Institute for the Study of
Healthcare Organizations & Transactions











Increasingly, pharmaceutical benefit management (PBM) companies have developed chronic disease management programs. Many chronic illnesses such as depression, hypertension, and diabetes require more than pharmacologic therapy and noncompliance with medications and other interventions is a major issue. Theories of compliance behavior suggest which interventions/mechanisms can be employed in order to successfully manage individuals who suffer from chronic illnesses. A study sample of 34 firms, national in scope, is derived from cross-sectional, archival data from 1994. The firms in the sample do seem to offer several of the services necessary to conduct a disease management program, however, the actual capability of the information systems to collect data and report information is not clear. Even if the firms' information systems prove to be adequate, a conceptual issue arises: Given what is known from the theoretical and empirical literature on medication compliance, it is questionable whether the techniques employed by these firms alone can be effective in managing chronic illnesses that require multi-faceted approaches to care. Therefore, PBMs may need to enter into interorganizational relationships with other managed care organizations if they are to be successful at managing chronic illnesses.



Pharmaceutical benefit management (PBM) companies, which represent one form of managed care in the drug industry, emerged as a separate organizational structure in the late 1980s and early 1990s. They specialize in managing prescription benefits for employers and for other managed care organizations (MCOs) by utilizing mechanisms that verify employee/subscriber eligibility for benefits, process claims, and manage communications with retail pharmacists (McGahan, 1994).

In 1993, MCOs accounted for 50 percent of the drug market; by the year 2003, the figure is projected to increase to 90 percent (O'Reilly, 1993). As the influence of PBMs grows, it is important to note that PBMs can have both positive and negative effects on the delivery system. On the one hand, they are specialized organizations that have specific ready-made programs, e.g., formularies, report cards, mail-order prescription services, to offer employer clients. In addition, they operate through a large network of pharmacies and therefore they can obtain volume discounts on drugs. Moreover, many have the financial resources to purchase and maintain state-of-the-art technology such as reporting systems and databases (Reissman, 1995).

On the other hand, PBMs can add another layer to the health care delivery system and, as a result, they may have the net effect of increasing administrative costs. In addition, although the technology such as the database may be state-of-the-art in principle, it is not clear that the information collected is adequate to conduct such important tasks as comprehensive outcomes programs. Finally, PBMs may experience the double agent problem because they may not always be clear about who is the primary client/interest to be served. For example, a PBM may contract with an employer or an MCO to provide managed pharmacy services to employees or subscribers. At the same time, a PBM may be involved in an interorganizational relationship with a drug manufacturer. These sets of relationships may present conflicting interests for the PBM.



The study sample consists of 34 PBM firms and represents national data obtained through a survey conducted in 1994 by the trade publication, Business Insurance. It includes information on 8 of the 10 largest PBMs in the country ("Doubts Emerge", 1995) as well as on relatively small firms.

PBMs are based in every geographic region, with the highest concentration (47%) located in the central region of the country. However, PBMs tend to serve a national market through a network of pharmacies. In 1994, the mean number of pharmacies in direct contract networks with PBMs was 33,000 (Kihlstrom, 1996).

Nine core services were identified but only 25 firms (74%) offered all of them (Kihlstrom, 1996). Those services include: claims processing, prescription education for users and physicians, concurrent and retrospective utilization review, formulary review/management, monitoring of physician prescribing patterns, mail order drug plans, assistance with retail pharmacy networks, and consulting services regarding the design of pharmacy benefit plans.

Three core information reporting mechanisms were offered by 30 firms (88%) and included: prescribing patterns/physician profiling, generic vs. brand utilization, and patient/employee utilization reports (Kihlstrom, 1996).



In 1994, nine of the PBMs reported carving out particular diseases for closer scrutiny and used the phrase "disease management" to label these programs (Kihlstrom, 1996). In principle, disease management programs, which evolved out of the continuous quality improvement movement, are efforts to provide cost-effective care for a chronic condition by emphasizing treatment protocols and changes in personal habits. They are designed to focus on the whole spectrum of care for a particular condition, including outpatient, inpatient, and ancillary services. Advocates of these programs suggest that if the entire illness and its natural progression are not considered in cost containment efforts, then those efforts are doomed to failure. By encouraging the use of practice guidelines, measuring the outcomes when guidelines are employed, and providing feedback to physicians, health plans, and medical groups, disease management advocates contend that variations in care can be reduced, better patient outcomes will be produced, and more effective cost saving will be achieved (Terry, 1995).

