Health Plans and Purchasers:
May 1997 - Number Two
By Julianna S. Gonen, PhD
By 1995, approximately three quarters of Americans with employer-based health insurance were enrolled in some form of managed health care.1 Managed care has initiated fundamental changes in the way consumers receive care and providers give care, and as a result has been criticized on general and specific bases, from elevating the wrong stakeholders to primacy (stockholders and business managers rather than physicians and patients) to encouraging practices deemed counter to traditional medical ethics (shortened hospital stays, reduced physician autonomy in clinical decision making). But few would argue that the pace of rising health care costs, due in part to overutilization and unnecessary care, warranted maintenance of the status quo. It is therefore necessary to take a critical yet balanced view of the changes brought about by managed care, particularly in regard to how structural features of managed care plans affect the delivery of primary care to women.
Managed care organizations (MCOs) can be grouped broadly into health maintenance organizations (HMOs) and preferred provider organizations (PPOs). There have been four traditional HMO models: group, staff, network, and independent practice association (IPA), although mixed model plans are now the most common. Changes within managed care make it no longer a simple matter to classify managed care plans according to these types, as plans increasingly defy traditional forms, often as a result of mergers or diversification into multiple lines of business. It is therefore more instructional to focus on the methods MCOs use to manage care and costs. These include 1) defined provider networks, with various provider compensation and gatekeeping mechanisms, which enable the use of practice guidelines and provider education initiatives; 2) information systems which facilitate prevention efforts and performance monitoring; and 3) benefit package designs and coverage options, including carve-outs and pharmaceutical formularies. Each is explored here to assess its impact on women's primary care.
HMOs assume a number of configurations in their relationships with their providers. Pure staff model HMOs, of which few remain, employ physicians directly on a salaried basis, and these physicians see only enrollees of the HMO. In the group model, the health plan contracts with one large multi-specialty group, and physicians in this group might have non-HMO patients as well. In a network model, the plan has such contracts with more than one group. The dominant model now is the independent practice association (IPA), in which the health plan contracts with IPAs, physician entities formed for such contracting purposes, as well as with solo practitioners. Physicians who are part of IPAs, or who contract directly with a health plan retain their own offices; this accounts for much of the growth of this arrangement, as it requires significantly less capital and investment in infrastructure by the health plan.
Less recognizable, but more prevalent, than HMOs are PPOs. One fundamental difference is that PPOs themselves are not insurers, whereas HMOs are insurer and provider in one. PPOs are entities (either loosely formed networks or actual organizations of providers) through which self-insured employers or insurance companies contract to purchase health care services for beneficiaries from a defined group of providers.3 Such arrangements function like a POS option in that care outside the network is covered (albeit at a reduced rate), but PPOs generally do not employ primary care gatekeepers, and enrollees are therefore free to self-refer to any provider. In this sense PPOs lack much of the care coordination that is found within HMOs.
Managed care plans use various provider compensation mechanisms to control their costs. One is capitation, whereby a physician (usually the primary care physician) or a physician group is paid a flat rate per enrollee per month regardless of the individual's use of services and is then responsible for providing all necessary care to that enrollee. This passes on some financial risk to the plan's providers and renders health care costs more predictable for the plan. Plans also negotiate fee discounts with providers, a common mechanism used in PPOs. Because primary care physicians' financial compensation is sometimes contingent upon reducing referrals to specialists, concerns have been raised over possible detrimental effects on patient care.
Provider Roles. An emphasis on primary care has long been a hallmark of managed care. Requiring care coordination by these providers (physicians, or sometimes advanced practice nurses or other providers) encourages preventive care and discourages seeking more costly specialist care by making that care contingent on the approval of the PCP. With the assumption of such a large degree of responsibility for overall care by PCPs, attention must be focused on who is providing primary care for women.
A 1993 Commonwealth Fund survey comparing women in fee-for-service and managed care plans found little difference in the distribution of women by type of PCP.4 Several physician specialties deliver primary care, and for women the picture is further complicated by reproductive health care and its relationship to overall primary care. As explained by Bartman (1996):
Managed care plans have debated whether to designate ob-gyns as PCPs or as specialists, and there is indeed no consensus among ob-gyns themselves on this issue.6 Many plans have developed compromises, whereby ob-gyns are considered specialists but women can self-refer to them for certain routine gynecologic care. According to the American Association of Health Plans (AAHP), at least 81% of HMOs either allow ob-gyns to be designated as PCPs or allow women to self-refer to them. The issue has been of enough consumer salience to inspire a wave of state legislative efforts to mandate these options.
