This Article Abstract Reprint (PDF) Submit a response to this article Alert me when this article is cited Alert me when Comments are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to Citation Manager Reprints & Permissions Citing Articles Citing Articles via ISI Web of Science (5) Citing Articles via Google Scholar Google Scholar Articles by Sparer, M. S. Articles by Gray, B. H. Search for Related Content PubMed PubMed Citation Articles by Sparer, M. S. Articles by Gray, B. H. Related Collections Access To Care Managed Care - Medicaid Medicaid

UpDate

SPECIAL REPORT

Promising Practices: How Leading Safety-Net Plans Are Managing The Care Of Medicaid Clients

Abstract

 
Health plans formed by safety-net providers serve large numbers of Medicaid beneficiaries. Through a series of case studies, we examined the care management tools used by leading safety-net plans. These plans do not rely on the coercive, command-style tools of managed care. They rely instead on tools that emphasize partnership with providers: sharing data about practice patterns, using provider profiles and financial bonuses to encourage particular practice patterns, and developing disease management programs that encourage patient compliance with treatment decisions that the plans make little effort to shape. The evidence suggests that these are promising practices but that even these leaders still have a long way to go.

This paper reports results of our second empirical study of safety-net health plans.³ We identified eighty-three of these plans around the nation, with a median Medicaid enrollment of just over 30,000. This paper is based primarily on interviews with plan officials and providers at six plans that we visited during 2000, plus telephone interviews conducted in 2001 with officials at four additional plans, supplemented with survey results we obtained from fifty-six plans in 2000.

We used two main criteria to select the six plans for our site visits. First, we sought plans with high scores on quality-oversight evaluations. We hoped to visit and learn from these plans about good care-management practices. Second, we wanted to examine plans that represented a range of sponsoring organizations. Following these guidelines, we chose three hospital-sponsored plans (Mercy Care Plan in Phoenix, Health Partners in Philadelphia, and Care Source in Dayton); two health center–sponsored plans (Neighborhood Health Plan of Rhode Island and Care Oregon); and one plan established by a consortium that included hospitals, health centers, and a health department (the Bronx Health Plan).

During the site visits we focused on several practices (or tools) of managed care: efforts to evaluate (or profile) the performance of providers and plans, use of financial incentives, prior-authorization requirements, use of practice guidelines, assessments of baseline health status, and disease/case management programs.

Gathering Information: Profiling The Physician Network

Top Gathering Information: Profiling... Managing Care: Establishing... Shifting Power Toward Providers The Unrealized Promise Of... Evaluating Health Plans'... Adding Up The Care... NOTES

 
Having accurate information about enrollee services is an essential starting point for plan managers’ efforts to improve quality, reduce costs, or both. Plans construct provider profiles to learn which providers order or deliver how much of what type of care for which patients; to compare providers with each other; and to assess their performance against various norms in the professional literature. Sometimes the goal is efficiency, sometimes improvement of quality, sometimes both at once. One provider (or clinic) may provide enrolled infants and toddlers with all appropriate immunizations, while another may not. Or there could be wide variation in tests ordered, drugs prescribed, or referrals authorized.

Despite the undisputed importance of profiling, however, most officials we met reported frustration with the limited utility of their provider profiles. One problem is that when providers are paid by capitation (or salary), they do not submit claim forms to obtain payment, and they have little incentive to submit so-called shadow claim forms that can be used to generate utilization data. Overburdened and understaffed safety-net clinics often do a particularly poor job of supplying these needed encounter data. Partly in an effort to improve these data submissions, some health plans are returning to fee-for-service (FFS) payment. A second approach is to pay providers a small fee for the shadow-claim submissions. There also are efforts to ease the administrative burden of the submission process, both by encouraging electronic (rather than paper) submissions and by allowing providers to submit standard forms required for other government programs.

