Since the mid-1990s the costs of prescription. How Effectively Do Managed Care Organizations Influence Prescribing and Dispensing Decisions?

Transcription

1 How Effectively Do Managed Care Organizations Influence Prescribing and Dispensing Decisions? Norman V. Carroll, PhD Objective: To examine the extent to which managed care organizations (MCOs) use formularies, therapeutic interchange, and prior approval and to determine how effectively these tools influence prescribing and dispensing decisions. Research Design: Literature review. Patients and Methods: Studies relating to effectiveness were identified through a comprehensive literature review using the MEDLINE and International Pharmaceutical Abstracts databases. Only peer-reviewed studies done in outpatient settings were included. Studies measuring extent of use were taken primarily from published and widely available marketing research reports. Results: Closed formularies were found to be effective in decreasing the utilization, but not necessarily the cost, of prescription drugs. Just under half of health maintenance organizations (HMOs) and 10% of employer-sponsored health plans use closed formularies. Prior approval programs have been shown to reduce use of target drugs and drug costs in a small number of drug classes. Nearly all HMOs and most employersponsored health plans use prior approval programs. How extensively the programs are used is not reported. About half of HMOs and employer-sponsored health plans use therapeutic interchange. Voluntary programs have been shown to be successful in staff-model HMOs. Mandatory, but not voluntary, programs have been shown to be successful in independent practice association-model HMOs. Conclusion: The literature indicates that most MCOs have had limited success using formularies, therapeutic interchange, and prior approval to influence prescribing and dispensing decisions. Although these tools have been effective in some situations, their impact has been limited by their low rate of utilization. (Am J Manag Care 2002;8: ) by at least 10% every year since In 1999 the growth rate was 16.9%. 1 As this growth has occurred, the employers and health plans that pay for prescription drugs have increasingly turned to health maintenance organizations (HMOs) and pharmacy benefit managers (PBMs) for help in controlling prescription drug costs. (In this article, HMOs and PBMs will be referred to as managed care organizations [MCOs]. It is reasonable to consider HMOs and PBMs together in the context of drug benefit programs because both use the same types of tools to control pharmaceutical expenditures and because more than 90% of HMOs use PBMs in their drug benefit programs. 2 ) One of the major means by which MCOs have attempted to control drug costs has been through influencing physicians prescribing and pharmacists dispensing decisions. The primary tools that MCOs have used to influence prescribing and dispensing are formularies, therapeutic interchange, and prior approval. The formulary is a continually updated list of medications and related information, representing the clinical judgment of physicians, pharmacists, and other experts in the diagnosis and/or treatment of disease and promotion of health. 3 The goal of a formulary is to identify drug products and therapies that are the most medically appropriate and costeffective to best serve the health interests of a given patient population. 3 A formulary can reduce drug costs directly by guiding physicians and pharmacists to use less expensive, equally effective products and Since the mid-1990s the costs of prescription drugs have been rising at a rapid and increasing rate. In 1993, national expenditures on prescription drugs at the retail level were $51.3 billion; by 1999 they had grown to $99.6 billion. 1 In addition, the overall rate of spending on drugs has grown From the School of Pharmacy, Virginia Commonwealth University, Richmond, Va. Address correspondence to: Norman V. Carroll, PhD, Professor of Pharmacy Administration, School of Pharmacy, Virginia Commonwealth University, Box , Richmond, VA nvcarrol@vcu.edu. VOL. 8, NO. 12 THE AMERICAN JOURNAL OF MANAGED CARE 1041

2 indirectly by preventing them from using less effective or more dangerous products. Formularies can be open or closed. An open formulary is a list of drug products that the MCO suggests that physicians use. Although the MCO prefers that physicians prescribe and pharmacists dispense products on the formulary, it will provide reimbursement for nonformulary products. A closed formulary is an exclusive list of drug products for which the MCO will provide reimbursement. If the physician prescribes a product that is not included on a closed formulary, the patient has to pay out of pocket for that product. Many formularies are partially closed. In other words, they are closed for some drug categories (eg, antilipid or antiulcer drugs) and open for others. Therapeutic interchange occurs when a pharmacist dispenses a product that is therapeutically equivalent to, but chemically different from, the one prescribed by the physician. A therapeutically equivalent product is one that is expected to have the same therapeutic results with a similar side effect profile. Therapeutic interchange differs from generic substitution. Generic substitution involves dispensing a different brand of a product that is chemically identical to the one prescribed. Therapeutic interchange involves dispensing a chemically different product. Because a chemically different product is dispensed, the pharmacist (in most settings outside hospitals or staff-model HMOs) must contact the patient s physician and receive the physician s approval for the switch before dispensing the product to the patient. Many MCOs have formal therapeutic interchange programs that contact physicians, pharmacists, and patients to request that they switch to products preferred by the MCO. Frequently, a pharmacist employed by the MCO makes the contact by telephone. Other MCOs mail letters to physicians, pharmacists, and patients. As with formularies, therapeutic interchange can reduce drug costs by persuading physicians and pharmacists to use products that are less costly, more effective, or both. Managed care organizations use prior approval programs to control utilization and expenditures of selected products. Typically, these products are very expensive or have serious side effects. A prior approval program requires that a pharmacist or physician receive explicit approval from the MCO before dispensing or prescribing one of the selected products. Reimbursement is not provided unless prior approval is obtained. Because patients are unlikely to purchase prescriptions for which their third-party payer does not provide reimbursement, prior approval programs generate savings by reducing the use of expensive products and by inducing physicians and pharmacists to switch patients to less expensive products. The common goal of formularies, therapeutic interchange, and prior approval is to persuade physicians and pharmacists to switch the drugs they prescribe or dispense to therapeutically equivalent products that are preferred by the MCO. Typically, the preferred products are less expensive than those they replace. In this article, these programs will be referred to collectively as drug preference programs. It is widely assumed that MCOs can and do influence prescribing and dispensing decisions through use of drug preference programs. Although a few studies have examined this assumption in defined and specific instances (eg, the effectiveness of a closed formulary in a given health plan), the assumption needs to be examined in a more comprehensive manner. It is important that this be done because of the critical relationship between MCOs ability to influence prescribing and dispensing decisions and their ability to control pharmaceutical use. The purpose of this article is to provide a comprehensive examination of the effects of drug preference programs on prescribing and dispensing decisions. Both the extent to which MCOs use drug preference programs and the effectiveness of these programs in influencing prescribing and dispensing decisions are examined. Also discussed are the outcomes when these programs are used effectively, including changes in drug- and health-related costs and outcomes. Disease management programs were not considered in this study. These programs are broad, multidisciplinary, patient-focused programs that attempt to better manage a patient s disease over the patient s life span. Attempts to influence prescribing and dispensing decisions are only one component of disease management programs. In contrast, formularies, therapeutic interchange, and prior approval programs focus directly on influencing prescribing and dispensing decisions. The intended effects of these tools on prescription drug use are direct and straightforward. These programs are intended to decrease the use of nonpreferred products and increase the use of preferred products. The effects of disease management programs on drug use are neither direct nor straightforward. Although they may direct physicians and pharmacists to use selected products, disease management programs emphasis 1042 THE AMERICAN JOURNAL OF MANAGED CARE DECEMBER 2002

