Under global capitation a health plan pays a primary care physician (PCP) risk unit a fixed monthly payment to cover all costs of providing health care to all capitated patients cared for by the risk unit's physicians regardless of the cost of the services they require.1,2 (A PCP risk unit is a group of primary care physicians sharing financial risk in a capitated managed care contract.) To be financially successful under capitation, a PCP risk unit must spend less on care than the amount of its capitation budget. With total capitated medical budgets in various geographic markets ranging from $95 to $130 per member per month for a commercial (under age 65) population, a bone marrow transplant, a complicated coronary artery bypass graft procedure, or the care of a premature infant can stress a risk unit's finances, even with reinsurance protection. Global capitation thus creates an incentive for PCP risk units to attract and retain healthy patients while avoiding the high cost of treating patients with severe illnesses.3–9
Although health plans and preferred provider organizations (PPOs) may include a broad array of specialists in their networks, PCP risk units typically refer patients within a small circle of specialists with whom they have developed ongoing referral relationships to facilitate the coordination of care, maintain quality, and control resource utilization.10,11 Specialists outside this circle—to whom the PCPs in the risk unit rarely refer—have no explicit incentive to be cost-effective in the delivery of care, since all expenses are borne by the PCP risk unit. Specialists at academic medical centers (AMCs), in addition to being outside most community referral circles, also represent greater financial risk, because costs are generally higher at AMCs than in the community.12–14 Some community PCPs also prefer the more direct (at time, face-to-face) communication they have with local specialists.
TRANSFER OF SICK PATIENTS TO ACADEMIC NETWORKS
During three years of capitation experience (1997–1999), we have observed that patients who have catastrophic illnesses often seek out what they perceive to be the highest-quality specialty care, irrespective of the imposed boundaries of health plan networks, preferred provider organizations, or a PCP's referral circle of selected specialists. Some of these patients request referrals to specialists at AMCs. Community PCPs, however, may be reluctant to refer sick patients to specialists at AMCs because of the significant financial risk associated with referring a patient with a severe illness, such as cancer, to an out-of-circle specialist at a high-cost institution. Although a community PCP may be confident that specialty care can be adequately provided within the unit's usual circle of local specialists, some patients demand a broader choice of providers for both secondary and tertiary care.
We have observed that patients' demands for referrals to AMC specialists, coupled with the reluctance of community PCPs to make these referrals, are driving some patients to disenroll from their community PCPs. Moreover, these patients are transferring their care to PCPs affiliated with AMCs to assure access to the AMC-affiliated specialists they seek. Based on our experience in global capitation, we believe the primary care risk units developed by AMCs in response to the market forces associated with managed care15–18 may impose a financial liability on AMCs under capitation. Transfers of sick patients to AMC-affiliated risk units create a financial burden known as “adverse selection” for AMCs participating in global risk capitation.19
The following case example is illustrative:
A 65-year-old central Massachusetts man with prostate cancer requests a second opinion from a Boston AMC urologist. Granted a referral for this consultation by his local community PCP, the patient is told by the Boston urologist that he is a candidate for a “nerve-sparing” total prostatectomy that could be performed at the Boston AMC with authorization by his community-based PCP. Citing the availability of high-quality prostate surgery locally, the patient's long-standing PCP denies the referral to Boston for surgery. Although the Boston urologist is listed within the patient's health plan network of providers, the Boston urologist is outside the PCP's usual referral circle of local specialists. The patient's PCP suggests that if the patient wishes to be referred to the Boston urologist, he should find a new PCP directly affiliated with the Boston urologist's AMC.
Determined to have the “nerve-sparing” procedure in Boston, yet avoid the significant out-of-pocket expense of having the procedure done in Boston without authorization from his PCP, the patient switches to a PCP affiliated with the same AMC as the urologist. The patient's new PCP (in a risk unit affiliated with the Boston AMC) meets her new patient when requested by the patient to grant a referral for total prostatectomy. When questioned about his motivation for obtaining primary care 75 miles away from his home, the patient explains that this is a mechanism by which to access (and have insurance pay for) the specialty surgery he desires. Further, he indicates no intention of maintaining a primary care relationship in Boston, but intends to switch back to his original PCP after the surgery.
