SPECIAL SUPPLEMENT TO M A N A G E D Care Recent Advances in Care: Treatment of Acid-Related Disorders HIGHLIGHTS • The Changing Landscape of Health Care • Current Treatment Trends in Gastric-Reflux Disorders • Effective Approaches to Acid Suppression • A Comparison of Proton Pump Inhibitors Volume 10, No. 10 October 2001 M A N A G E D Care Editor JOHN A. MARCILLE Managing Editor MICHAEL D. DALZELL Senior Editor FRANK DIAMOND SPECIAL SUPPLEMENT M A N A G E D Care October 2001 Senior Science Editor PAULA SIROIS Senior Contributing Editor PATRICK MULLEN Contributing Editors BOB CARLSON JOHN CARROLL DAVID COLEMAN, JD JEFFREY J. DENNING MIKE FOLIO, JD MICHAEL LEVIN-EPSTEIN JACK MCCAIN KAREN TRESPACZ, JD Recent Advances in Care: Treatment of Acid-Related Disorders Health Care in the United States: Current and Future Challenges ..............................................2 Ronald T. Luke, PhD, JD Design Director PHILIP DENLINGER Editorial Advisory Board Chairman ALAN L. HILLMAN, MD, MBA Senior Fellow Center for Health Policy Leonard Davis Institute of Health Economics University of Pennsylvania, Philadelphia Impact of Acid-Related Disorders in the United States.................................................................7 Michael Brown, MD, and Russell D.Yang, MD, PhD Comparative Pharmacology of Proton Pump Inhibitors....11 Joseph A. Barone, PharmD, and John R. Horn, PharmD Group Publisher TIMOTHY J. STEZZI Publisher Optimizing Acid-Suppression Therapy ...............................17 TIMOTHY P. SEARCH, RPH Jeffrey L. Barnett, MD, and Malcolm Robinson, MD Midwest Sales Manager TERRY HICKS SCOTT MACDONALD Evidence-Based Health Care: Making Health Policy and Management Decisions............22 Director of Production Services Nimish B.Vakil, MD Eastern Sales Manager WANETA PEART MANAGED CARE (ISSN 1062-3388) is published monthly by MediMedia USA Inc. at 275 Phillips Blvd., Trenton, NJ 08618. This is Volume 10, Issue 10. Periodicals postage paid at Trenton, N.J., and at additional mailing offices. POSTMASTER: Send address changes to MANAGED CARE, MediMedia USA, 275 Phillips Blvd., Trenton, NJ 08618. Prices: $10 per copy, $93 per year in the USA; $120 per year elsewhere. Send letters to the editor c/o Frank Diamond, MANAGED CARE, 275 Phillips Blvd., Trenton, NJ 08618. Letters may be edited for length and clarity. E-mail: [email protected]. Phone: (609) 671-2100; fax (609) 882-3213; circulation inquiries, (609) 671-2100. Copyright ©2001 MediMedia USA Inc. This MANAGED CARE Special Supplement is derived from the proceedings of recent managed care summit meetings and is supported by educational grants from Eisai Inc. and Janssen Pharmaceutica Inc. The opinions expressed herein are those of the participants and faculty, and do not necessarily reflect the views of MediMedia USA Inc., or the publisher, editor, or editorial board of MANAGED CARE. Health Care in the United States: Current and Future Challenges RONALD T. LUKE, PHD, JD Research & Planning Consultants, LP, Austin, Texas P rior to 1980, the paradigm of health care in this country was based on indemnity health plans. Insured patients could choose freely among available providers. Insurers rarely questioned physicians’ decisions. Care was paid for on a fee-forservice basis with providers largely determining the fees. Few health insurance companies saw their job as managing care. Most saw their job as paying the bills. Yet the 1980s witnessed significant changes. The rapid growth of health care costs and insurance premiums generated pressure from purchasers of health care, including both private payers and the government, to slow the rate of growth. Through the ’80s, health care spending continued to rise as a percentage of the gross domestic product (GDP), and by 1990, health care was taking about 12 percent of the total economy. Health care spending was increasing at double-digit rates.1 U.S. health care spending was at about 12 percent of its GDP, while in other developed countries it was about 8 percent.2 Yet U.S. health outcomes were not significantly better; median life expectancy was about the same, and the infant mortality rate was substantially worse.2 Managed care was offered as the way to reduce costs. While managed care organizations (MCOs) were originally formed in the 1940s, their national growth began with the passage of federal health maintenance organization legislation in 1973. Rapid growth of managed care enrollment nationwide occurred in response to the rapid increases in health care costs in the 1980s. Although quality-improvement issues were part of the arguments in favor of managed care, the primary motivation for employers and government to adopt managed care was to contain expenditures. Author correspondence: Ronald T. Luke, PhD, JD President Research & Planning Consultants, LP 7600 Chevy Chase Drive, Suite 500 Austin,TX 78752 Phone: (512) 371-8100 Fax: (512) 371-0327 2 MANAGED CARE / SUPPLEMENT What happened? Managed care enrollment increased dramatically during the 1990s. By the late 1990s, about 85 percent of the commercially insured population participated in some form of managed care (e.g., HMO, preferred provider organization, point-of-service plan). In addition, Medicaid recipients were being moved to managed care settings and there was growth in Medicare managed care. Managed care slowed the growth of health care costs considerably. The annual inflation rate for health care plummeted in the early 1990s, from 4.5 percent above the general rate of inflation