3 MCO Influence on Medication Use on prevention and early and aggressive treatment may result in increased use and drug costs for both preferred and nonpreferred drugs. METHODS Studies that examined the effectiveness of drug preference programs were identified through an exhaustive review of MEDLINE and International Pharmaceutical Abstracts databases. MEDLINE was searched for the period from 1966 through December 3, 2001, using the terms drug product selection, formulary, therapeutic interchange, therapeutic substitution, prior authorization, and prior approval. The search was limited to studies that were conducted in humans, published in English, and included one of the search terms in the title or abstract. The search identified 848 articles. The International Pharmaceutical Abstracts database was searched for the period from 1970 through December 3, 2001, using the same terms. The search of article titles identified 930 articles. The author identified additional relevant articles by scanning the references of those identified by the computerized search. In addition, experts in the field were contacted and asked to identify relevant articles. The author selected data-based research articles in peer-reviewed journals that examined the effectiveness of formularies, prior approval, or therapeutic interchange in outpatient settings similar to those served by MCOs. Excluded were studies conducted in inpatient hospital settings, long-term care facilities, or indigent-care clinics and studies that were based on modeling rather than actual data. Studies conducted in outpatient hospital settings were included in the review because they are similar to the programs operated by staff-model HMOs. Both serve ambulatory populations; use formularies, prior approval, and therapeutic interchange to influence pharmaceutical use; and serve patients who get most of their prescriptions through inhouse pharmacies because drugs are provided there at reduced cost. The review excluded studies of therapeutic interchange in which patients were switched to a different dosage form of the same drug or that included a switch from one drug to a combination of drugs. Also excluded were studies of state Medicaid formularies that excluded entire classes of drugs. Formularies operated by MCOs typically exclude products within classes, but not entire therapeutic classes. Fifty-six studies, spanning the years 1984 to 2001, met these criteria. Studies of closed formularies were more prevalent early in this period, and studies of therapeutic interchange were more prevalent in recent years. Most of the information on utilization of drug preference programs came from reports published annually by marketing research companies (Table 1). Table 1. Marketing Research Reports Used to Estimate Frequency of Use of Drug Preference Programs Company Conducting Research Sample on Which Report Title and Compiling Report Sponsor of Report Report Is Based The Wyeth-Ayerst Prescription Pharmacy Benefit Management Wyeth-Ayerst 446 employers representing Drug Benefit Cost & Plan Design Institute, Inc 15 million employees and dependents Survey Report: 2000 Edition 4 Managed Care Digest Series 2000: SMG Marketing Group Aventis Census of the 820 HMOs operating HMO-PPO/Medicare-Medicaid in the United States in 1999 Digest 2 Novartis Pharmacy Benefit Report: Emron/IMS America Novartis Sample of HMOs representing Facts & Figures, 2000 Edition million members Novartis Pharmacy Benefit Report: Emron/IMS America Novartis 50 HMO pharmacy directors Trends and Forecasts, and 27 medical directors 1998 Edition 6 representing 70 health plans and 19 million enrollees Pharmacy directors of 12 PBMs serving 23.5 million consumers HMO indicates health maintenance organization; PBM, pharmacy benefits manager. VOL. 8, NO. 12 THE AMERICAN JOURNAL OF MANAGED CARE 1043

4 These reports are widely available and well known within the pharmacy field. They are funded by major pharmaceutical companies but are produced independently, and without editorial input, from the sponsors. It is unlikely that sponsors bias the results because of the nonproprietary, non product-specific nature of most of the information provided in these reports. The reports consist of basic statistics describing the managed care industry and use of pharmacy benefit management techniques by MCOs. Other sources included reports published by professional associations, benefits management companies, and government agencies. Because use of drug preference programs has grown steadily over the past 10 years, this review focused on the most recent utilization reports available. These included reports published between 1998 and RESULTS The literature review results are presented by type of program. For each type, evidence of both its effectiveness and the frequency with which it is used is presented. Formulary Effectiveness Little information is available about the effectiveness of open formularies. The available information suggests that open formularies have minimal impact on prescribing and dispensing decisions. 7,8 This is probably because open formularies provide few incentives for pharmacists, physicians, or patients to use listed products. In addition, open formularies may be less rigorously or scientifically determined and, therefore, may be perceived as providing less appropriate guidance. A number of studies have examined the effectiveness of closed and partially closed formularies (Table 2). Most have indicated that increasing the restrictiveness of a formulary will result in lower utilization of restricted products Some have noted increased use of formulary products, 9,10,13-15 although others did not examine use of specific products. 11,12 Lower utilization has not always been shown to result in lower drug expenditures, however. Two studies have documented lower drug costs in a closed formulary. 9,12 Others have indicated no statistically significant differences between costs in formulary and nonformulary groups. 10,13,15 A few studies have suggested that restricting formularies may lead to negative health consequences. These consequences have included increases in inpatient hospitalizations 11,12,16,17 and emergency room visits. 16,17 Two of these studies 16,17 have been widely criticized for deficiencies in research design. Major criticisms focused on the small number of data points (ie, MCOs) included in the study, use of an after-only study design without randomization, and ignoring discounts and rebates (which may be substantial in MCOs) when calculating drug costs Two studies have reported that more restrictive formularies are associated with lower rates of continuation with drug therapy. 9,18 Schweitzer examined the relationship between the generosity of a health plan s formulary and consumers satisfaction with the plan. 19 The measure of formulary generosity considered both the number of products covered and whether expensive products were excluded from the formulary. The results indicated no significant relationship between formulary generosity and health plan satisfaction. Utilization of Formularies The most recent surveys indicate that 74% of employer-sponsored plans 4 and virtually all HMOs 2,5 use formularies. Use of closed formularies is much less common. A census of all HMOs operating in 1999 indicated that 49% of HMOs use closed formularies. 2 Another survey indicated that 49% use closed formularies and 18% use partially closed formularies. 5 Formulary use is substantially lower outside HMO settings. Various surveys indicate that only about 8% to 18% of employer-sponsored health plans use closed formularies. 6,23,24 Between 24% and 42% of employer-sponsored plans use partially closed 6 or incented formularies. 23 (An incented formulary provides some coverage for nonformulary products, but requires the patient to pay a higher amount to receive these products.) Two other studies, although not providing numerical estimates, also suggest that closed formularies are rarely used by PBMs. 25,26 Use of more restricted formularies has increased substantially over the last several years. The percentage of HMOs using closed formularies grew from 23% in 1992 to 49% in ,27 Use of closed formularies in employer-sponsored plans has remained relatively stable since 1995, but use of incented formularies has grown from a few plans in 1996 to 25% of responding plans in ,28 Effectiveness of Prior Approval Four published studies have evaluated the effectiveness of prior approval programs. All found posi THE AMERICAN JOURNAL OF MANAGED CARE DECEMBER 2002