The disease management programs that have been developed by PBMs have several generic characteristics. First, the PBM reviews employee/subscriber claims to discern which individuals are taking drugs that are related to specific chronic diseases. The usage patterns are then examined to determine if that individual has received the correct drugs at the proper dosage. And, by reviewing the refill patterns, it is possible to learn if the person is taking the medications as prescribed. An individual who may be experiencing difficulty with a prescription may then be sent a letter which offers to provide further information about the chronic illness that requires medication management. If the individual is interested, the PBM will send information in an effort to promote drug compliance. The PBM also may send information to the individual's physician regarding the drug's optimal use. The goal seems to be to decrease the acute episodes that require hospitalization (due to non-compliance with the prescribed medication).

The exact nature of the disease management program will be dependent upon the specific disease to be managed. Several chronic diseases have been targeted by PBMs for management including asthma, ulcers, depression, hypertension, arthritis, heart failure, and osteoporosis. A few PBMs have initiated programs that are designed to manage all of the prescription drugs of the elderly who are enrolled in client health plans.

As one example, one of the largest PBMs, Merck-Medco Managed Care, Inc., has launched a diabetes disease state management program (Gebhart, 1995). Using the medical literature and a expert panel, the company established specific clinical objectives. Individuals who use insulin are identified by using the database of a health plan who may be a client of Merck-Medco. Those individuals are sent a letter asking if they would like more information about diabetes management. Should they consent, they then receive a quarterly mailing from the Merck-Medco pharmacy which includes a personalized letter, program brochures, a toll-free telephone number, a self-administered status survey, and an eight-page newsletter. Individuals also receive regular mailings with other materials such as a blood glucose monitoring diary, food-exchange worksheets, etc. The articles that are distributed in the packet of information focus on motivation, exercise, weight loss, diet, the importance of regular physician visits, and family involvement. In addition, several Merck-Medco pharmacists who received specialized training in diabetes management and telephone counseling techniques provide clinical support by telephone.

The emergence of PBMs into the managed care field is a relatively recent development and not much is known about them, and since disease management itself is an evolving concept, it is not clear how successful PBMs will be at managing chronic illnesses through disease management programs. Although 88% of the PBMs in the study sample seem to offer the information reporting services necessary to conduct a disease management program, e.g., prescribing patterns/physician profiling, generic vs. brand utilization, and patient/employee utilization reports, it is too early to tell if the firms' information systems will be able to collect and report all of the required data that will be needed to successfully conduct a disease management program. However, even if the firms' information systems prove to be adequate, a conceptual issue arises: Given what is known from the theoretical and empirical literature on medication compliance, it is questionable whether the techniques described above, if employed by themselves, can be effective in managing chronic illnesses that often require multi-faceted approaches to care.



There are at least five different orientations that bear on the issue of treatment compliance: 1) a biomedical model; 2) operant behavior and social learning theory; 3) rational belief theory; 4) a communication approach; and, 5) the self-regulative systems theory. Each of these perspectives is thoroughly reviewed in a 1987 article by Leventhal and Cameron.

Of these five broad frameworks, the one that most closely corresponds to the approach employed by the PBMs is the communication approach. From this theoretical perspective, compliance with a treatment regimen seems to depend upon at least six major steps: (1) generation of the message, including specific goals and ways of reaching those goals; (2) reception of the message by the individual; (3) message comprehension; (4) message retention; (5) acceptance or belief in the content of the message; and (6) compliant action (Leventhal, et al., 1984; McGuire, 1980, 1985). Based on these principles, Ley (1981, 1989) developed a model of compliance that is depicted below.