down to about 1 percent.3 During the 1990s, there was a change in the relationship between the rates of growth of the economy and the growth of health care expenditures (Figure 1).1 In the middle and latter parts of the decade, the two came together and, in some years the rate of growth of the economy actually exceeded that of health care expenditures. This caused health care spending as a percentage of the economy to plateau, or even fall slightly. By the late 1990s, decreasing health care inflation rates and a strong economy reduced the pressure on government and employers to control health care costs. Other issues assumed higher business and political priority. How did managed care reduce costs? There are only two ways to reduce health care costs: reduce the amount paid per unit of service* and reduce the per capita utilization of services. Most of the reduction in health care costs by managed care plans in the 1990s was due to the negotiation of lower payment rates to health care providers. The development of new payment systems by Medicare for hospitals and for physicians assisted managed care plans in negotiating lower prices. The Medicare prospective payment system (PPS) replaced cost-based reimbursement for inpatient hospital services beginning in 1983. The Medicare Resource Based Relative Value System (RBRVS) replaced “usual, customary, and rea* This can be accomplished by reducing the price for individual services and by shifting from more-expensive to less-expensive services to accomplish the same purpose. Both affect the weighted average price per unit of service. sonable” (UCR) reimbursement for physician services beginning in 1992. Medicare became a benchmark for MCOs in their dealings with providers. Some plans have adopted the Medicare fee schedules as a pricing formula. Depending on a plan’s market power, some plans have been able to negotiate contracts at less than 100 percent of Medicare rates. Even when a different pricing formula is used, Medicare pricing serves as an important benchmark. Yet, in recent years, limits on plans’ ability to achieve price reductions seem to have been reached. For many reasons, health care providers have been able to negotiate more substantial price increases in the past few years. These reasons include consolidation of hospital systems, creation of independent practice associations to negotiate for physicians, reduced ability of plans to guarantee volume to health care providers due to more inclusive networks, and increased experience in contract negotiation among health care providers. Utilization management measures by MCOs include, among others, preauthorization, restrictions on direct access to specialists, and selection of physicians with conservative practice styles. While utilization management measures by MCOs have been the focus of political attention, they have never been the main source of cost reductions. Most of the impact of utilization management has been on the use of inpatient hospital services. There has been some impact on the use of physician specialists. Yet the low copayments of managed care plans for office visits to primary care physicians and for prescription drugs may have increased utilization of these services. Where do we stand today? cians and patients. By the late 1990s, health care providers had won the battle for public opinion and had built a public sense of entitlement to unlimited health care regardless of cost and regardless of contract. The backlash from physicians stemmed from reductions in their income and from the constraints MCOs put on their professional autonomy. While most patients seem to be satisfied with their level of care, there was a dissatisfied minority of patients, a dissatisfied majority of physicians, and a few real horror stories to gain the attention of elected officials. In the booming economy of the 1990s, control of health care costs seemed less important politically. The political goals increasingly became patient choice and restricting the mechanisms managed care plans could use to control costs. This resulted in state and federal legislation for mandated benefits and patient bills of rights. The number of mandates passed by states increased dramatically during the past five years (Figure 2).4 These mandates include any-willing-provider legislation, direct access to specialists, the right to sue health plans and employers, provider antitrust exemptions, expanded definitions of medical necessity, and benefit mandates. While each of these mandates contributes just a fraction of a percentage point in increased costs, together they ultimately lead to an increase in the amount charged for health insurance. For example, in Texas, one of the states with the most mandates, mandates account for an additional increase of health care premiums of between 7 percent and 8 percent. While some of this legislation provides genuine protection for patients, much of it is better seen as protection for health care providers. None of the legislation provides any benefits for people who lose health insurance coverage as a result of higher premium costs. Growth rate (%) Managed care has undoubtedly imposed inconveniences on health care providers and on some patients. Nevertheless, by FIGURE 1 Gross domestic product and health care expenditure any objective measurement, the qualannual growth rates, 1990–1999. ity of health care delivered through MCOs is as good or better than that Health care expenditures 12 delivered through unmanaged inGross