5 MCO Influence on Medication Use Table 2. Studies of Effectiveness of Closed Formularies Study Setting and Sample Research Design Measure Result Horn et al 16 National sample of After-only without Degree of formulary Increased formulary restrictiveness patients from randomization restrictiveness was associated with higher drug 6 managed care costs and increased emergency organizations department visits and hospitalizations Horn et al 17 Same as above Same as above Same as above Same as above; in addition, some associations were stronger for elderly patients Motheral and 5890 continuously Before and after Closed vs open Smaller increases in drug use and Henderson 9 enrolled patients in with control group formulary expenditures in the closed 2 employer-sponsored formulary plan drug plans Motheral et al patients in 2 Case-control, before Closed vs open Smaller increase in drug use in government-employee and after with formulary closed formulary plan; smaller, drug plans control group but not statistically significant, increases in drug costs in closed formulary plan Kreling et al 13 Wisconsin Medicaid Before and after Dropping Slight increase in expenditures on program; sample of without control group propoxyphene analgesics; substantial decreases in more than napsylate from number of analgesic prescriptions recipients in each formulary and recipients time period Smith and 137 Michigan Survey of patients How patients 46% discontinued the drug, McKercher 14 Medicaid recipients taking 1 or more responded to having 23% switched to an alternate drug, products dropped their medicines 31% paid out of pocket for drug from formulary dropped from the formulary Lingle et al patients Before and after Open vs closed Increased use of drugs, physician continuously enrolled without control group formulary office visits, and outpatient hospital in the South Carolina services and decreased inpatient Medicaid program hospitalizations after opening the formulary Kozma et al 12 Subset from the Same as above Same as above Increased use of and expenditures for sample above of drugs, physician office visits, and patients outpatient hospital services, and having at least 1 claim decreased inpatient hospitalizations for a prescription drug and expenditures after opening during the study period the formulary Soumerai et al patients Time series analysis Changes in use and Use of and expenditures for the enrolled in the expenditures after 12 ineffective products dropped New Jersey Medicaid dropping reimburse- substantially; however, increases in program ment for 12 ineffective use of substitute products completely but commonly used offset decreased expenditures for drug products the ineffective products Streja et al patients from a Regression analysis of Formulary with only Patients with only 1 SSRI on single group practice patients newly paroxetine vs formulary were 80% less likely to in California prescribed a study formulary with complete therapy antidepressant paroxetine and fluoxetine Schweitzer health plans in Correlation of A measure of formulary No significant association between California formulary generosity generosity defined to formulary generosity and satisfaction and consumer include both the with the health plan satisfaction with number of products the plan covered and the price of the products SSRI indicates selective serotonin reuptake inhibitor. VOL. 8, NO. 12 THE AMERICAN JOURNAL OF MANAGED CARE 1045

6 tive, but limited, evidence of their effectiveness in reducing drug costs (Table 3). Two studies evaluated the effectiveness of prior approval programs for brand name nonsteroidal anti-inflammatory drugs (NSAIDs) in state Medicaid programs. 29,30 Both studies found that prior approval programs substantially decreased overall utilization of NSAIDs, decreased utilization of brand name NSAIDs, increased use of preferred generic products, and yielded significant savings even after accounting for changes in use of related healthcare services. However, authors of both studies caution that their results may not be widely applicable. Both studies examined prior approval programs only for NSAIDs. This drug class has the following characteristics: it includes generic products, brand name products in the class have no significant therapeutic advantages over generics, drugs in the class are used to treat a symptomatic problem of mild to moderate severity, and there is little risk in trying a different agent from the class because a therapeutic failure is unlikely to result in hospitalization, use of other expensive healthcare services, or serious damage to the patient s health. Authors of both studies suggest that it may not be reasonable to generalize their results to therapeutic classes that do not share these characteristics. A third study evaluated drug savings effected by Iowa Medicaid s prior approval program. 31 This program required prior approval for brand name NSAIDs, brand name benzodiazepines, nonsedating antihistamines, and high doses of ulcer medications (compared with the lower doses needed for maintenance therapy). The program generated substantial savings on drug costs, primarily through increased use of preferred products. The majority of savings came from increased use of generic NSAIDs and benzodiazepines. The benzodiazepine class, like the NSAIDs, includes generic products that are therapeutically equivalent to brand name products and is Table 3. Studies of Effectiveness of Prior Approval Programs Study Setting and Sample Research Design Measure Result Smalley et al 29 More than Interrupted time series Net cost savings from PA 53% decrease in expenditures on patients in Tennessee for brand name NSAIDs NSAIDs with no concomitant Medicaid Program increases in use of other pain relievers, emergency room visits, or hospitalizations; 19% decrease in NSAID use Kotzan et al continuously Interrupted time series Net cost savings from PA Large savings on NSAIDs due to eligible patients in the for brand name NSAIDs decline in use of brand name Georgia Medicaid NSAIDs and overall decrease program in use of all nonnarcotic analgesics Kotzan et al 32 Georgia Medicaid Cross-sectional Differences in market The market share of PA products program and a share of PA products for cash patients was 5.6% of comparison group between the Medicaid prescription numbers and 10.9% of cash-paying population and the of dollars; the market share for patients; sample sizes control group of Medicaid patients was 2.4% of were not reported cash-paying patients prescription numbers and 7.9% of dollars Phillips and Larson 31 Iowa Medicaid Before and after Changes in drug Decreases in drug costs of between program served more without control group expenditures after $2.51 and $3.83 million for these than implementation of PA classes in the 12 months after enrollees at the time program for brand name program implementation of the study; sample NSAIDs and benzosize was not reported diazepines, nonsedating antihistamines, and high-dose ulcer products NSAIDs indicates nonsteroidal anti-inflammatory drugs; PA, prior approval THE AMERICAN JOURNAL OF MANAGED CARE DECEMBER 2002

