Correspondence
The Marketplace in Health Care Reform
N Engl J Med 1993; 329:883-885September 16, 1993
- Article
To the Editor:
In the Special Report “The Marketplace in Health Care Reform” (Jan. 14 issue),1 Kronick et al. suggest that managed competition will not work in smaller metropolitan or rural areas that “would require alternative forms of organization and regulation of health care providers to improve quality and economy.” This difficulty could be surmounted easily if insurers defined suitably large regions in their requests for competitive bids from provider organizations. This would leave the details of local arrangements in sparsely populated areas to the provider organizations, but would ensure competitive prices, since the care of people located in the sparsely populated regions would be aggregated and priced in common with the care of those in the more densely populated areas.
A corollary advantage of this approach would be to encourage provider organizations to locate their more expensive staff and facilities, such as those typically associated with tertiary care, more centrally within the region, thereby leading to lower unit costs and higher quality.
The authors' analysis of the minimal region size required, on the basis of requirements for physicians and hospital beds, is a particularly useful one; however, a similar analysis must be made of how large a region must be to have an acceptable actuarial risk. If we look forward to a shift from fee-for-service to capitation, provider organizations must be large enough for this risk to be held to acceptably low levels to avoid unmanageable financial difficulty in the event of adverse experience.
1 ReferencesMelville H. Hodge
TDS Healthcare Systems, San Jose, CA 951341
Kronick R ,Goodman DC ,Wennberg J ,Wagner E . The marketplace in health care reform -- the demographic limitations of managed competition. N Engl J Med 1993;328:148-152
Full Text | Web of Science | Medline
To the Editor:
Managed competition does assume a different character in rural areas. In areas of low population density, competition between provider networks will largely consist of competition between the administrative frameworks of different health maintenance organizations (HMOs) that have negotiated contracts with the same hospital and many of the same physicians. The accountable health plans will not be totally independent. The full impact of competition among providers may be compromised.
However, the purchasing side will also have a different character in rural as compared with urban areas under managed competition. Urban areas may have many large purchasers of health care in the form of large employers and multiple health insurance purchasing cooperatives. Rural areas may have one such cooperative as the purchaser of health care. Rural states may approximate systems with a single payer, a payer whose power to demand and monitor health care of good value could compensate for the compromised degree of competition among providers.
A. Collier Smyth, M.D.
New Hampshire Oncology-Hematology Professional Association, Hooksett, NH 03106To the Editor:
Kronick et al. conclude that managed competition cannot be extended to small metropolitan and rural areas. In forming this conclusion, the authors use the geographic boundaries defined by the Office of Management and Budget and the U.S. Bureau of the Census1.
We agree that providers must be divided into competing economic units defined by some geographic boundary. However, the size of the health care market may be defined more broadly to include many surrounding small areas. When this broader definition is used, demographic factors will not limit the full implementation of managed competition. Instead, the limiting factor now is whether competing health plans will extend their current areas of network coverage from existing larger metropolitan areas to include small areas. The answer is yes, if the proper financial incentives are used under managed competition.
For instance, the federal government could offer financial assistance to encourage existing health plans to extend their coverage areas to the new geographic boundaries established under managed competition. Authorized grants of up to $1 million could be offered for initial network-development costs1. Another approach could be to implement a surcharge -- say, on 5 percent of all medical bills -- affecting health plans that fail to expand their networks to all areas within the geographic boundary of their health care market2.
A qualified health plan must offer the services provided by primary care physicians, specialists, acute care hospitals, and tertiary care hospitals. We here define one way that health plans could expand their networks to small areas. To provide full coverage within an established geographic boundary, health plans must target a minimal number of enrollees in each small area inside that geographic boundary. Kronick et al. show that 2000 persons are required to support one primary care physician. This is not a large number; consequently, health plans should be able to establish extensive independent primary care networks. For most populations, 10 percent of the people spend about 70 percent of the health care dollars3. This finding implies that 90 percent of the people in small areas can be cared for through a primary care network. The remaining 10 percent will require more resource-intensive care by specialists and hospitals. For these patients, the health plan will need to develop programs by which specialist physicians can travel4. This group of patients will also require acute care hospital services3. These services can be delivered by building mini-hospitals (about 45 beds). Essentially, the mini-hospitals would be expanded surgical centers and would provide mainly maternity services and some general surgical services. Patients requiring the services of a tertiary care hospital can be treated through a gatekeeper program providing coordinated care. Physicians in the primary care network will closely monitor a patient's natural disease process, develop treatment plans, and coordinate all care delivery.
4 ReferencesDouglas G. Cave, Ph.D., M.P.M.
Hewitt Associates, Newport Beach, CA 92660-29351
Dorsey JL . The Health Maintenance Organization Act of 1973 (P.L. 93-222) and prepaid group practice plans. Med Care 1975;13:1-9
CrossRef | Web of Science | Medline2
HMOs facing higher charges until signing Medicaid pacts. Press release of the Capital District Business Review, Albany, N.Y., July 20, 1992.