HIGH ILLNESS BURDEN OF PATIENTS SWITCHING TO AMC-AFFILIATED PCPS
To estimate the types of illnesses prompting patients to transfer their primary care affiliations, we asked the 100 PCPs affiliated with our AMC (who manage approximately 20,000 full-risk managed care patients) to identify new patients who had transferred their care with the express purpose of accessing care from our specialists. Of 45 cases identified by our PCPs over a one-year period, 53% had transferred for cancer treatment (or cancer-related surgery, such as mastectomy), 27% for other surgical procedures such as coronary artery bypass grafting, 13% for treatment of complicated medical conditions such as cirrhosis, 7% for orthopedic surgery, and 2% for complicated obstetrics.
We also analyzed the medical expenditures incurred for patients in risk contracts who had transferred from community risk units to the risk units of Massachusetts General Hospital and Brigham and Women's Hospital, the two primary AMCs within Partners HealthCare System. These expenses were compared with those of patients who had transferred out of these teaching hospitals' risk units into community risk units within our system. This analysis indicated that for one major risk contract, the mean annual health plan payment for patients who had switched to the teaching hospitals' primary care risk units averaged $3,730 more in annual expenses than the mean annual plan payment of those who had switched out ($5,350 versus $1,605). In a second contract, the average incremental expense per patient was $1,911 ($3,219 versus $1,267). Complete financial data were not available for patients who had transferred to or from community groups who were not part of Partners HealthCare System. However, if the rate of transfers to our teaching hospitals were similar for community groups outside our system, then the net impact would be an expense of $7.29 per member per month for the first contract and $2.47 per member per month for the second contract.
These findings suggest that not only does the net flow of seriously ill patients from community risk groups to academic PCPs create a financial burden for AMCs (already under pressure from the Balanced Budget Amendment of 199720), but also that these transfers disrupt established patient—doctor relationships. By switching to an AMC-affiliated PCP a patient may have gained access to excellent specialty care, but in the process may have sacrificed a long-standing primary care relationship and given up access to primary care close to home.
In our experience there have been four common reasons why a community-based patient will switch to a PCP with the same institutional affiliation as his or her tertiary care specialist:
* Patients sometimes feel that their care, now complicated by catastrophic illness, is best coordinated by physicians within the same institutions as their treating tertiary care specialists.
* Patients have concerns about the quality of care provided by their previous physicians.
* Patients are encouraged (intentionally or unintentionally) by their tertiary care specialists to switch PCPs.
* Patients desire access to new therapies available only through participation in academic clinical research trials.
As illustrated in the example case above, our experience in capitated contracts highlights a fifth scenario. Faced with the administrative and financial burdens of paying for care provided at an academic tertiary care institution,21,22 PCPs from community networks inadvertently or intentionally encourage patients to switch to PCPs affiliated with the desired specialists.
The financial consequences of caring for a population with a disproportionate burden of illness could be ameliorated if capitation payments could be adequately risk-adjusted. Significant efforts have been made to predict resource utilization based on health-status measures, prior utilization, diagnoses, severity-of-illness measures, and other case-mix variables.23–32 However, when applied retrospectively, the best methods predict less than 50% of the observed variance in expenditures; if used prospectively to predict future resource utilization, performance falls significantly.33 As a result, few risk-adjustment methods are in general use today. Typically set to regional or statewide norms with minimal risk-adjustment (except for age and sex), capitation budgets do not attempt to mirror health status.
Options for AMCs
Ironically, AMCs' success in carrying out their mission—to provide high-quality, complex tertiary care—is what makes them vulnerable to adverse selection in a capitated environment. There are, however, a number of alternatives for AMCs to consider in addressing these problems:
1. Eschew primary care network development. Instead, develop only specialty care services. Without their own PCP networks, AMCs will not be subject to the potential deficits associated with global risk capitation. However, AMCs will then be totally dependent on referrals from community-based PCPs. In addition to the financial risks, this strategy also threatens academic teaching and research missions.
2. Limit enrollment of new patients into capitated health care plans by capping enrollment at a target panel size. Although limiting panel size reduces actuarial stability and does not deter adverse selection, it does limit the overall financial vulnerability of the institution.