Linking services to individual providers. Even when safety-net clinics submit encounter data, however, it is often difficult to link services to individual providers. Enrollees formally assigned to a specific primary care provider often receive care from another provider without the plan’s knowledge. This is especially common in academic medical centers (AMCs), where residents provide much of the care (supposedly under the tutelage of an attending physician). Preparing individual provider profiles is also expensive. Mercy Care is the only safety-net plan we saw that uses an individual profiling system similar to that used by many commercial HMOs. The plan pays around $200,000 annually to a vendor that produces quarterly reports on all primary care physicians in the network who have at least twenty Mercy enrollees as patients. The reports compare use by provider, on a wide range of variables, adjusted by case-mix (age, sex, and disease).

As a result, the majority of health plans with which we spoke rely on site-based profiles, having concluded that drilling down to the individual doctor is too expensive or too difficult, or both. But the utility of these more modest efforts is also challenged, however, especially by sites with questionable profiles that insist that their patient population is sicker or less compliant than average and that the profile failed to adjust properly for this "fact." Plan officials concede that the science of risk adjustment still has far to go but insist that the complaints are exaggerated.

Prevalence of profile feedback. Despite these obstacles, health plans managers are proceeding with efforts to develop and use provider profiles. According to our survey, more than 60 percent of safety-net plans provide some form of profile feedback to their physician network.⁴ The site visits confirmed the prevalence of the practice. The site visits also suggest that the profiles are most often used as a conversation starter with individual providers (or the medical director or a clinic), to point out questionable practice patterns.

Some plans also use profiles to create a sense of partnership between the provider and the health plan. Mercy Care sends provider-relations representatives to primary care physicians’ offices three times a year to review the profile results, offer suggestions to poor performers, and identify best practices from the more successful. The plan also hosts dinners for network physicians at which aggregate data about plan performance are presented and strategies for improvement are discussed. To encourage attendance, the health plan pays physicians $100 to attend. Plan officials are pleased with the results: The dinners are popular, the medical director and staff get to know large segments of the provider network, plan officials get feedback from providers about health plan practices, and the plan is able to convey useful information.

Changing physician behavior. It is difficult, however, for Mercy or any other plan to move from informational dinners (or, more commonly, newsletters) to focused interventions that change providers’ behavior. One problem is the changing composition of provider networks. As a safety-net plan increases its membership, it often adds office-based physicians to a network previously composed of hospital-based clinics and CHCs. Health plan managers therefore must seek to improve care in two very different delivery systems: traditional safety-net clinics and mainstream office-based practices. Health plan officials told us that they find it harder to influence office-based physicians with relatively few plan enrollees than clinic-based providers.

Another hurdle is uncertainty and disagreement over the value of financial incentives. Should health plans reward or punish clinics based on their performance on profiles? Some plan managers argue that incentives can encourage good outcomes, prompting physicians and clinic administrators to press for improved performance. Others believe that financial incentives cause dissension in the provider network by creating winners and losers based on data whose accuracy is disputed. Moreover, some providers’ poor performance could be improved with additional resources (a new computer system) but would decline if resources were withdrawn because of penalties.

Rewarding good outcomes. According to our survey, most safety-net health plans do not link physician compensation to patient satisfaction, financial targets, or information adequacy.⁵ The site-visit health plans, however, are pushing forward with some pilot initiatives, distinguishing themselves from most of their counterparts. The strategy is to communicate clearly with providers about the terms of the incentive effort and to reward good outcomes rather than to punish poor performers. The most generous incentive program we encountered is at Neighborhood Health Plan of Rhode Island (NHP-RI), where participating health centers can earn up to $4 per member per month for each of several quality benchmarks: expanded office hours, electronically submitted encounter data, improved office procedures, and a completed application for Joint Commission on Accreditation of Health-care Organizations (JACHO) accreditation. Care Oregon had something similar, the Quality Bonus Program, in which part of any year-end surplus is distributed to primary care clinics based on performance on quality indicators. A couple of years ago the plan was distributing roughly $3 million per year.