7 MCO Influence on Medication Use a class for which a therapeutic failure is unlikely to result in serious health consequences or increased use of more expensive healthcare services. Kotzan et al used a market share analysis to estimate cost savings due to prior approval in the Georgia Medicaid program. 32 They compared market shares of products that required prior approval in the Medicaid program with the market shares of those same products among noninsured patients. Medicaid patients were 44% less likely to take a drug requiring prior approval than were noninsured patients. Based on the substantially higher cost of drugs requiring prior approval, they concluded that the prior approval program had resulted in savings for the Georgia Medicaid program. The authors conclusions are based on 2 related assumptions. The first is that market shares of prior approval products would have been equivalent in the cash and Medicaid markets in the absence of the prior approval program. The second is that the differences in market shares were primarily a result of the program. Given the likelihood of differences in socioeconomic and healthcare status between the 2 groups and the absence of baseline measures for either group, it is difficult to accept these assumptions. Utilization of Prior Approval Programs Between 80% 2 and 88% 5 of HMOs use prior approval programs. Use is lower among employersponsored plans. About 66% of these plans use prior approval programs. 4 The cited sources report the percentages of health plans that use prior approval, but most do not indicate how extensively those programs are used. Whether the plans have limited programs that cover only 1 or 2 drug classes or extensive programs that cover many classes is not reported. However, data from the Novartis Pharmacy Benefit Report: Facts & Figures, 2000 Edition indicate that about 3% of HMO prescriptions are changed or not dispensed as a result of prior approval. 5 The percentage of HMOs using prior approval increased from 14% in to 80% in Among employer-sponsored plans, use has grown from 43% of plans in 1996 to 66% in ,28 Effectiveness of Therapeutic Interchange Programs A number of studies have examined the cost savings and clinical effects of therapeutic interchange programs in outpatient settings Most examined the interchange of a single product in a small sample of patients attending a hospital outpatient clinic. The majority examined switches of statins, angiotensin-converting enzyme inhibitors, or calcium channel blockers. These studies found substantial savings as a result of the switch. In addition, almost all reported an equivalent or an improved clinical effect after patients were switched. However, limitations in the research designs of these studies make it difficult to draw firm conclusions about the effectiveness of therapeutic interchange programs. The samples in many of these studies are biased by study inclusion criteria. They excluded patients who did not return to the clinic for evaluation after the switch, patients who were switched back to the original drug due to adverse reactions or lack of therapeutic effect, and/or those who stopped taking the drug for any reason. Eliminating noncompliant patients and patients who did not respond, or responded adversely, to the switched medication substantially weakens the conclusion that switching had no detrimental effects. A few studies found less positive results after interchange programs. Amidon et al, 56 Nelson et al, 57 and Condra et al 58 found failure rates between 14% and 28% for a proton pump inhibitor switch. Oatis and Stowers reported a 15% failure rate for a calcium channel blocker switch. 59 Stock and Kofoed reported a 37% failure rate for switches among selective serotonin reuptake inhibitors. 60 Several studies also found higher costs after the switch. Oatis and Stowers 59 and Mamdani et al 40 found higher drug costs after a calcium channel blocker switch as a result of higher doses of the switched product 40 and higher use of concomitant heart medications. 59 Amidon et al found that the high failure rate (28%) for a proton pump inhibitor switch offset any savings. 56 A number of surveys have measured patient perceptions of the outcomes of therapeutic interchanges. Surveys of patients switched from loratidine to cetrizine 61 and from omeprazole to lansoprazole 57,58 found decreased control of symptoms and lower levels of patient satisfaction with drug treatment after the switch. Gardner et al found that 25% of women switched from conjugated to esterified estrogens experienced a return of menopausal symptoms and 15% experienced new symptoms. 62 Table 4 describes several large evaluations of therapeutic interchange programs that do not suffer from the subject inclusion limitations of the studies discussed above. All indicate that therapeutic interchange programs can result in substantial savings on drug costs. Most indicate that therapeutic interchange programs result in savings even after VOL. 8, NO. 12 THE AMERICAN JOURNAL OF MANAGED CARE 1047

8 Table 4. Studies of Therapeutic Interchange Programs Study Setting and Sample Research Design Measure Result Brufsky et al 63 Staff-model HMO with Interrupted Changes in market share Cimetidine s share increased by 54% members and monthly time of cimetidine after in the staff model and by 10% in group-model HMO series with a voluntary program the group model; net savings of with members comparison series consisting of prescriber $1.06 million education, therapeutic reevaluation of patients, and performance feedback Baluch et al 64 Staff-model HMO with Before and after Voluntary switch of patients 89% of patients were switched from members without control from CE to EE; the program CE to EE; after 6 months, 6.5% were group was promoted through switched back to EE; net savings were newsletters mailed to patients $32 per patient per year and prescribers, and pharmacist presentations at clinics Gardner et al 62 Same as above Survey of patients Patient evaluation of and 9% of those switched and maintained on estrogens experiences with the switch on EE and 25% of those switched to 3 years after from CE to EE EE then switched back to CE rated the therapeutic switch experience as negative; 25% of interchange those switched and maintained on EE intervention reported return of menopausal symptoms and 15% reported new symptoms; of those switched to EE and then switched back to CE, 52% reported return of symptoms and 43% reported new symptoms Andrade et al 66 Mixed-model HMO Before and after Switch of patients from CE 72% of those originally on CE were with members with control to EE; switch program,which switched to EE; those switched group lasted for 18 months, began experienced an excess of 20 office with a newsletter, continued visits per 100 patients in the postswitch with lists of patients who period compared with the preswitch needed to be switched period; at the end of 2 years, 15% of accompanied by phone calls those switched had switched back to from pharmacists, and finally CE. Estimated monthly savings in drug required physicians to com- ingredient costs were $2.43 per patient plete standardized forms to justify the need for conjugated estrogen therapy for the patient to be dispensed conjugated estrogen tablets Good et al 65 Veterans Administration Before and after Switch of patients from 18% of patients initially switched to clinic; study of 704 without control nizatidine to cimetidine; cimetidine were switched to other patients group switches were implemented antiacid agents within 6 months; there by the pharmacy service were no increases in GI-related hoswithout the intervention of pitalizations, clinic visits, or diagnostic individual clinicians tests between the 6-month pretest and posttest periods. Estimated monthly savings on drug costs were $12.60 Benedetto et al 67 Four IPA-model HMOs; Before and after Changes in market share of Market share of fexofenadine increased membership in HMOs with control fexofenadine at 4 HMOs. by 46% at HMO A, by 6% at HMO B, was , , group HMO A used a mandatory by 3% at HMO C, and by 2% at , and switch program;hmo B HMO D, the control HMO mailed letters to physicians and Within 6 months, more than 50% of patients to inform them that switched patients had switched back fexofenadine was the preferred to their original antihistamine at product; HMO C mailed at HMOs B, C, or D letters to physicians only; HMO D had no intervention CE indicates conjugated estrogen; EE, esterified estrogen; GI, gastrointestinal; HMO, health maintenance organization; IPA, independent practice association THE AMERICAN JOURNAL OF MANAGED CARE DECEMBER 2002