3. Maintain a primary care network development strategy but refuse to participate in global risk capitation. In a highly penetrated managed care environment where capitation continues to spread, this strategy could constrain an institution's market share.
4. Develop unique contract arrangements with health plans to increase compensation to the AMC-affiliated risk groups through a series of “carve outs.” In this scenario the health plan retains the risk for catastrophic and costly illnesses and spreads this risk across its entire network. Some health maintenance organizations (HMOs) have adopted this strategy with in-vitro fertilization and HIV care. Implementing comprehensive carve-out payment methods would probably be very complex, however.
5. Tertiary centers accept the deficits from primary care global capitation as a “loss leader” and count on other lines of non-risk referral business to cover the differential. This presents obvious long-term financial hazards, particularly in an era of declining reimbursement and growing provider risk-sharing.
6. Establish surrogate PCPs for patients from outside networks receiving care at the AMC. On-site surrogate PCPs would maintain communication with the patient's community PCP and assure efficient resource utilization at the AMC. This solution helps address the community PCP's communication concerns, while leaving the risk with the referring PCP risk unit.
Options for Health Plans
Although it is the AMCs that are vulnerable to adverse selection under capitation, we believe that a comprehensive solution will require collaboration between AMCs, community PCPs, and health plans. Rather than simply allowing market forces to undermine academic centers, health plans have the opportunity to eradicate the perverse incentives limiting access to care for the seriously ill. There are several options for health plans and insurers to contribute to a solution:
1. Require that the expenses that accrue during the first six months after a patient has transferred from one PCP to another be shared equally between the risk units of the two PCPs. Although this requires a mechanism to track patients as they change PCPs, it would discourage PCPs from casting off sick patients to other risk units.
2. Limit patient enrollment to PCPs within a certain geographic range of their home or work sites. Some plans encourage patients to choose PCPs within 30 miles of their homes. Based on the observation that traveling great distances to see a physician may be a proxy for severity of illness,34 this mandate would have the effect of constraining a subset of patients who would otherwise choose PCPs at great distances from their homes in order to access tertiary care.
3. Establish coverage of tertiary and quaternary care from a risk pool common to the health plan's entire population. Individual PCP risk units would not be at risk for the costs of catastrophic illnesses. Instead, a health plan would distribute this risk over a very broad patient base. This solution would blunt the incentive for PCP risk units to shun severely ill patients.
4. Allow patients to change PCPs only during a defined open enrollment period once per year. This rule could be applied to all PCP changes, or to the subset of changes to PCPs in AMC-affiliated risk units. Under this model, academic PCPs would have limited protection from patient transfers prompted by the onset of severe illness.
5. Employ a case-rate, or “contact capitation,” payment method for tertiary care.35 Like fee-for-service, case-rate arrangements target funds to the institution and physicians providing care, yet, like capitation, the payment of a fixed fee for an episode of care removes the incentive for overutilization. Developing case rates can be complex, however, especially for prolonged episodes of care such as cancer therapy.
6. Risk-adjust each risk unit's capitation payment based on the health status and disease burden of its population. Although imprecise, risk-adjustment methods would reallocate funds across a health plan's entire network of enrollees, providing higher budgets to risk units with sick populations and proportionately lower budgets to risk units caring for relatively healthier populations. This solution would require an educational component to reassure community-based risk units whose capitation budgets might decrease.
SUMMARY AND CONCLUSIONS
As stated earlier, AMC-affiliated PCPs are vulnerable to severe financial losses if they participate in capitated contracts yet attract patients who disproportionately require tertiary care. When healthy, patients may choose health insurance products based on price, with little regard for the specialty services that will be available if needed. However, we have observed that seriously ill patients will change PCPs and provider networks rather than constrain their choices of specialists to community-based PCPs' circles of preferred specialists. Our experience in capitated contracts suggests that the incentives of capitation make community-based PCPs hesitant to refer patients to specialists at AMCs, thus driving sick patients to switch from their community-based PCPs to AMC-affiliated PCPs as a means of gaining access to AMC specialists. We would note that there are no data that suggest that patient outcomes have been harmed by community PCPs' efforts to keep care within the community risk pools; however, there has clearly been considerable patient dissatisfaction.