3
Cave DG ,Tucker LJ . Preventing hospital readmissions -- a new direction for case management programs. Compens Benefits Manage 1991;7:21-274
No quick fix: Americans are looking to Canada for ideas about reforming health care -- but Canadians are facing problems too. Press release of Washington Post Foreign Service, Montreal, July 23, 1991.
To the Editor:
The article by Kronick et al. on the demographic limitations of managed competition relied on what is probably a substantial underestimate of the need for primary care practitioners (one per 2000 enrollees). Other Western medical systems that have achieved some of the goals set for our health care reform, such as the systems in Canada, Germany, Sweden, and the United Kingdom, all use approximately one generalist per 1000 enrollees.
An examination of the physicians employed in large staff-model HMOs (one of which provided the estimate for the analysis by Kronick et al.) shows low staffing ratios in comparison with the rest of the United States and other Western nations. Although the relative proportion of generalist physicians is much higher in the HMOs (accounting for some 42 to 62 percent of staff physicians) and therefore more comparable to the proportions seen in other Western nations, the overall staffing of physicians per 1000 people is remarkably low. However, it is important to avoid extrapolation from the case of HMOs to society as a whole, as these authors did, because in the HMOs the relatively low staffing is probably possible only because of successful selective marketing to younger, relatively healthy people, who simply require less medical care than the population as a whole.
Realizing the improved access, cost containment, and quality enhancement promised under managed competition depends on a well-developed infrastructure to support primary care. I submit that the demographic limitations outlined by Kronick et al. will be markedly compounded by staffing shortages and distribution limitations.
Gregg S. Meyer, M.D., M.Sc.
Massachusetts General Hospital, Boston, MA 02114To the Editor:
Kronick et al. make a number of assumptions about managed-care organizations that relate to traditional staff-model or group-model HMOs.
In the late 1980s, we began successful experiments that showed similar organizational performance with new, different models of provider relationships. Although it is rooted in the staff model, Group Health Cooperative has branched out into smaller communities with similar expectations and principles regarding performance, but it has contracts with community providers of primary care and specialty care (as opposed to full-time, salaried physicians) and with hospitals.
The costs and other performance measures of the delivery system outside the major urban center appear similar to those achieved with the staff model. At the same time, our competitors in the Puget Sound area have also branched out into smaller communities. Information on expected cost and use of services can be computed accurately for small populations (less than 25,000), except for occasional outliers. From this information, various contracting arrangements can be made and the outcomes measured and compared with those of our staff model or other benchmark expectations of performance.
This preliminary positive experience suggests that managed care and perhaps managed competition may be expandable to serve a broader population than suggested in the article. In any event, there is an urgent need for demonstrations of managed-care competition involving urban and nonurban settings.
Phil Nudelman, Ph.D.
Al Truscott, M.D.
Ed Wagner, M.D., M.P.H.
Group Health Cooperative of Puget Sound, Seattle, WA 98121Author/Editor Response
The authors reply:
To the Editor: The letters from Hodge, Cave, and Nudelman and colleagues each suggest that health plans based in urban areas can and should be encouraged to expand their service areas to include more sparsely populated rural areas. We agree that providers in rural areas can and should be linked with such organizations and that links with health plans based in densely populated urban areas are often desirable.
In sparsely populated areas, however, competition among plans will not create a dynamic in which the supply of health care resources is likely to match the health care needs of the population. Consider a community of about 100,000 persons, which according to our estimates would need three cardiologists and one 120-bed hospital. There is no reason to expect that vigorous competition among health plans will result in this desired allocation. There is simply no mechanism available by which competing organizations can cooperate in creating and implementing a plan of resource allocation. Some form of health planning for the public sector is needed. In contrast, in a city 10 times larger that needed 30 cardiologists and 1200 hospital beds, competition among plans might well result in such an allocation: a health plan that used many more resources would be forced to have higher prices than its competitors and would lose members; a plan with too few resources available would presumably produce lower-quality care and lose enrollment as dissatisfaction grew among consumers and providers.
We agree with Dr. Smyth that concentrated purchasing power can potentially facilitate a health planning process in which health care resources and needs are well matched and in which health care providers have incentives and opportunities to use the available resources relatively well. However, health planning efforts for the public sector are often heavily influenced by the interests of providers, rather than patients, and single-payer systems that perpetuate fee-for-service payment may do little to create either the opportunities or the incentives for providers to use the available resources efficiently. Fortunately, in rural areas and smaller cities with fewer providers and, typically, a stronger sense of community, public-sector health planning is likely to produce better outcomes than it will in densely populated, heterogeneous urban areas.
Richard Kronick, Ph.D.
University of California, San Diego, La Jolla, CA 92093-0622David C. Goodman, M.D.
John Wennberg, M.D.
Dartmouth Medical School, Hanover, NH 03755