9 MCO Influence on Medication Use accounting for implementation costs and changes in related healthcare costs The studies suggest that the success of voluntary programs is highly dependent on the organizational settings in which they occur. Brufsky et al 63 and Baluch et al 64 found that voluntary interchange programs resulted in high rates of switching to the preferred product and substantial savings in staff-model HMOs. The voluntary program evaluated by Brufsky et al also was successful, but much less so, in a group-model HMO. A study of therapeutic interchange programs in 4 independent practice association (IPA)-model HMOs found that voluntary programs achieved little in the way of switching or savings. 67 In comparison, mandatory switch programs were found to be successful in an IPA-model HMO, 67 a mixed-model HMO, 66 and a Veterans Administration outpatient clinic. 65 These studies also indicate that financial incentives may be important. Brufsky et al and Baluch et al point out that physicians in staff-model HMOs may have been willing to voluntarily switch patients to preferred products because they had an economic incentive to do so; most physicians realized that their compensation and staffing levels were tied to the HMO s financial success. 63,64 Physicians in the group-model HMO studied by Brufsky et al were paid by capitation. The only successful program in the study of IPA-model HMOs was a mandatory switch program that provided reimbursement only for the preferred product. 67 Utilization of Therapeutic Interchange Between 38% 2 and 57% 68 of HMOs use therapeutic interchange. About 75% of PBMs offered therapeutic interchange programs in However, the percentages of employer and HMO clients that actually used these services were not reported. About 56% of employer-sponsored plans used therapeutic interchange in As with prior approval programs, the literature does not indicate how extensively therapeutic interchange was used. Use of therapeutic interchange has increased in recent years, but somewhat less than the increase in use of prior approval and restricted formularies. Use has grown from 23% of HMOs in to 38% in Among employer-sponsored plans, use has grown from 36% in to 56% in Surveys of Drug Preference Programs in the Population Table 5 describes studies that examined the impact of drug preference programs in the population These studies examined any type of drug switching that occurred when a MCO refused to pay for the originally prescribed product, but was willing to pay for a therapeutic alternate. Thus, they covered drug switching that resulted from closed formularies, prior approval, and therapeutic interchange. All indicated that the incidence of drug switching is low. The highest estimate indicated that 5.2% of consumers enrolled in MCOs in California had been forced by their health plan to change medications in the past year. 73 The lowest estimate indicated that only 0.25% of third-party prescriptions dispensed in community pharmacies in Virginia were switched. 70 Two studies suggest that drug switching may have negative health consequences for some patients. 70,71 These studies found that between 5% and 11% of consumers chose not to have their prescription filled when faced with a drug switch. That is, they received neither the originally prescribed product nor a therapeutic alternate. It is difficult to know whether this is a significant health problem because the studies provide little detail about the types of products involved. DISCUSSION If MCOs were successful in their attempts to influence prescribing and dispensing decisions, then drug preference programs should be effective when used and they should be used commonly. Further, if drug preference programs are to have favorable effects on costs and quality, their use should not be associated with negative health outcomes. This case cannot be made from the published literature. Although several studies indicate that some MCOs have exerted substantial influence over prescribing and dispensing decisions, the bulk of the literature indicates that most MCOs have not. The literature indicates that closed formularies are effective in decreasing use of nonformulary products. However, it also indicates that closed formularies are used infrequently. In addition, there is some suggestion that implementing closed formularies may result in higher rates of inpatient hospitalization 11 and drug therapy discontinuation. 6,18 Therapeutic interchange and prior approval are used in the majority of health plans, but how extensively they are used within these plans is not known. Further, these interventions have been shown to be effective only in limited circumstances. Prior approval programs have been shown to be effective VOL. 8, NO. 12 THE AMERICAN JOURNAL OF MANAGED CARE 1049

10 Table 5. Studies of the Incidence of Drug Switching in the Population Study Setting and Sample Method of Data Collection Results Carroll community Pharmacy students observed Only 54% of drug switches suggested by third-party pharmacies in community pharmacists payers resulted in a drug switch; of the remainder, Virginia; sample of the third-party payer approved 23% to be dispensed as third-party originally prescribed, patients paid out of pocket to prescriptions receive the originally prescribed drug for 12%, and 11% were not dispensed The rate of prescriptions actually switched was 2.5 per 1000 third-party prescriptions dispensed Schauffler et al insured Telephone survey 5.1% of PPO members, 5.2% of IPA/network HMO Californians members, and 2.4% of staff/group HMO members were forced to change medications by their health plans in the past year Nystrom et al physicians and Mail survey 3% of prescriptions written by physicians and 0.75% 380 community of prescriptions dispensed by pharmacists were pharmacists in switched to a product chemically different from the Virginia one originally prescribed Pyles et al households Telephone survey Only 3.8% of respondents had been asked to accept in Virginia a drug switch in the past year; of those asked, 70% accepted the switch, 11% paid for the originally prescribed drug out of pocket, 5% received no medication, and 12% obtained approval from the insurer to get the originally prescribed drug HMO indicates health maintenance organization; IPA, independent practice association; PPO, paid provider organization. primarily with drug classes that include generic products, which are used to treat symptomatic problems of mild to moderate severity, and in which brand name products have no significant therapeutic advantages over generics. Many therapeutic classes do not meet these criteria. Voluntary therapeutic interchange programs have been shown to be successful in staff-model HMOs, but not in IPA models. One could reasonably assume from this that voluntary programs also would not be successful in preferred-provider organizations or other plans with less organizational control than that provided by staff-model HMOs. Compared with other types of HMOs and health plans, staff-model HMOs enroll relatively few patients. 2 This fact suggests that voluntary therapeutic interchange programs may have little impact on the vast majority of patients covered by prescription drug insurance. Given employers reluctance to implement closed formularies, it seems unlikely that they would subject their employees to mandatory therapeutic interchange programs. The few available studies of the incidence of drug switching provide additional evidence that programs to influence prescribing and dispensing are not widely used, are not highly effective, or both. Estimates of the rate of drug switching resulting from all techniques (ie, closed formularies, therapeutic interchange, prior approval) range from 0.25% of third-party prescriptions dispensed in community pharmacies 70 to 5.2% of consumers in MCOs having to switch medicines in a year. 73 A major limitation of this review is the paucity of research on the effectiveness of drug preference programs. This limitation is especially problematic in light of the number of factors that could influence the effectiveness of these programs. These factors include the type of drug preference program used, the organizational setting in which it is implemented, the patient population affected, the therapeutic use of the drugs involved, and the geographic setting in which the program is implemented. It is rare that more than one study has examined the effectiveness of the same type of drug preference program under the same conditions. This limits the confidence with which one can make conclusions about the effectiveness of these programs. The one case in which there has been replication prior approval programs suffers from another limitation. All studies of prior approval programs have been done under nearly identical conditions THE AMERICAN JOURNAL OF MANAGED CARE DECEMBER 2002