First, in this model, compliance, as the desired end result, is defined as "the extent to which the patient's behavior (in terms of taking medication, following diets, or other life style changes) coincides with medical advice" (Haynes, et al, 1979). This model hypothesizes that 1) the extent to which individuals understand the cause of their illness and the correct location of the relevant organ involved in the illness or the processes involved in the treatment of the illness (Boyle, 1970; Roth, 1979) will influence the degree of compliance (the greater the understanding, the higher the level of compliance with the advice given); and, 2) the ability to remember the information given by the health care provider is a factor that influences compliance.

Although understanding and memory may directly influence compliance, both of these factors may also be mediated by the level of satisfaction with the health care encounter. Satisfaction is derived from various components of the interaction with the health care professional (Ley, 1988, 1989; Haynes, et al., 1979) including the degree of emotional support and understanding provided during the encounter (the higher the level of emotional support and understanding provided, the higher the level of satisfaction with the encounter), behavioral aspects such as, was the explanation given by the provider adequate?, and the perceived competence of the health care professional (the greater the perceived competence, the greater the satisfaction). Studies (Korsch and Negrete, 1972; Francis, Korsch, and Morris, 1969) that were conducted suggest that these components of patient satisfaction are positively related to compliance with clinical advice.

In summary, disease management programs, whose theoretical underpinnings are familiar, have their origins in the continuous quality improvement movement. The goal of these programs, at least as it has been articulated in the trade journals, is to improve patient compliance with the prescribed medications by enhancing patient understanding of the illness and of the medication regimen. And, the way in which patient understanding is enhanced in these programs is by using educational materials and by seeking to improve the communication channels between several critical actors: the patient-pharmacist, the physician-pharmacist, and ultimately the patient-physician.



While receiving, understanding, remembering, and accepting information about a particular chronic illness and about the treatment of it through the use of prescribed medications are essential for compliance, these elements are not sufficient (Leventhal and Cameron, 1987). The communication approach does not account for the ways in which information about the health threat actually affects attitude change or if attitude change occurs, what specific mechanism leads to compliance with treatment instructions. Referring back to the original model as constructed by Ley, Leventhal and Cameron (1987) seem to object to the direct paths that lead to compliance and argue that the causal mechanisms of compliance are not fully accounted for in the model.  Taking into account these criticisms, perhaps the model is more like the below.

In addition, Haynes, Wang, and Gomes (1987) suggest that compliance with short-term treatments (less than 2 weeks) can be improved by the use of clear instructions, by special "reminder" pill containers and calendars and through educational efforts in general. However, they argue that compliance with long-term treatments is more difficult to achieve. Long-term medication compliance for chronic illness such as hypertension, diabetes, depression, etc., requires a combination of interventions: provision of clear instructions, patient self-monitoring of compliance and/or treatment outcomes, enhancement of social support, contingency contracting and rewards or reinforcement for high levels of compliance, group discussion, and supervised self-management (Haynes, Wang, Gomes, 1987). No single intervention has proved to be effective. Moreover, these interventions must continue as long as compliance is required.

Finally, Conrad (1985) examined the medication practices of individuals with epilepsy and suggests that variation from a prescribed regimen, or noncompliance, may depend more on the social meaning of medication and self-regulation for the individuals who suffer from a chronic illness than on their perceptions about the clinical information given to them. He identified four primary reasons for variation from the prescribed medication regimen: testing, control of dependency, destigmatization, and "practical" considerations.

First, individuals may take themselves off of medication or alter the dosage as an experiment to see what the effect might be. For individuals with certain chronic illnesses, this kind of testing is a way to evaluate how the disorder is progressing (Conrad, 1985). Second, while taking medication increases self reliance, e.g., by reducing the likelihood of seizures, taking medications can be experienced by the individual as a threat to self-reliance because medications sometimes become symbolic of the dependence created by having certain chronic illnesses (Conrad, 1985). Third, for many individuals chronic illness and the treatment involved, e.g., taking medication, is stigmatizing and is a constant reminder of the illness. Therefore, avoiding medications might be viewed, by the individual, as a way of avoiding the stigma that is often attached to some chronic illnesses such as epilepsy and depression. Finally, some individuals with epilepsy reported that they altered the medication regimen to correspond with their perceptions of the social environment. For example, some reduced medication to avoid problems that might occur from mixing alcohol and drugs. Others reported altering the dosage during periods of "high stress" (Conrad, 1985). Given these findings, simply providing information to individuals about the chronic illness and the required medications, as the communication approach suggests, may not be sufficient. The notion of the social meaning of noncompliance may need to be explored with individuals.