Use of non-physician providers in primary care roles also varies among managed care plans. Approximately 70% of plans in a 1994 GHAA survey used advanced practice nurses and physician assistants as PCPs, about half including them on provider lists from which enrollees could select a personal clinician.7 A GHAA/Kaiser Family Foundation study, also from 1994, found that while over half of responding staff model HMOs allowed members to select a nurse practitioner as a primary care provider, the numbers fell to 32%, 19%, and 3% respectively for group, network, and IPA model plans. An in-depth study of 23 HMOs by Mathematica that same year found that the structural emphasis on multispecialty groups and medical center settings in group/staff models is more conducive to utilizing care teams and mid-level practitioners.8 Often the designation of the PCP is made at the group practice or IPA level, and therefore not dictated or tracked by the health plan. The availability of these providers and the functions which they perform are also affected in large part by state regulations on prescriptive authority and degree of practice autonomy.
Gatekeeping and Referral Processes. A major duty of PCPs in managed care is coordinating access to other providers and services. In pure HMO configurations, the PCP serves as a gatekeeper. The Mathematica 23-HMO study found that while ob-gyns and some other subspecialists may deliver primary care, as do nurse practitioners and physician assistants, they are not often used as gatekeepers. With regard to scope of primary care practice, there was no consensus on conditions or treatments for which PCPs are responsible, but health plans have adopted some mechanisms to define the scope, including specifying responsibilities under capitation arrangements and providing financial incentives to increase individual physicians' range of primary care practice competencies.9
The study also found "wide variations in primary care strategies, from open access to primary care and loose referral policies to tight gatekeeping systems." New enrollees to the plans are generally asked to choose a medical group or IPA or health center in the network, and then further asked to select one primary care provider. Half of the network/IPA model plans delegate the selection rules for the individual PCP to the contracted group or IPA, as the plan often capitates the group or IPA and not the individual providers. Many plans typically allow for some crossover, allowing enrollees either to use various health centers or different providers within a group.
A 1995 Commonwealth Fund survey found that managed care enrollees were more likely than fee-for-service members to rate as fair or poor their ease of changing doctors (25% v. 6%), their choice of doctors (25% v. 5%), access to specialty care (23% v. 8%), and waiting time for a routine appointment (28% v. 11%).10 Largely in response to consumer dissatisfaction, some health plans have begun to relax gatekeeping requirements and ease the process of referrals. New mechanisms include self-referral to ob-gyns for routine visits, self-referral to other specialists for a set copayment per visit, referrals by primary care physicians without plan approval, and open access to in-network specialists.11
Provider Education and Guidelines. MCOs can promote practice standards among their providers, from training to the adoption of formal practice guidelines for clinical conditions to enhanced screening protocols to education about the importance of communication skills. Plans thus have many avenues through which to influence provider behavior in ways beneficial to women's health.
Many health plans conduct training for their providers. Working from the assumption that women themselves can determine who they prefer as a PCP, Kaiser Permanente in Northern California determined that its role was to ensure that any provider serving in a primary care role possess certain competencies. The plan adopted a "cross training" approach, using in-house continuing medical education in the primary care needs of particular types of patients that physicians see. Some HMOs have even begun to undertake physician training at the graduate medical education level, often in partnership with medical schools.12 As managed care's market share continues to grow, its ability to influence medical education curricula will likely grow as well. This is thus a critical time for advocates of curriculum changes related to competency in providing women's health care to dovetail their efforts with those of managed health care plans seeking to influence provider training in the direction of ambulatory and primary care.
Many MCOs actively use practice guidelines (as many as 86% of HMOs, according to AAHP), some adopting those developed elsewhere (by the U.S. Preventive Services Task Force or the Agency for Health Care Policy and Research, for example) and some developing their own. A 1996 report of the U.S. General Accounting Office (GAO) described how managed care plans use practice guidelines to improve physician practice in "problem areas--that is, those services or conditions that are high cost, high medical liability risk, and high incidence for their patient population," as well as in areas of great practice variation.13 The GAO found that most plans modify external guidelines to some extent, primarily to allow for local clinical input from their providers and for the plans' organizational needs. Some experts express concern that such alteration may compromise guideline integrity. Of equal import is the extent to which guidelines are being developed with women's health concerns in mind.