11 MCO Influence on Medication Use The small number of research studies limits the study in another way. There is considerable variability among studies in the way drug preference programs were implemented. For example, studies of formularies included 1 study of the deletion of a single product from the formulary, 2 studies in which several products were deleted from the formulary, and 4 studies in which most or all of the formulary was closed. Because of the substantial differences in the way the formularies were implemented, it is difficult to make inferences about the effectiveness of the typical or average formulary. Because there are only a few studies in each group, it is risky to make inferences about the effectiveness of different ways of implementing formularies. The limited evidence for the effectiveness of drug preference programs is not conclusive proof that they are ineffective. The literature indicates that some programs have been very successful at influencing prescribing and dispensing. Undoubtedly other programs have been successful, but have not been evaluated and reported in the literature. Although the existing research is not adequate to support firm conclusions about the effectiveness of drug preference programs, it is sufficient to conclude that they are used sparingly in most drug benefit plans. I could find no published research on this topic; however, a number of sources have suggested that the more restrictive drug preference programs are not widely used because plan sponsors are reluctant to burden and inconvenience plan beneficiaries Employers seem especially reluctant to use these programs in low-unemployment labor markets. 76 It will be interesting, and a test of this hypothesis, to see whether use of more restricted programs increases as unemployment increases. A related possibility is that MCOs do not implement drug preference programs because of resistance from patients, pharmacists, and physicians. Drug preference programs typically are associated with changes in prescriptions after the prescription has been written. This results in extra work for pharmacists and physicians and in inconvenience for patients. As a result of patient and provider dissatisfaction, a number of states and the federal government have introduced legislation to limit or ban the use of drug preference programs A recent study indicates that MCOs may not use capitation more frequently because of provider resistance. 86 A similar type of resistance could limit the use of drug preference programs. It also is possible that MCOs have tested drug preference programs and found that they were ineffective. However, the managed pharmacy literature 87,88 and the major PBMs 89,90,91 continue to stress the fundamental importance of drug formularies and related activities in controlling drug use. It is unlikely that they would do so if they knew that these programs did not work. FUTURE RESEARCH This review has raised a number of issues for future research. The need for more research on the effectiveness of drug preference programs and the reasons that MCOs do not use them more frequently already have been mentioned. In addition, this review found that increased use of drug preference programs might lead to decreased adherence to drug therapy or to increased use of other healthcare services. This is an important area for future research. Policy makers and payers need to know whether, and to what extent, programs designed to control drug costs result in poorer patient health and increased nondrug costs. Finally, little research has examined consumer preferences for drug preference programs. Because consumer acceptance may be a key to the success of drug preference programs, more needs to be known about their preferences. CONCLUSIONS The employers and health plans that pay for prescription drugs depend on MCOs to control pharmaceutical costs. One of the primary ways that MCOs have attempted to do so has been through influencing physicians prescribing and pharmacists dispensing decisions using formularies, therapeutic interchange, and prior approval. A review of the literature indicates that these tools can control costs when they are implemented, but that they are used infrequently. This suggests that most MCOs, as they currently operate, exert little control over prescribing and dispensing decisions and, consequently, over pharmaceutical costs. The problem is aggravated by several factors. First, the demand for prescription drugs is increasing dramatically. This is due to the aging population, improved diagnostics that result in earlier detection and treatment of diseases, 92,93,94 and new drugs that either treat diseases formerly untreatable with pharmaceuticals or that have milder side effects. 92,93,94 Second, the costs of drugs are rising because new VOL. 8, NO. 12 THE AMERICAN JOURNAL OF MANAGED CARE 1051

12 drugs are substantially more expensive than those they replace. 89,94-96 Even when the new drugs are more effective or have milder side effects, the net result is an increase in the costs of pharmaceuticals. Finally, pharmaceutical manufacturers make substantial efforts to influence prescribing and dispensing decisions. Pharmaceutical companies spent $15.7 billion on promotional activities in In recent years, pharmaceutical companies have begun advertising directly to consumers through television, newspapers, and magazines. Spending on this method of promotion has grown from $0.8 billion in 1996 to $2.5 billion in ,98 Unless MCOs can counteract these trends, there is little likelihood that they can restrain rising spending on pharmaceuticals. One obvious change that MCOs could make is to expand the use of more restrictive drug preference programs such as closed formularies and mandatory therapeutic interchange programs. The most frequently cited reason that MCOs do not use these programs more frequently is payers reluctance to inconvenience consumers. If this is true, then use of more restrictive programs may increase as rising drug costs put increasing pressure on employers and government payers. The growth in the use of more restrictive programs over the past decade provides some evidence that this is already happening. MCOs also might better control costs through developing and implementing other methods of influencing prescribing and dispensing. One promising method is the use of tiered copayments. Over the last few years, the use of this tool has grown dramatically. 23 Tiered copayment programs are designed to allow consumers to use and be reimbursed for any drug product they choose, but to provide financial incentives for them to choose lower cost products. These programs appear to offer substantial promise for controlling costs by influencing drug choice, but have yet to be adequately studied. Whether through more aggressive implementation of drug preference programs or through use of new programs like tiered copayments, MCOs must change the way they operate if they are to successfully control drug costs. REFERENCES 1. Heffler S, Levit KR, Smith S, et al. Health spending growth up in 1999: Faster growth expected in the future. Health Aff. 2001;20(2): Aventis Pharmaceuticals. Managed Care Digest Series 2000: HMO-PPO/Medicare-Medicaid Digest. Bridgewater, NJ: Aventis Pharmaceuticals; Academy of Managed Care Pharmacy. Principles of a sound drug formulary system. Available at: Accessed November 21, Pharmacy Benefit Management Institute, Inc. The Wyeth- Ayerst Prescription Drug Benefit Cost & Plan Design Survey Report: 2000 Edition. Scottsdale, Ariz: Pharmacy Benefit Management Institute, Inc; Novartis Pharmaceuticals. Novartis Pharmacy Benefit Report: Facts & Figures, 2000 Edition. East Hanover, NJ: Novartis Pharmaceuticals; Novartis Pharmaceuticals. Novartis Pharmacy Benefit Report: Trends and Forecasts, 1998 Edition. East Hanover, NJ: Novartis Pharmaceuticals; Harris N. Managing drug formularies. Am Drug. 1995;212: Lipton HL, Gross DJ, Stebbins MR, Syed LH. Managing the pharmacy benefit in Medicare HMOs: What do we really know? Health Aff. 2000(2);19: Motheral BR, Henderson R. The effect of a closed formulary on prescription drug use and costs. Inquiry. 1999;36: Motheral BR, Delate TA, Shaw JA, Henderson R. The effect of a closed formulary in the face of real-life enrollment and disenrollment patterns. J Manag Care Pharm. 2000;6: Lingle EW, Reeder CE, Kozma CM. Impact of an open formulary system on the utilization of medical services. J Res Pharm Econ. 1990;2: Kozma CM, Reeder CE, Lingle EW. Expanding Medicaid drug formulary coverage: Effects on utilization of related services. Med Care. 1990;28: Kreling DH, Knocke DJ, Hamme RW. The effects of an internal analgesic formulary restriction on Medicaid drug expenditures in Wisconsin. Med Care. 1989;27: Smith DM, McKercher PL. The elimination of selected drug products from the Michigan Medicaid formulary: A case study. Hosp Formulary. 1984;19: Soumerai SB, Ross-Degnan D, Gortmaker S, Avorn J. Withdrawing payment for nonscientific drug therapy: Intended and unexpected effects of a large-scale natural experiment. JAMA. 1990;263: Horn SD, Sharkey PD, Tracy DM, et al. Intended and unintended consequences of HMO cost-containment strategies: Results from the Managed Care Outcomes Project. Am J Manag Care. 1996;2: Horn SD, Sharkey PD, Phillips-Harris C. Formulary limitations and the elderly: Results from the managed care outcomes project. Am J Manag Care. 1998;4: Streja DA, Hui RL, Streja E, McCombs JS. Selective contracting and patient outcomes: A case study of formulary restrictions for selective serotonin reuptake inhibitor antidepressants. Am J Manag Care. 1999;5: Schweitzer SO. Differences in managed care drug formularies: What can consumers learn? Med Care Res Rev. 2000;57: Curtiss FR. Drug formularies provide a path to best care. Am J Health Syst Pharm. 1996;53: Blumberg MS. Six HMOs in search of a study: Comments on two papers by Horn et al. Am J Manag Care. 1996;2: Gibaldi M. Drug formularies and health care costs: In God we trust; all others must show data. PORPP Rep. 1996;2: Pharmacy Benefit Management Institute, Inc. The Takeda and Lilly Prescription Drug Benefit Cost & Plan Design Survey Report: 2001 Edition. Scottsdale, Ariz: Pharmacy Benefit Management Institute, Inc; Academy of Managed Care Pharmacy. Common Practices in Formulary Management Systems. Alexandria, Va: Academy of Managed Care Pharmacy; THE AMERICAN JOURNAL OF MANAGED CARE DECEMBER 2002