The NHP-RI and Care Oregon programs both sought to influence the decisionmakers at clinics (such as medical directors); staff physicians generally are on a fixed salary, and their compensation is not directly affected by the incentives. In contrast, a few years ago Mercy Care Plan implemented a Physician Recognition Program, under which the plan distributed roughly $500,000 among primary care physicians who scored high on several quality measures, had no member complaints, and showed up at plan-sponsored meetings. The program seemed promising because the provider network had relatively few salaried clinic doctors and a large number of solo (or small- group) practitioners.

Whether these programs actually change practice patterns is not clear to plan leaders. Most are cautiously optimistic, although they lack tangible evidence that demonstrates an effect. Physicians who perform well may have done so anyway, and those with many managed care contracts seem unlikely to change behavior because of incentives offered by a single plan. Health services research does little to resolve the debate: The literature on the impact of financial incentives on physician behavior is rather thin.⁶ Regardless of their true impact on quality, plan managers believe that the programs are popular with providers (especially if there are only winners and no losers). Even so, most plan officials consider them a luxury that is feasible only in good financial times. Both Care Oregon and Mercy Care Plan suspended their programs when a year-end surplus from which they could be funded was no longer available.

Managing Care: Establishing Disease Management Programs

Top Gathering Information: Profiling... Managing Care: Establishing... Shifting Power Toward Providers The Unrealized Promise Of... Evaluating Health Plans'... Adding Up The Care... NOTES

 
Nearly every safety-net plan has a brochure that trumpets its latest disease management initiative. Roughly 90 percent of these plans report that they have a disease management initiative in place: 80 percent have a program for children with asthma, 75 percent have an initiative covering high-risk pregnant women, and 60 percent, a program for members with diabetes.⁷ Plans cite these initiatives as evidence that Medicaid beneficiaries can fare better in managed care than they did in the old FFS environment, which rarely gave high-risk beneficiaries coordinated and continuous care.

The health plan officials we met concede, however, that these disease management initiatives tend to be modest. They do not involve hiring a cadre of nurse care managers to monitor physicians’ interventions. Instead, the plans try to identify enrollees with certain chronic conditions, mail them educational materials about their condition, and encourage them to seek appropriate care (perhaps with reminder postcards or coupons that entitle compliant patients to small rewards). Some plans also send physicians state-of-the-art practice guidelines in the mail that summarize the latest research in the field and the best practices for treatment, but few appear to inquire closely whether these are followed.

In these disease management programs the task of actually managing the patient’s care remains with the patient’s regular care providers. The goal is to increase compliance with physician treatment recommendations, not to supplant the physician’s discretion. These providers are assumed to have more leverage with patients than does the distant (and perhaps mistrusted) HMO and are better able to teach and promote healthy behavior. As a result, despite lofty rhetoric about innovative disease management efforts, none of the plans we visited have disease management departments that contain more than one or two nurses, and relatively few clients benefit from active care management.

For instance, Care Oregon uses claims data to identify enrollees with diabetes, sends them educational materials, and encourages them to seek regular care. Physicians manage the actual care as part of their regular responsibilities to the patient. Mercy Health Plan, Bronx Health Plan, and NHP-RI all report similar approaches. NHP-RI, for example, sends instructional mailings to roughly 2,500 enrollees with asthma and also sends identifying information to their primary care physicians. The plan also has one nurse who works with asthma patients, but her caseload is relatively small (fewer than 100) and is limited to those patients who either are noncompliant or need special attention. Health Partners, still more intent on delegating disease management to its provider networks, awarded $50,000 grants in 2000 to several of its large provider networks to encourage them to expand their disease and care management departments.

Plans’ reluctance to push aggressively an expansive disease management agenda of their own is plain even where state law mandates the initiative. Ohio, for example, recently required health plans in the Medicaid market to have a disease management program for asthmatic children. As of November 2000 Care Source had enrolled about 40 percent of eligible youngsters in its asthma program and hoped to add the others over the next year. The program aims to persuade targeted members to follow the physician’s treatment protocol. The job of health plan case managers is to help clients overcome barriers between them and the physician’s office and to supplement the physician’s treatment protocol, not to promote or impose external practice guidelines. Families that are hungry or lack a car are less likely to follow an ambitious treatment plan, so case managers might get them into the food stamp program or arrange transportation to medical appointments. Plan leadership tends to steer clear of clinical terrain.