Helping providers learn to screen better for conditions and behaviors for which women may be at particular risk is another means for plans to improve outcomes through provider education. Kaiser Permanente in Oregon is developing a regionwide program for the secondary prevention of domestic violence by increasing the screening of women patients by primary care clinicians. Similarly, HealthPartners in Minnesota is planning to teach staff throughout the HMO how to recognize and treat domestic abuse.14 Several MCOs, such as HealthSystem Minnesota, have begun in-house training programs in patient communication, in light of increasing evidence that good clinician-patient communications result in better outcomes and higher patient satisfaction.15 Such programs are of particular salience to women, who place heavy emphasis on the communication component of care. As organized systems with defined provider networks, managed care plans can thus affect many health care providers at once, influencing their behavior in both clinical and non-clinical areas of care.
Women's Health Centers. Women's health centers deserve special attention, as many have traditionally provided primary care and are now, like other community health centers, threatened by restrictive managed care networks. In 1994 there were an estimated 3,600 women's health centers nationwide, 12% of them classified as primary care centers; an additional 71% were reproductive health centers, of which 26% planned to expand into primary care. These centers were estimated to have served over 14 million women in 1993, and about half of those served used the centers as their usual source of care.16 Displacement of clientele previously served by centers that do not survive is one outcome of managed care market dominance that has a clear adverse impact on some women.
One survey revealed that 66% of women enrolled in HMOs would be more likely to choose a health plan if its network included a women's health center.17 The Bureau of National Affairs reported in early 1996 that "integrated women's services [are] increasingly important to [MCOs] refining their products."18 Some MCOs may opt to develop centers rather than contract with them; New York's Community Health Plan, for example, has a women's primary care department. More loosely structured IPA-model plans, however, are more likely to contract with existing centers, as these plans by definition construct delivery networks from existing providers and institutions.19 An example is Harvard Pilgrim Health Care in Boston, which has contracted with the for-profit Spence Centers to provide services to its enrollees who desire care at a women's center.
Women's health centers are initiating many strategies to position themselves for managed care, including expanding into primary care and mid-life health services, extending services to men or children, contracting for niche services such as abortion, and including hospital-sponsored centers in the hospital's marketing plan.20 Centers are using consultants and attorneys, reviewing model contracts, recruiting staff with managed care and finance expertise, educating staff in the mechanics of managed care, recruiting certified providers, forming alliances with other providers and provider groups, upgrading information systems, and determining unit costs. Concerns include ensuring adequate time for patient-provider communication in the face of pressures for increased efficiency, and retaining a women-centered approach to care.
The above strategies have met with some success. For example the Chico Feminist Health Center in California, seeking to reposition itself in the managed care market, began recruiting a local primary care physician and stepped up its marketing to primary care gatekeepers and MCOs. As a result, one mainstream managed care plan contracted with the Center, allowing its enrollees to access its services without referral or prior authorization.21 Other niche programs and clinics, such as those serving primarily lesbian women or ethnic minorities, face similar issues. Programs such as the Lesbian Services Program (LSP) of Washington, D.C.'s Whitman-Walker Clinic have had a strong community orientation and have financed their services, many provided free or on a sliding-scale basis, through public funding and their own fundraising efforts. They are now also learning the ropes of managed care contracting. LSP is marketing itself to MCOs by emphasizing its willingness to serve all women, regardless of sexual orientation, while striving to retain its traditional identity.22 Thus the growth of managed care plans means that both individual and institutional providers are having to gain access to restrictive provider networks to retain their patient bases.
Preventive Services and Patient Education. Managed care plans have both the ability and the incentive to assist their enrollees in adopting healthy behaviors. Using computer tracking systems allows plans to generate screening reminders and identify at-risk individuals. Plans also offer health promotion classes and provide enrollees with health advice through mailings and newsletters. In these ways, information systems can help plans and providers coordinate primary care services. HIP Health Plan of New Jersey, for example, designed the Women's Health Profile, "an assessment tool to help health care organizations not only understand the risk profile of their female membership, but also to screen, segment and target interventions (including education) for their female members." The survey of women enrollees was designed to identify individual risk factors, which are then forwarded to the women's primary care providers, allowing the PCP and the plan to focus intervention efforts more effectively.24
A Mathematica study published in 1997 by the Commonwealth Fund detailed how several plans have implemented preventive services for women, focusing specifically on breast and cervical cancer screening initiatives because "as the most developed preventive programs for women, they have the most to tell about plans' efforts and experiences in this area."25 The study found that MCOs used two broad strategies for improving the delivery of breast and cervical cancer screening: data-driven (using administrative data, medical charts, patient surveys, and laboratory and pharmacy data) and office-based (assisting office staff in monitoring preventive services). The seven programs in the case study used one or more types of interventions to increase the use of screening, including patient reminders, provider reminders, performance feedback to providers, financial incentives for providers, and office staff reminders and procedures. More data is needed, however, to assess the effectiveness of such tools in improving health outcomes.