In summary, the requirements of a comprehensive disease management program can be daunting. For example, the provider network should include disease specific as well as primary care providers. Practice guidelines and referral patterns for the targeted disease must be established. Mechanisms that can identify and recruit individuals into the programs must be instituted and should include appropriate incentives and disincentives. Also, it has been argued that specialized disease managers who understand the chronic illness and its treatment should be an integral part of any program. Finally, one of the most important products of PBMs' information systems is the data that are available for outcomes research. However, as noted above, it is not clear how well the information systems within PBMs will perform.

Disease management programs that seek to oversee the use of prescription medications for chronic illnesses will undoubtedly continue to grow. The communication approach that is employed by PBMs in their chronic disease management programs may offer several of the steps necessary to promote medication compliance. However, this approach does not fully account for critical mechanisms such as the relationships between information processing and attitude change and between attitude change and behavior change. Therefore, in addition to this approach and its mechanisms, perhaps other models should be explored by PBMs that more thoroughly address these issues. As just one example of an alternative, Leventhal and Cameron (1987) offer the self-regulative system theory. But there are other approaches that should be explored as well.

In addition, if Haynes, Wang, and Gomes (1987) are correct that long term compliance requires a variety of interventions and if, as Conrad (1985) believes, medication noncompliance can have social meaning for individuals, then services such as support groups and/or discussion groups and mechanisms such as contingency contracting need to be available as part of ongoing disease management programs. However, many PBMs may not have the organizational capacity to offer all of the necessary services by themselves and therefore they may need to enter into relationships with other organizations such as managed behavioral health care organizations or HMOs if they are to be successful at managing chronic illnesses.

One example of this type of relationship has already emerged (Muirhead, 1996). Value Health Inc, a large managed care organization, has combined the resources of its PBM, ValueRx, and its mental health and substance abuse management division, Value Behavioral Health, to manage individuals' use of psychotropic drugs for depression. In general, this disease management program operates much like the diabetes management program described above. However, in this case, interventions may be initiated by the mental health professionals within Value Behavioral Health who can detect difficulties with an individual's progress in therapy which, in turn, may suggest medication problems or by the pharmacists within Value Rx who may note drug-use problems through a review of the medication use profile (Muirhead, 1996). However, these arrangements are still emerging and, as a result, not much is known about the extent to which they will offer successful solutions to medication noncompliance.

Because disease management programs within PBMs are still emerging, very little data exist that indicate the extent to which these programs actually alter patient outcomes. Much more work needs to be conducted on the efficacy of the various approaches, like the communication approach for example. Another more interesting study might be designed along the lines of a randomized clinical trial. One group of individuals might receive information through various mailings and telephone conversations with pharmacists while a second group would receive information and support using alternative approaches. An impact analysis (Schalock, 1995) that examines the cost of the two programs as well as the outcomes could be conducted in an attempt to determine the differences in outcomes.

Finally, the development of the interorganizational relationships that involve PBMs, HMOs, and managed behavioral health care firms, along with the ways in which disease management programs are structured within these alliances, should be examined. Since investigators do not have control over events and the focus is on contemporary dynamic events, the most fruitful approach might be one that examines several of these relationships as part of a multiple case study strategy (Yin, 1994). One interesting question might be, How do these relationships form and sustain themselves? Another might be, How do the organizational structures and mechanisms influence the operation of the disease management programs and ultimately patient outcomes?

PBMs are a relatively new and growing concern in the managed care world. Not much is known about the cost and quality of the services provided by them but because the management of chronic illness is a complex process and may have an impact on members of vulnerable populations such as the elderly, in the future, this line of research will be crucial. However, the research effort will require a multi-faceted approach and should include clinical researchers, service researchers, and information system managers. The research findings should be of interest to health care managers as well as policymakers.