Shifting Power Toward Providers

Top Gathering Information: Profiling... Managing Care: Establishing... Shifting Power Toward Providers The Unrealized Promise Of... Evaluating Health Plans'... Adding Up The Care... NOTES

 
The rise of managed care fundamentally changed the relationship between providers and insurers. The requirement that physicians receive permission for referrals to specialists, hospital admissions, surgical procedures, and the like prompted howls of protest from providers about the alleged evils of 800 numbers, denials of needed care, a plethora of forms and rules, and ruthless cost cutting. Obliging physicians to treat patients according to utilization review protocols based on practice guidelines was said to risk turning the art of medicine into cookbook formulas.

Managed care officials saw the reduction in physician autonomy and increased oversight of practice patterns as an overdue antidote for a medical system gone awry. Extraordinary variation in the practice of medicine within and between communities and clinical populations and evidence of overuse of services was seen as evidence of why plans need to find ways to improve practice patterns.

Whatever the validity of this rationale for managed care, the health plan officials with whom we met consistently voiced disappointment with the results of these traditional medical management tools, as did the large majority of plans that responded to the survey. Indeed, only 25 percent of respondents reported that the savings generated by specialty referral requirements exceeded the cost.⁸

Despite their dissatisfaction, however, the plan officials are reluctant to abandon prior authorization of specialty care (much less of hospital care). Their preference is to instruct patients to ask for referrals from primary care providers but to eliminate the requirement that doctors put these referrals in writing and receive plan approval. The goal is to return referral authority to the primary care physician, without empowering consumers to bypass general practitioners on their way to specialists. Several officials offered us the analogy to how empty police cars on the side of the highway deter speeders; likewise, the referral requirement (even when loosely enforced) supposedly deters inappropriate use of specialty care.

The Unrealized Promise Of The Baseline Assessment

Top Gathering Information: Profiling... Managing Care: Establishing... Shifting Power Toward Providers The Unrealized Promise Of... Evaluating Health Plans'... Adding Up The Care... NOTES

 
Nobody disputes that it is a good idea for plans to assess the health status of new enrollees. An initial assessment can take the form of questions, posed via telephone or written questionnaire, designed to encourage the self-reporting of illness or risk status. Plans also generally encourage a baseline physical performed by a primary care physician. The benefits are obvious: Enrollees with a chronic illness can be referred to disease management programs, pregnant women can be encouraged to obtain prenatal care, and children who need preventive services can be identified.

Many states require health plans to conduct initial assessments, while others delegate the assessment function to enrollment brokers or county welfare workers. Roughly 75 percent of safety-net health plans make an effort to conduct such an assessment—and 86 percent of those plans do so at least in part because it is a regulatory requirement.⁹

Despite the popularity of this practice, however, the plan officials with whom we spoke have doubts about its value. These officials reported a response rate of less than 30 percent from new enrollees. Many Medicaid beneficiaries do not have a telephone. Others no longer live at their listed address, are deterred from responding by language barriers, or are unwilling to respond to a seemingly burdensome request. This low response is consistent with the results reported by health plans that develop programs to educate new Medicaid enrollees about the workings of managed care. Generally unwilling to sift through enrollment materials, attend new member orientation sessions, or read disease management materials, beneficiaries may wait until sickness strikes before contacting the plan or seeking medical care.

Fearful that the cost exceeded the benefit, Care Oregon has decided against an initial assessment program. Mercy Care Plan began such a program but soon scaled back the effort. And while the other plans we visited have early assessment programs, usually because government regulators require one, they make little effort to contact beneficiaries who do not respond quickly.