While health promotion and disease prevention are philosophically in keeping with the managed care concept and could be cost effective for health plans in the long run, because healthier enrollees are less costly, in markets with high enrollee turnover it is difficult for plans to see a return on such investments in the short run. In a 1995 consumer survey sponsored by the Commonwealth Fund, more than 50% of managed care enrollees surveyed had been in their current plan less than three years, compared with 37% for those in fee-for-service plans. The same survey found that the percentages of members who had not received certain preventive services in the previous year were in fact similar in managed care and fee-for-service: pap test (25% v. 24%); pelvic exam (25% v. 24%); mammography screening (31% v. 31%); and blood pressure reading (14% v. 12%).26
Performance Measurement. Use of the systems described above allows for measurement of various aspects of health care delivery, but ease of tracking is not automatic and plans vary in their level of information system sophistication and capabilities. The most widely used set of performance measures is the Health Plan Employer Data and Information Set (HEDIS), a project of the National Committee for Quality Assurance (NCQA), a private organization that accredits MCOs. Both the accreditation process and reporting HEDIS measures are voluntary, but with both private and public purchasers increasingly looking to quantifiable measures of health care value, accreditation standards and the HEDIS measures serve an important role in setting priorities for improvement. They thus stand as an important point of leverage for women's health advocates seeking to influence health plan and provider behavior. In the current version of HEDIS (3.0), a number of measures relate to primary care services, some with particular relevance to women, such as how many enrollees received breast and cervical cancer screening. HEDIS also measures access to ambulatory or preventive health care and availability of primary care providers.
NCQA's accreditation standards also evaluate primary care services. While performance measures focus on the delivery of discrete types of care, the accreditation process measures how the overall MCO functions and the mechanisms that it has in place to ensure quality care. The accreditation process assesses whether the MCO "has established standards for the availability of PCPs and access [to various types of care]." NCQA also calls for MCOs to adopt practice guidelines for the use of preventive services for the full spectrum of populations enrolled, and at least annually the MCO must monitor and evaluate at least two of these services.
Neither the HEDIS performance measures nor the accreditation standards set benchmarks that plans must attain. Rather, they measure services or check for the existence of monitoring mechanisms without holding plans to any externally imposed standards. HEDIS measures are also only slowly moving towards measuring outcomes and currently focus primarily on process.
Access and Coverage
Industry data on benefits collected by AAHP typically report on a health plan's "best-selling benefits package" in order to capture what services are available to most enrollees. Data on coverage show that "all HMOs cover primary care visits for members in their best-selling benefits package. Most require copayments for primary care visits, but 98.7% charge no deductibles and 97.7% no coinsurance. About 13% of members covered by best-selling packages pay no copayment for primary care visits. Of members who do make copayments, 49% pay $5 per visit and another 23% pay $10. In no responding HMO did members covered by the best-selling package pay more than $15."27 Nearly all plans also reported some limits on coverage beyond enrollee cost sharing, including visit and dollar limits per year.
Within the past year controversy has arisen over the effects of coverage limitations on quality of care and the physician-patient relationship. Reports have surfaced of clauses in MCO contracts with physicians that prohibit providers from discussing treatment or referral options with patients that the plan does not cover. Although evidence is largely anecdotal, the industry has been scrambling to assure consumers that their physicians are not being prevented from discussing all clinical options available, whether or not the health plan will provide and pay for all services. Federal legislation banning such clauses failed in 1996 but has been reintroduced. To the extent that such restrictions are imposed by plans on providers they are likely to have particular impact on PCPs, already under pressure to limit referrals and use of costly treatments.
A survey of 200 large employers by the Washington Business Group on Health that focused on women's health services revealed that MCOs consistently provided the most coverage for routine screening services, particularly preventive screenings and health promotion for women. The women's health care services receiving the most attention were breast cancer screening and prenatal services, and the least commonly covered screenings among all types of insurance (managed care and fee-for-service) were for depression, osteoporosis, and family violence. Smoking cessation programs were the most offered health promotion programs. Fee-for-service arrangements offered significantly less coverage of health promotion programs than did MCOs.28
Carve-Outs. "Carve-outs" are services separated from a comprehensive health benefits package and either not covered or covered through a separate plan or contract. Examples of services that are often carved out of managed care benefits packages are mental health services, pharmacy, vision care, dental care, chiropractic care, and sometimes reproductive health services. In many instances, MCOs or purchasers contract with specialty managed care networks set up to meet only the specialty services needs of the insured population. We now see, for example, dental HMOs, managed behavioral health organizations (MBHOs), chiropractic networks, and pharmacy benefits managers (PBMs).