13 MCO Influence on Medication Use 25. Grabowski H, Mullins CD. Pharmacy benefit management, cost-effectiveness analysis and drug formulary decisions. Soc Sci Med. 1997;45: Teitelbaum F, Parker A, Martinez R, Roe C. Express Scripts 1999 Drug Trend Report. Maryland Heights, Mo: Express Scripts; Aventis Pharmaceuticals. Managed Care Digest Series 2000; Managed Care Trends Digest Bridgewater, NJ: Aventis Pharmaceuticals; Pharmacy Benefit Management Institute, Inc. The 1996 Prescription Drug Benefit Cost & Plan Design Survey Report. Scottsdale, Ariz: Pharmacy Benefit Management Institute, Inc; Smalley WE, Griffin MR, Fought RL, Sullivan L, Ray WA. Effect of a prior-authorization requirement on the use of nonsteroidal antiinflammatory drugs by Medicaid patients. N Engl J Med. 1995;332: Kotzan JA, McMillan JA, Jankel CA, Foster AL. Initial impact of a Medicaid prior authorization program for NSAID prescriptions. J Res Pharm Econ. 1993;5: Phillips CR, Larson LN. Evaluating the operational performance and financial effects of a drug prior authorization program. J Manag Care Pharm. 1997;3: Kotzan JA, Perri M, Martin BC. Assessment of Medicaid priorapproval policies on prescription expenditures: Market-share analysis of Medicaid and cash prescriptions. J Manag Care Pharm. 1996;2: Hoechst Marion Roussel, Inc. Managed Care Digest: HMO- PPO Digest:1995. Kansas City, Mo: Hoechst Marion Roussel; Patel RJ, Gray DR, Pierce R, Jafari M. Impact of therapeutic interchange from pravastatin to lovastatin in a Veterans Affairs Medical Center. Am J Manag Care. 1999;5: Korman L, Borysiuk L. Replacing lovastatin with pravastatin: Effect on serum lipids and costs. Am J Health Syst Pharm. 1995;52: Fugit RV, Resch ND. Conversion of patients from simvastatin to lovastatin in an outpatient pharmacy clinic. Am J Health Syst Pharm. 2000;57: Kinnon AL, Bourne J, Blizzard S, Hunter D, Phillips C. Outcome analysis of a formulary transition from nifedipine to felodipine at a Veterans Affairs Medical Center. J Manag Care Pharm. 1999;5: Krantz SR, Rase RS, Piepho RW. Retrospective analysis of formulary transition at a large metropolitan HMO: Nifedipine GITS to Felodipine ER. J Manag Care Pharm. 1996;2: Clay DR, Bourg MP, Lawrence DB. Outcomes of an amlodipine-to-felodipine therapeutic interchange program. Am J Health Syst Pharm. 2000;57: Mamdani MM, Resig CJ, Stevenson JG. Cost analysis of therapeutic interchange of calcium channel blockers for the treatment of hypertension: Unexpected results from a conversion program. J Manag Care Pharm. 2000;6: Gustin G, White WB, Taylor S, Daragjati C. Clinical outcome of a mandatory formulary switch for dihydropyridine calcium channel blocker therapy at a Veteran s Administration Medical Center. Am J Hypertens. 1996;9: Alexis G, Sessler NE, Conlin PR. Safety and efficacy of a mandatory formulary switch from nifedipine GITS to amlodipine. J Manag Care Pharm. 1999;5: Walters J, Noel H, Folstad J, Kapadia V, White CM. Prospective evaluation of the therapeutic interchange of felodipine ER for amlodipine in patients with hypertension. Hosp Pharm. 2000;35: Lindgren-Furmaga EM, Schuna AA, Wolff NL, Goodfriend TL. Cost of switching hypertensive patients from enalapril maleate to lisinopril. Am J Hosp Pharm. 1991;48: Briscoe TA, Dearing CJ. Clinical and economic effects of replacing enalapril with benazepril in hypertensive patients. Am J Health Syst Pharm. 1996;53: McDonough KP, Weaver RH, Viall GD. Enalapril to lisinopril: Economic impact of a voluntary angiotensin-converting enzymeinhibitor substitution program in a staff-model health maintenance organization. Ann Pharmacother. 1992;26: Hilleman DE, Mohiuddin SM, Wurdeman RL, Wadibia EC. Outcomes and cost savings of an ACE inhibitor therapeutic interchange. J Manag Care Pharm. 1997;3: Cantrell WC, Kimber GH, Morrill GB. Evaluation of blood pressure and adverse effects in patients converted from lisinopril to benazepril. J Manag Care Pharm. 1999;5: Nadel HL. Formulary conversion from glipizide to glyburide: A cost-minimization analysis. Hosp Pharm. 1995;30: Alexis G, Henault R, Sparr HB. Conversion from glipizide to glyburide: Long-term follow-up of a cost-impact survey focusing on the elderly. Clin Ther. 1993;15: Jansen DJ, Resch ND, Hoffman RM. Doxazosin-to-terazosin switch for benign prostatic hyperplasia. Am J Health Syst Pharm. 1999;56: Parra D, Beckey NP, Korman L. Retrospective evaluation of the conversion of amlodipine to alternative calcium channel blockers. Pharmacotherapy. 2000;20: Hilleman DE, Wurdeman RL, Lenz TL. Therapeutic change of HMG-CoA reductase inhibitors in patients with coronary artery disease. Pharmacotherapy. 2001;21: Gerson LB, Hatton BN, Ryono R, et al. Clinical and fiscal impact of lansoprazole intolerance in veterans with gastroesophageal reflux disease. Aliment Pharmacol Ther. 2000;14: Landry FJ, Horwhat JD, Tomich D, Pinski L. Felodipine as an alternative to more expensive calcium antagonists in mild to moderate hypertension. South Med J. 1996;89: Amidon PB, Jankovich R, Stoukides CA, Kaul AF. Proton pump inhibitor therapy: Preliminary results of a therapeutic interchange program. Am J Manag Care. 2000;6: Nelson WW, Vermeulen LC, Geurkink EA, Ehlert DA, Reichelderfer M. Clinical and humanistic outcomes in patients with gastroesophageal reflux disease converted from omeprazole to lansoprazole. Arch Intern Med. 2000;160: Condra LJ, Morreale AP, Stolley SN, Marcus D. Assessment of patient satisfaction with a formulary switch from omeprazole to lansoprazole in gastroesophageal reflux disease maintenance therapy. Am J Manag Care. 1999;5: Oatis GM, Stowers AD. Conversion from amlodipine to felodipine ER: Did the change fulfill expectations? Formulary. 2000;35: Stock AJ, Kofoed L. Therapeutic interchange of fluoxetine and sertraline: Experience in the clinical setting. Am J Hosp Pharm. 1994;51: Jean CD, Triplett JW. Investigating patient experiences after a formulary change. Am J Hosp Pharm. 2000;57: Gardner JS, Scholes D, Baluch WM, Krauss RH. Acceptability of a substitution of estrogen replacement therapy to women enrolled in a health maintenance organization. J Women s Health.1998;7: Brufsky JW, Ross DD, Calabrese D, Gao X, Soumerai SB. Shifting physician prescribing to a preferred histamine-2-receptor antagonist. Effects of a multifactorial intervention in a mixedmodel health maintenance organization. Med Care. 1998;36: Baluch WM, Gardner JS, Krauss RH, Scholes D. Therapeutic interchange of conjugated and esterified estrogens in a managed care organization. Am J Health Syst Pharm. 1999;56: Good CB, Fultz SL, Trilli L, Etchason J. Therapeutic substitution of cimetidine for nizatidine was not associated with an VOL. 8, NO. 12 THE AMERICAN JOURNAL OF MANAGED CARE 1053