Evaluating Health Plans’ Performance

Top Gathering Information: Profiling... Managing Care: Establishing... Shifting Power Toward Providers The Unrealized Promise Of... Evaluating Health Plans'... Adding Up The Care... NOTES

 
The National Committee for Quality Assurance (NCQA) has developed the most widely used performance standards for managed care organizations. The NCQA also serves as an accrediting organization, evaluating health plans’ quality improvement processes, the adequacy of their provider networks, preventive health activities, medical records, and policies and practices on members’ rights and responsibilities. States generally do not require NCQA accreditation for HMOs, but this indicator of plan quality has been valuable to plans in the commercial market. Large employers often rely on the NCQA seal of approval as evidence that a plan provides good-quality care and has systems in place to track and document care.¹⁰ Many employers include only accredited plans in their benefit programs.

For health plans that concentrate on the Medicaid market, however, the benefits of NCQA accreditation are less obvious. States generally do not restrict participation to accredited plans, and beneficiaries are not thought likely to use accreditation in choosing a plan. Most plan officials emphasize the deterrents to accreditation, beginning with the steep cost. One plan estimates that it would cost at least $100,000 to apply ($50,000 for the application itself, $30,000 to cover the cost of the audits, and $20,000 for miscellaneous expenses). The process is also labor-intensive. NHP-MA, which was planning to seek accreditation at the time of our interview, was sending fifty staffers on a two-day retreat just to plan its application. These costs are particularly distasteful to plans that also must comply with different and rigorous state-administered quality review processes.

Safety-net plan officials also complain that the accreditation process is especially difficult for a plan in the Medicaid market. For example, the rapid turnover of Medicaid enrollees hampers provision and thus documentation of preventive services over time. Similarly, reviewers may focus on services (such as mammogram screening) that are regularly delivered to adult commercial-plan enrollees but rarely to the children that Medicaid plans enroll in large numbers.

As a result, few safety-net plans have voluntarily sought NCQA accreditation. Of the plans that have done so, many have not repeated the process because they believe that most of the benefits were gained the first time out.¹¹ Mercy Care, for example, had NCQA accreditation from 1994 to 1998 but chose not to seek renewal; Arizona has its own health plan oversight system, there seemed to be little competitive advantage in the designation, and the NCQA review teams seemed unsympathetic to the plan’s needs.

The NCQA might well have lost even more relevance to the Medicaid market except that a handful of states have enacted legislation requiring all health plans to seek accreditation and to obtain it within a specified time. Massachusetts, for example, included in its 2000 Patient Protection Act just such a requirement, which prompted the NCQA planning retreat at NHP-MA.

To be sure, even plans that doubt the economic value of NCQA accreditation concede that the application process has benefits. Although Mercy Care abandoned its formal accreditation effort in 1998, it continues to conduct an annual internal mock NCQA process, bringing in members of the plan’s board of directors to play "outside reviewers." NHP-RI officials acknowledge that their NCQA accreditation was spurred by a state mandate, but they reported that they have become far more sophisticated as a result of the application effort and have invested in quality improvement efforts that might otherwise have gone unrealized.

Adding Up The Care Management Scorecard

Top Gathering Information: Profiling... Managing Care: Establishing... Shifting Power Toward Providers The Unrealized Promise Of... Evaluating Health Plans'... Adding Up The Care... NOTES

 
Safety-net health plans do not fit the stereotype of managed care organizations seeking to change behavior and practice patterns with coercive and command-style tools of managed care.¹² Plan executives think that it is ineffective and counterproductive to try to impose their judgments on medical decisions, so they generally do not try. Few think it worth battling about practice guidelines with a provider community that values autonomy and case-by-case treatment decisions. Nor is it cost-effective to second-guess provider referral patterns—so long as primary care providers stick to specialists in the provider network. Health plans are therefore holding their breath and loosening the reins on some techniques with which to manage outpatient use. Health plan nurses still review inpatient admissions and must still authorize certain kinds of elective surgery, but, at least outside the hospital, the trend is clearly toward returning to the provider the discretion and the autonomy that managed care seemed to threaten.

Safety-net plans increasingly pin their hopes on managed care tools that emphasize partnership with providers— most notably, sharing data about practice patterns and using provider profiles and financial bonuses to encourage particular practice patterns. This trend also is apparent in the new generation of disease management programs that encourage patients’ compliance with providers’ treatment decisions that the plans make little effort to shape.