domestic product demnity plans. The percentage of en10 rollees satisfied with their health plans has also remained high. Deliv8 ering the same quality of care for a lower cost is good for society as a 6 whole. At the margin, more people will have health insurance when pre4 miums are stable or declining. Fewer people will have health insurance 2 when premiums rise. Nevertheless, managed care has 0 become the villain in the minds of 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 many consumers and elected officials as a result of backlash from physi- Adapted from CMS.1 SUPPLEMENT / MANAGED CARE 3 What issues do we face as a nation? predict health care expenditures will rise about 8 percent annually over the next few years and then fall to about 7 percent per year toward the end of the decade. These increases would occur when the economy is projected to grow at about 3 percent (Figure 3).1 So we will again see health care expenditures rise as a percentage of GDP. Health care costs will be a growing problem in the next 30 years, as more individuals reach age 65. Today, 13 percent of the population is 65 years or older. In 2030, that figure will rise to 20 percent.5 Medicare projections from the Centers for Medicare and Medicaid Services, (CMS, formerly HCFA) place per capita spending at about $5,700 by the year 2008. Research & Planning Consultants LP projects that with current trends, the figure will be about $9,000 in 2030 (in year 2000 dollars). If accurate, that level of spending would increase health care spending from about 13 percent to about 22 percent of the GDP. These figures are driven not only by the proportion of the population that is older than 65 years, but also by the growth rate of health care FIGURE 2 Number of mandates passed by states, 1990–1998. expenditures per capita in all age groups. Although significant changes in the U.S. health care system occurred in the mid-1990s, many changes were undone in the last part of the decade, putting us close today to where we were in 1990 (Figure 3).1 In 1990, per capita spending on health care was about $2,000. Today, it is about $4,700. U.S. spending per capita is still more than twice the median for other developed countries,2 as is spending as a percent of national GDP (Figure 4). Yet our infant mortality, life expectancy at birth, and life expectancy at age 65 are worse or not significantly better than other developed countries. So our increased spending is not justified by better health outcomes. This time, though, public opinion is against the precepts of managed care rather than in favor of them. We are again seeing acceleration in health care expenditures. Recent quotations for health insurance premiums are reported to be 10 to 30 percent higher than in the prior year. Projections from the Social Security Administration 100 86 Number of mandates 90 80 70 60 53 50 40 38 35 27 30 20 10 0 1990 1992 1994 1996 1998 Adapted from Seward and Henderson4 with permission. Annual increases (%) FIGURE 3 Health care spending during the last decade. 10 9 8 7 6 5 4 3 2 1 0 What does this mean to the nation as a whole? Essentially, if we repeat the experience of the last two decades, we will have health care expenditures growing more rapidly than the economy, leading us back to all the problems we experienced at the end of the 1980s. And as we spend more on health care, dollars are limited for other expenditures that also affect our quality of life. Future health care expenditures may crowd out spending on other social priorities such as education, research and development, and business investment in capital goods. For example, the percentage of the GDP spent on education has not grown since 1970. Where do we go from here? 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Adapted from CMS.1 4 What is the future? MANAGED CARE / SUPPLEMENT The demand for health care services is essentially unlimited, as long as someone else is paying for them. There are four ways to control percapita health care expenditures: (1) restrict who has access; (2) restrict the treatment; (3) restrict how much care is delivered; and (4) restrict how much providers are paid. Health care at any cost, almost regardless of benefit to the patient, is gaining status as Percentage of GDP spent on health care a political right. To the extent this oc- FIGURE 4 Health care spending as a percentage curs, market forces are less effective of gross domestic product. and less acceptable as a means of lim16 iting expenditures. The alternative is a mix of market mechanisms and raUnited States 14 tioning by government. New Zealand Canada While explicit rationing of health Sweden Spain 12 care is considered politically taboo, Germany Portugal Switzerland France our current systems implicitly ration Greece 10 Netherlands care. 6 The publicly insured (e.g., Iceland 8 Czech Medicaid, children’s health insurance Denmark Republic Australia program) have access to a very broad Norway 6 range of services but may have limJapan Ireland Italy Poland Belgium ited access to providers. They may Mexico Korea Finland 4 Austria have to wait for care. Those privately Turkey United Kingdom insured have access to a broad range 2 of services and providers. The uninsured have very limited access to 0 emergent and urgent services from a 0 5 10 15 20 25 30 limited set of providers. And if you Per capita GDP (thousands of dollars) are older than 65, you have essentially 2 unlimited access to care through Adapted from OECD Health Data. Medicare with