14 increase in healthcare utilization. Am J Manag Care. 2000;6: Andrade SE, Gurwitz JH, Cernieux J, Fish LS. Evaluation of a formulary switch from conjugated to esterified estrogens in a managed care setting. Med Care. 2000;38: Benedetto SR, Sloan AS, Duncan BS. Impact of interventions designed to increase market share and prescribing of fexofenadine at HMOs. Am J Health Syst Pharm. 2000;57: Novartis Pharmaceuticals. Novartis Pharmacy Benefit Report: Facts & Figures, 1999 Edition. East Hanover, NJ: Novartis Pharmaceuticals; Marion Merrell Dow. Managed Care Digest HMO Edition Kansas City, Mo: Marion Merrell Dow Inc; Carroll NV. Therapeutic interchange in community pharmacies in Virginia. Am J Health Syst Pharm. 2000;57: Pyles MA, Carroll NV, Holdford DA. Study to determine the impact of the PBM practice of therapeutic interchange on citizens of the Commonwealth of Virginia. In: An Examination of Pharmacy Benefit Manager Practices and Therapeutic Interchange for Citizens in the Commonwealth of Virginia, Final Report of the Virginia Department of Medical Assistance Services, House Document No. 4. Richmond, VA: Commonwealth of Virginia, Nystrom SV, Stover K, Crain WM. Physician and retail pharmacist perceptions of the practice of therapeutic interchange in the Commonwealth of Virginia. In: An Examination of Pharmacy Benefit Manager Practices and Therapeutic Interchange for Citizens in the Commonwealth of Virginia, Final Report of the Virginia Department of Medical Assistance Services, House Document No. 4. Richmond, VA: Commonwealth of Virginia, Schauffler HH, McMenamin S, Cubanski J, Hanley HS. Differences in the kinds of problems consumers report in staff/group health maintenance organizations, independent practice association/network health maintenance organizations, and preferred provider organizations in California. Med Care. 2001;39: Pharmacy Benefit Management Institute, Inc. The Wyeth- Ayerst Prescription Drug Benefit Cost & Plan Design Survey Report: 1999 Edition. Scottsdale, Ariz: Pharmacy Benefit Management Institute, Inc; Merck-Medco Managed Care. Managing Pharmacy Benefit Costs: New Insights for a New Century. Franklin Lakes, NJ: Merck- Medco Managed Care, LLC; Johnson S. Defined contribution is coming, albeit slowly. Manag Healthc. 2000;10: Moran DW. Prescription drugs and managed care: Can freemarket détente hold? Health Aff. 2000(2);19: Rx discounts affected by employer reluctance to hassle members PCS. Health News Daily. 2000;12: Rose JR. Lawmakers have HMOs in their gunsights. Med Econ. 1997;74: Ukens C. Sub attack: Virginia takes aim at drug switching. Drug Top. 1997;141: North Carolina has enacted bills affecting pharmacy practice. AMCP News. August 1999;11: Texas has issued a regulation requiring managed care organizations to notify physicians, providers and enrollees 90 days before a proposed formulary revision. AMCP News. February 1999;11: Gebhart F. Charting a new course: With passage of HMO law, California weighs Rx benefit regs. Drug Top. 1998;142: New Jersey bans closed formularies. Drug Benefit Trends. 2001;13: Academy of Managed Care Pharmacy. Ohio HMO reform law covers state-regulated health plans. Weekly News. November 13, Chernew M, Cowen ME, Kirking DM, et al. Pharmaceutical cost growth under capitation: A case study. Health Aff. 2000(6);19: Wertheimer AI, Navarro RP. Managed Care Pharmacy: Principles and Practice. Binghamton, NY: Haworth Press, Inc; Navarro RP. Managed Care Pharmacy Practice. Gaithersburg, MD: Aspen Publishers, Inc; Teitelbaum F, Martinez R, Parker A, et al. Express Scripts 2000 Drug Trend Report. Maryland Heights, Mo: Express Scripts; AdvancePCS. What we do: Advance solutions. Available at: Accessed August 3, Merck-Medco Managed Care, LLC. The Merck-Medco Drug Trend Report. Franklin Lakes, NJ: Merck-Medco Managed Care, LLC; Dubois RW, Chawla AJ, Neslusan CA, Smith MW, Wade S. Explaining drug spending trends: Does perception match reality? Health Aff. 2000;19: Kleinke JD. Just what the HMO ordered: The paradox of increasing drug costs. Health Aff. 2000;19(2): Berndt ER. The US pharmaceutical industry: Why major growth in times of cost containment? Health Aff. 2001;20(2): RxHealthValue. Prescription drug expenditures increase more than 24%. Available at: docs/research_ pdf. Accessed November 26, Findlay S. Prescription Drug Expenditures in 2000; The Upward Trend Continues. Washington, DC: National Institute for Health Care Management Research and Educational Foundation; Kreling DH, Mott DA, Wiedenholt JB, Lundy J, Levitt L. Prescription Drug Trends A Chartbook Update. Menlo Park, Calif: The Henry J. Kaiser Family Foundation: Adams C. FDA plans to review policy allowing direct-toconsumer drug ads for TV. The Wall Street Journal Online. March 1054 THE AMERICAN JOURNAL OF MANAGED CARE DECEMBER 2002