One plausible explanation for the partnership model is that provider-sponsored plans would be expected to be friendly to their provider sponsors. Prior research, however, found considerable tension between these health plans and their provider sponsors, especially when the plan looks to expand the provider network.¹³ Moreover, many safety-net plans report that they tried to implement (albeit unsuccessfully) some of the command-and-control tools of managed care that now seem to be out of favor.

A more persuasive explanation is that these plans are concluding that a partnership model is the best way to influence the practice of medicine. Provider profiles, disease management programs, and practice guidelines are all promising practices, but they work best if the goal is to supplement, not redesign, physicians’ practice patterns and if the data are sufficiently accurate to persuade doctors to listen. As health plans improve their medical information systems and develop quality management programs, the likelihood that managed care will improve care for Medicaid beneficiaries should increase.

Safety-net health plans continue to operate in an unsettled market. In this turbulent environment the challenge for safety-net plans is to demonstrate that they offer the best system for providing beneficiaries with good care at affordable prices. Our evidence suggests that some of these plans are implementing promising practices but that even these leaders still have a long way to go.

Editor's Notes

 
Michael Sparer is an associate professor and Larry Brown is a professor at the Mailman School of Public Health, Columbia University. Michael Gusmano is with the International Longevity Center–USA. Cathy Rowe is with the Division of Health and Science Policy at the New York Academy of Medicine, and Brad Gray is director of that division.

Support for this study was provided by the Bureau of Primary Care, Health Resources and Services Administration, and by the Center for Health Care Strategies.

NOTES

Top Gathering Information: Profiling... Managing Care: Establishing... Shifting Power Toward Providers The Unrealized Promise Of... Evaluating Health Plans'... Adding Up The Care... NOTES

 

1. Kaiser Commission on Medicaid and the Uninsured, "Fact Sheet on Medicaid and Managed Care" (Washington: Kaiser Commission, December 2001).
2. The commercial plans were discouraged by lower-than-expected payment rates, onerous regulatory requirements, unexpected challenges in serving the Medicaid population, and deep financial losses in their core commercial business. T. Coughlin et al., "Commercial Health Plan Participation in Medicaid Managed Care: An Examination of Six Markets," Inquiry (Spring 2001): 22–34.
3. The first study produced two articles. B.H. Gray and C. Rowe, "Safety-Net Health Plans: A Status Report," Health Affairs (Jan/Feb 2000): 185–193; and M.S. Sparer and L.D. Brown, "Uneasy Alliances: Managed Care Plans Formed by Safety-Net Providers," Health Affairs (July/Aug 2000): 23–35.
4. M. Gusmano et al., "The Evolving Role and Care Management Approaches of Safety-Net Medicaid Managed Care Plans" (Unpublished paper, New York Academy of Medicine, 2002), Table 6.
5. M. Gusmano et al., Safety-Net Health Plans: Managing Patient Care, Report to the Center for Health Care Strategies and the Health Resources and Services Administration (New York: New York Academy of Medicine, 2001), Table 25.
6. B.S. Armour et al., "The Effect of Explicit Financial Incentives on Physician Behavior," Archives of Internal Medicine 161, no. 10 (1998): 1261–1266.
7. Gusmano et al., "The Evolving Role and Care Management Approaches," Table 7.
8. Ibid., 19.
9. Gusmano et al., Safety-Net Health Plans, Table 24.
10. J.K. Iglehart, "The National Committee for Quality Assurance," New England Journal of Medicine 335, no. 13 (1996): 995–999.[Free Full Text]
11. Gusmano et al., "The Evolving Role and Care Management Approaches," 25.
12. Commercial HMOs seem also to be reducing their reliance on traditional medical management tools (from gatekeepers to preadmission review). C. Lesser and P. Ginsburg, Back to the Future? New Cost and Access Challenges Emerge, Issue Brief no. 35 (Washington: Center for Studying Health System Change, February 2000).
13. Sparer and Brown, "Uneasy Alliances."

                      What's this?