An imbalance in the distribution of sick patients has been documented in other capitated systems. After Tennessee Medicaid converted its entire program to a capitated system (TennCare), Tennessee AMCs had a significantly higher prevalence of patients with high-cost medical conditions that did other types of health care institutions in Tennessee.36 Similarly, an analysis of California Medicare HMOs by the United States General Accounting Office (GAO) found that sick patients were more likely to rapidly disenroll from HMOs than were healthier Medicare patients. In their report the GAO hypothesized that the high disenrollment rate among sick patients “might indicate greater access barriers and less satisfaction with HMOs for such beneficiaries.”37
A clinically appealing strategy to counterbalance the financial effects of adverse selection would be to employ a nimble health-status and severity-of-illness adjustment method that allowed a health plan to appropriately adjust each risk unit's capitation budget. Currently, however, risk adjustment remains a relatively blunt instrument and can represent only a partial solution to adverse selection. Paying for tertiary care using case rates is also appealing, as this strategy blends the beneficial aspects of fee-for-service and capitation payment, eliminating the incentives for both under- and over-utilization by paying a fixed fee for an episode of care.
We believe that in addition to developing risk-adjustment and tertiary case-rate methods, health plans should collaborate with AMCs to implement complementary strategies that will minimize systematic adverse selection and prevent unnecessary disruption of community-based primary care relationships during the course of an illness. Ideally, purchasers of health care, such as large self-insuring employers that define their own benefit plans, should also participate in the development of solutions. This may be difficult, however, as providers typically interact with health plans only and not directly with payers.
Although health plans could allow the financial impact of adverse selection to undermine the financial stability of AMCs, the health plans in our region have traditionally eschewed such tactics, favoring collaboration. Short-term financial gain for health plans from an overly aggressive strategy is likely to be offset by the creation of ill will among teaching hospitals and specialists, souring current and future relationships. For their part, AMCs will need to address their historically high cost structures. As suggested in the Sixth Report of the Council on Graduate Medical Education,38 AMCs must develop the capability to allocate teaching and research costs appropriately to be more competitive in the marketplace.
It is in the best interests of patients, physicians, payers, and health plans to modify the current counterproductive incentives39 that promote adverse selection at AMC-affiliated primary care risk units. To be successful, managed care must have fair capitated risk-sharing payment methods. Adequate budgets commensurate with the illness burdens of the populations being served and collaboration between provider groups and health plans could greatly reduce adverse selection among AMC-affiliated primary care practices.
1. Berwick DM. Payment by capitation and the quality of care. N Engl J Med. 1996;335:1227–31.
2. Swartz K, Brennan TA. Integrated health care, capitated payment, and quality: the role of regulation. Ann Intern Med. 1996;124:442–8.
3. Newhouse JP. Patients at risk: health reform and risk adjustment. Health Aff (Millwood). 1994;13:132–46.
4. Neff JM, Anderson G. Protecting children with chronic illness in a competitive marketplace. JAMA. 1995;274:1866–9.
5. Pearson SD, Sabin JE, Emanuel EJ. Ethical guidelines for physician compensation based on capitation. N Engl J Med. 1998;339:689–93.
6. Kassirer JP. Managed care and the morality of the marketplace. N Engl J Med. 1995;333:50–2.
7. Relman AS. Salaried physicians and economic incentives. N Engl J Med. 1988;319:784.
8. Abrams FR. The doctor with two heads—the patient versus the costs. N Engl J Med. 1993;328:975–6.
9. Woolhandler S, Himmelstein DU. Extreme risk—the new corporate proposition for physicians. N Engl J Med. 1995;333:1706–8.
10. Hillman A. Managing the physician: rules versus incentives. Health Aff (Millwood). 1991;10:138–46.
11. Kerr EA, Mittman BS, Hays RD, Siu AL, Leake B, Brook RH. Managed care and capitation in California: how do physicians at financial risk control their own utilization? Ann Intern Med. 1995;123:500–4.
12. Blumenthal D, Meyer US. The future of the academic medical center under health care reform. N Engl J Med. 1993;329:1812–4.
