insured Americans had considerably less trouble obtaining medical care than their uninsured counterparts nationwide. Just 15 percent of privately insured individuals said they had difficulty getting care. Although access for the insured also varied by community -- ranging from 12 percent in Lansing to 18 percent in Phoenix -- the degree of variation was narrower than the spread observed among uninsured populations.

Notably, the geographic pattern of access problems for the uninsured did not mirror the pattern for the privately insured. Communities where uninsured residents struggled most were not necessarily the ones where insured residents faced the greatest barriers. Orange County illustrated this disconnect clearly: uninsured people there had fewer access problems than uninsured people in most other sites, yet the privately insured in Orange County reported more difficulty than their insured peers elsewhere.

This finding suggested that different forces drove access problems for each group. Uninsured individuals cited cost concerns overwhelmingly -- 90 percent named affordability as the primary barrier. Privately insured people also pointed to costs (48 percent), but they additionally mentioned insurance-related hurdles and referral difficulties (28 percent) and issues with the convenience and ease of navigating the system (33 percent).

Policy Implications

Decision makers in both the public and private sectors now confronted the reality that access to necessary medical care was not uniform across the country. This Issue Brief did not attempt to identify the precise causes of these disparities, but the underlying factors were almost certainly complex. Simple, one-size-fits-all policy fixes seemed unlikely to close the gaps between communities.

The health care safety net -- often seen as the primary avenue for improving access among the uninsured -- was not a well-coordinated system in most communities. Instead, it consisted of a loose collection of public hospitals, community health clinics, local health departments, and other individual providers that served the medically indigent in a piecemeal fashion. Building effective interventions to strengthen such a fragmented network presented considerable challenges.

Little was known at the time about how specific elements of the safety net -- such as the availability of preventive and primary care through clinics and physician offices -- affected access for uninsured populations. Ongoing HSC research aimed to help policymakers better understand which health system factors, including managed care penetration and local uninsurance rates, shaped the ability of uninsured individuals to obtain charity care.

Communities also differed in how much priority they assigned to indigent care and in how they organized responses to the uninsured challenge. Some variation reflected local cultural attitudes, while other differences arose from distinctive volunteer efforts and community traditions. These factors often lay beyond the reach of formal public policy.

Providing health coverage to those who lacked it would address many of the barriers the uninsured faced and would produce more uniform access from one community to the next. An HSC simulation projected that the proportion of uninsured people reporting difficulty getting care would drop from 30 percent to roughly 21 percent if they received Medicaid or another form of public coverage, and to about 17 percent with private insurance. These improvements held relatively steady across study sites, regardless of whether a community's uninsured already experienced more or fewer access difficulties.

Geographic disparities in access for the uninsured were expected to continue -- and potentially widen -- as competition among providers intensified and public funding contracted. The already limited federal role in financing care for the medically indigent appeared likely to shrink further, driven by cuts in Medicaid and Medicare Disproportionate Share Hospital Payments. Consequently, policies addressing uninsured populations would remain driven largely at the state and local level. Understanding how local market conditions -- including managed care dynamics and community-level innovations -- affected access was essential to designing strategies that could improve care for the uninsured.

Data Source

This Issue Brief drew on data from the Community Tracking Study's Household Survey, a nationally representative telephone survey of the civilian, non-institutionalized population and 60 randomly chosen communities. In-person interviews with households lacking telephones supplemented the phone survey to ensure proper representation. Fieldwork was conducted between July 1996 and July 1997. The survey gathered observations on nearly 33,000 families and 60,000 individuals, of whom 7,200 were uninsured. The overall family response rate was 65 percent.

Information on health insurance coverage, health status, health care access and utilization, and demographic characteristics was collected for all adults in the household and one randomly selected child. The data were weighted to represent the continental U.S. population and adjusted for survey non-response. Standard errors used in tests of statistical significance accounted for the survey's complex sample design. All differences between estimates cited were statistically significant at the p<0.05 level.

This Issue Brief was adapted from "Ability to Get Medical Care for the Uninsured: How Much Does it Vary Across Communities?" by Peter J. Cunningham and Peter Kemper, published in the Journal of the American Medical Association, Vol. 280, No. 10 (September 9, 1998). A companion article, "How Well Do Communities Perform on Access to Care for the Uninsured?" by Peter J. Cunningham and Heidi Whitmore, appeared in a fall issue of HSC's Research Report series.

Sources and Further Reading

Kaiser Family Foundation — Employer Health Benefits Survey — Annual data on employer-sponsored health insurance.

CMS — Health Insurance Marketplace — Federal marketplace information.

Health Affairs — Peer-reviewed health policy research.

Robert Wood Johnson Foundation — Health policy research and programs.

Commonwealth Fund — Research on health care coverage.