Annual Directory of Prescription Benefit Management Companies. (1995, April 3). Business Insurance. Chicago, IL: Crain Communications, Inc.

Boyle, C.M. (1970). Differences between patients' and doctors' interpretations of common medical terms. British Medical Journal, 2, 286-9.

Conrad, P. (1985). The meaning of medications: Another look at compliance. Social Science and Medicine, 20, 29-37.

Doubts emerge about drug industry mergers. (1994, November). Business & Health, 53-61. [Correction: Business & Health (1995, February), 7.]

Francis, V., Korsch, B.M., & Morris, M.J. (1969). Gaps in doctor-patient communication: Patients' responses to medical advice. New England Journal of Medicine, 280, 535-540.

Gebhart, F. (1995, August 21). Medco expanding its mail-order diabetes management program. Drug Topics, 59-60.

Haynes, R.B., Sackett, D.L., & Taylor, D.W. (Eds.) (1979). Compliance in Health Care. Baltimore, MD: Johns Hopkins University Press.

Haynes, R.B., Wang, E., & da Mota Gomes, M. (1987). A critical review of interventions to improve compliance with prescribed medications. Patient Education and Counseling, 10, 155-166.

Kihlstrom, L.C. (1996, June). Managed care and mergers in the prescription drug industry. Presentation at the annual meeting of the Association for Health Services Research, Atlanta, GA.

Korsch, B. & Negrete, V. (1972). Doctor-patient communication. Scientific American, 227, 66-74.

Leventhal, H. & Cameron, L. (1987). Behavioral theories and the problem of compliance. Patient Education and Counseling, 10, 117-138.

Leventhal, H., Zimmerman, R., & Gutmann, M. (1984). Compliance: A self-regulation perspective. In D. Gentry (Ed.), Handbook of Behavioral Medicine (pp. 369-434). New York, NY: Pergamon Press.

Ley, P. (1981). Professional non-compliance: A neglected problem. British Journal of Clinical Psychology, 20, 151-4.

Ley, P. (1988). Communicating with Patients, London: Croom Helm.

Ley, P. (1989). Improving patients' understanding, recall, satisfaction, and compliance. In A. Broome (Ed.), Health Psychology. London: Chapman and Hall.

McGahan, A. M. (1994, November-December). Industry structure and competitive advantage. Harvard Business Review, 115-124.

McGuire, W.J. (1980). The communication-persuasion model and health-risk labelling. In L.A. Morris, M.B. Mazis, & I. Barofsky (Eds.), Product Labelling and Health Risks (pp. 99-122). New York, NY: Banbury.

McGuire, W.J. (1985). Attitudes and attitude change. In G. Lindzey & E. Aronson (Eds.) Handbook of Social Psychology: Vol. II, 3rd Edition (pp. 233-346). New York, NY: Random House.

Muirhead, G. (1996, August 19). Value Health to manage use of psychotropics. Drug Topics, 58.

O'Reilly, B. (1993, September 20). Why Merck married the enemy. Fortune, 60-64.

Reissman, D. (1995, July-August). Contracted pharmacy services: The value of a carve-out PBM. HMO Magazine, 57, 60-61.

Roth, H.P. (1979). Problems in conducting a study of the effects of patient compliance of teaching the rationale for antacid therapy. In S.J. Cohen (Ed.), New Directions in Patient Compliance (pp. 111-26). Lexington, MA: Lexington Books.

Schalock, R.L. (1995). Outcome-Based Evaluation. New York, NY: Plenum Press.

Terry, K. (1995, April). Disease management: Continuous health-care improvement. Business & Health, 64-72.

Yin, R.K. (1994). Case Study Research: Design and Methods, 2nd Edition. Thousand Oaks, CA: Sage.



Paper presented at the annual meeting of the American Public Health Association, November 19, 1996.

Acknowledgments: This paper was prepared with support, in part, from a National Institute of Mental Health grant, MH 15783, awarded to Yale University.


Lucy Canter Kihlstrom, Ph.D.

Copyright 2000 Institute for the Study of Healthcare Organizations & Transactions

Last modified:  04.08.2010 02:58 PM