13. Kassirer JP. Academic medical centers under siege. N Engl J Med. 1994; 331:1370–1.
14. Iglehart JK. Rapid changes for academic medical centers (second of two parts). N Engl J Med. 1995;332:407–11.
15. Clemmer TP, Spuhler VJ, Berwick DM, Nolan TW. Cooperation: the foundation of improvement. Ann Intern Med. 1998;128(12 part 1):1004–9.
16. Nolan TW. Understanding medical systems. Ann Intern Med. 1998; 128:293–8.
17. Jensen G, Morrisey M, Gaffney S, Liston DK. The new dominance of managed care: insurance trends in the 1990s. Health Aff (Millwood). 1997;16:125–36.
18. Blumenthal D, Weissman JS, Griner PF. Academic health centers on the front lines: survival strategies in highly competitive markets. Acad Med. 1999;74:1037–49.
19. Diehr P, Madden CW, Martin DP, et al. Who enrolled in a state program for the uninsured: was there adverse selection? Med Care. 1993;31:1093–105.
20. Iglehart JK. Support for academic medical centers—revisiting the 1997 Balanced Budget Act. N Engl J Med. 1999;341:299–304.
21. Gold M. Financial incentives: current realities and challenges for physicians. J Gen Intern Med. 1999;14(suppl 1):S6–S12.
22. Selker HP. Capitated payment for medical care and the role of the physician (editorial). Ann Intern Med. 1996;124:449–51.
23. Robinson JC, Luft HS, Gardner LB, Morrison EM. A method for risk-adjusting employer contributions to competing health insurance plans. Inquiry. 1991;28:107–16.
24. Wetzell S. Consumer clout. Minn Med. 1996;79:15–9.
25. Iezzoni LI. Dimensions of risk. In: Iezzoni LI (ed). Risk Adjustment for Measuring Health Care Outcomes. Ann Arbor, MI: Health Administration Press, 1994:29–118.
26. Starfield B, Weiner J, Mumford L, Steinwachs D. Ambulatory care groups: a categorization of diagnoses for research and management. Health Serv Res. 1991;26:53–74.
27. Epstein AM, Cumella EJ. Capitation payment: using predictors of medical utilization to adjust rates. Health Care Financ Rev. 1988;10:51–69.
28. Stewart AL, Hays RD, Ware JE. The MOS Short-Form General Health Survey: reliability and validity in a patient population. Med Care. 1988; 26:724–32.
29. Weiner JP, Dobson A, Maxwell SL, Coleman K, Starfield BH, Anderson GF. Risk-adjusted Medicare capitation rates using ambulatory and inpatient diagnoses. Health Care Financ Rev. 1996;17:77–99.
30. Newhouse JP, Manning WG, Keeler EB, Sloss EM. Adjusting capitation rates using objective health measures and prior utilization. Health Care Financ Rev. 1989;10:41–54.
31. Welch HG, Black WC, Fisher ES. Case-mix adjustment: making bad apples look good. JAMA. 1995;273:772–3.
32. Ellis R, Pope G, Lezzoni L, et al. Diagnosis-based risk adjustment for Medicare capitation payments. Health Care Financ Rev. 1996;17:101–28.
33. Fowles JB, Weiner JP, Knutson D, Fowler E, Tucker AM, Ireland M. Taking health status into account when setting capitation rates. JAMA. 1996;276:1316–21.
34. Welch HG, Larson EB, Welch WP. Could distance be a proxy for severity-of-illness? A comparison of hospital costs in distant and local patients. Health Serv Res. 1993;28:441–58.
35. Chesanow N. Can contact capitation conquer costs? Med Econ. 1998;75:110–20.
36. Bailey JE, Van Brunt DL, Mirvis DM, et al. Academic managed care organizations and adverse selection under Medicaid managed care in Tennessee. JAMA. 1999;282:1067–72.
37. Medicare: Fewer and Lower Cost Beneficiaries with Chronic Conditions Enroll in HMOs. United States General Accounting Office report to the chairman, Subcommittee on Health, Committee on Ways and Means, House of Representatives, August 1997. Washington, DC: GPO, 1997.
38. Council on Graduate Medical Education. Sixth Report: Managed Health Care: Implications for the Physician Workforce and Medical Education. Rockville, MD: Department of Health and Human Services, 1995.
39. Light DW. Good managed care needs universal health insurance. Ann Intern Med. 1999;130:686–9.