Mental health is one of the areas most often subject to carve-outs. As of 1996, approximately 300 MBHOs nationwide were serving about 100 million Americans; the top three such companies enrolled 42 million people.29 The proliferation of MBHOs has been extensive enough to warrant the creation of a separate set of accreditation standards by NCQA. Some recent studies have indicated that mental health services under managed care may be inadequate for women. A study of nearly 1,000 patients with depression found that those enrolled in managed care were properly diagnosed only 40% of the time, but their fee-for-service counterparts fared only slightly better, with a correct diagnosis rate of 50%. Appropriate use of efficacious antidepressants was low in both settings, but those in prepaid plans were twice as likely to be prescribed inappropriate mild tranquilizers.30 Another study highlighted problems with access, revealing that HMO enrollees were half as likely as fee-for-service patients to see a psychiatrist and that they had significantly shorter visits with their doctors.31 There are also indications that managed care plans are emphasizing drug treatment over psychotherapy, despite evidence that a combination is often more effective in treating many conditions that disproportionately affect women, such as eating disorders, panic attacks, obsessive-compulsive disorders, and depression.32
According to the AHCPR guidelines on treating depression in primary care, "despite the high prevalence of depressive symptoms and major depressive episodes in patients of all ages, depression is underdiagnosed and undertreated by primary care and other nonpsychiatric practitioners, who are, paradoxically, the providers most likely to see these patients initially." As many as one in eight individuals may require treatment during their lifetimes, with the direct and indirect costs amounting to approximately $16 billion a year (1980 dollars).33 Given such data, it is certainly not in the best interests of either consumers or health plans themselves to minimize women's mental health needs.
Pharmaceutical Formularies. Formularies are lists of prescription drugs that a health plan covers in its benefits package. "Closed" formularies require authorization and greater enrollee cost-sharing for drugs not on the list (similar to using out-of-network providers), while "open" formularies are simply lists of recommended medications. Most plans have provisions whereby physicians can prescribe outside the formulary if medically necessary. Restricting access to certain drugs is a mechanism to achieve cost savings, although the effectiveness of formularies in controlling costs has been questioned and concerns have been raised regarding their impact on quality of care. A controversial study published in 1996 concluded that restrictive formularies actually increase total medical costs by leading to poorer outcomes and more visits to providers and hospitals.34 Formulary adherents challenged these conclusions, citing the limited number of HMOs studied and the lack of an established cause-and-effect relationship between formularies and increased utilization.35 Restrictive formularies may have particular significance for women in terms of the availability of the full range of contraceptive options and anti-depressants.36
Summary and Questions for Further Consideration
Questions raised by this review that merit further attention include the following:
1. Health Affairs 16:1; 1997.
2. A 1995 Commonwealth Fund survey found that one in six managed care respondents had sought non-emergency care outside of their plan during the past year, usually because "they wanted to see a better doctor." The Commonwealth Fund, Patient Experiences with Managed Care, July 19, 1995.
6. A 1996 ACOG survey of ob-gyns found that 37 percent of respondents had little or no interest in serving as primary care physicians, another 37 percent indicated some or high interest, while the remainder expressed no preference. Most of those responding also said that they did not want to be responsible for providing general medical care.
12. Examples include Group Health Cooperative of Puget Sound and the University of Washington, Health Care Plan and the University of Buffalo, Humana and the University of Louisville, and Sierra Health Services and the University of Nevada.
14. Healthcare projects mulled to stem Minn. violence. Modern Healthcare, October 7, 1996.
17. Primary market research, Rynne Marketing Group, Evanston, IL.
18. Scott K. HMOs, hospitals look to women's centers as avenue to growth, patient satisfaction. BNA's Managed Care Reporter 2:7;1996.
33. Rush AJ, Golden WE, Hall GW, et al. Depression in primary care: volume 1. Detection and Diagnosis. Clinical Practice Guideline Number 5. Rockville, MD. U.S. Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research. ahcpr Publication No. 93-0550. April 1993.
35. Academy of Managed Care Pharmacy news release, March 20, 1996.