no significant probtal peer-review systems to require best practices and lem of access to providers. quality of care for inpatients is questionable. At this time, the Oregon Health Plan is the only pubThis creates two issues: First, people resent MCOs lic health program in the United States to explicitly adoverruling their individual physicians. The public has dress rationing of treatment paid by a health plan. Their been led to believe that personnel other than physicians system recognizes that while access to health care is a within MCOs are overruling their doctors. In fact, the basic right, there is a limit to the amount of money the laws in most states actually allow only physicians republic sector can devote to health care, and that limited tained by MCOs to disapprove requested treatment. Secresources necessitate choices about what treatments and ond, giving sole authority to the physician translates to conditions can be covered. At present, this approach apno review of his or her work, no preauthorization of pears to be the most logical basis for rationing and is contreatment to ensure that care is standard and approprisistent with increased emphasis on evidence-based medate, and no comparison of practice profiles of physicians icine. to ensure that each individual physician is practicing acWhile rationing can address some of the equity quescording to the quality standards set by the experience of tions as to how health care is provided, it does not fully a larger group. If managed care plans are not reviewing address questions of efficiency, efficacy, and quality of physician decisions and asking why best practices are care. Physician practice patterns are the most important not being used, who is? factor in all three concerns. Because physicians control The focus of state and federal governments on en85 to 90 percent of health care expenditures, changes hanced confidentiality of patient records may interfere within the medical profession have the potential to alter with the ability of MCOs to exchange the type of inforper-capita health care expenditures. Changes may involve mation needed to monitor care, to carry out disease how many physicians are trained, how they are paid, management, and to monitor accountability of and reand where they practice. sults by individual physicians. If managed care plans are We need to change the paradigm that glorifies indinot looking at the overall course of treatment and vidual physician autonomy to a systems paradigm that changes in the patient’s health status, who is? emphasizes evidence-based medicine and outpatient A political battle is inevitable over the scope of the quality-assurance systems. The lay public believes that health care entitlement. Can a health insurance plan the physician knows best, but the average patient has no limit its offering to a basic package of treatments for way to judge whether a physician is following best praccertain conditions established, like the Oregon plan, tices (doing the right thing) and is providing highthrough cost/benefit analysis and a budget constraint? quality services (doing the right thing right). Outside the People who want more can buy private supplemental inhospital, there is no peer review of physicians except surance. Alternatively, will the courts and the legislature that done by managed care plans. The ability of hospi- SUPPLEMENT / MANAGED CARE 5 say that no matter what the policy says, any insurance plan must pay for all treatment that could have any benefit to the individual regardless of cost or cost/benefit relationships? Increasingly health care services are improving quality of life, not simply maintaining life or basic functions. While everyone has an equal right to life, we do not recognize a right to equal quality of life. There has never been a right to equal access to other necessities — housing, food, transportation, etc. — beyond a basic level. A large policy issue will be defining which health care services are basic necessities and hence an entitlement, and which are discretionary expenditures to enhance quality of life. If commercial insurers cannot write limited policies and rely on regulators and courts to respect those limits, then premium rates must be set to cover the risks of unexpected obligations. This prices more people out of the private insurance market and increases the problem of a large and growing uninsured population. Managed care will survive the public backlash evident today, but in a weakened form. As the economy weakens, and health care costs increase, there will be a renewed interest in reducing expenditures. Younger individuals in this country have grown up within a managed care environment, both as patients and physicians, and accept the practices and general restrictions on care. As older people move out of the system, resistance will begin to dissipate. However, in the current legislative and litigation climate, health insurance plans will, on balance, retreat from managing health care. If managed care is not the answer to coming increases in U.S. health care costs, what is? Some favor restoring market discipline to individual health care decisions by medical savings accounts. Others favor a single-payer health care system like all other industrialized nations. Both sides have strengths and weaknesses. Neither side believes commercial managed care plans are the correct answer, nor have they determined how to balance physician autonomy with ensuring that patients are treated with best practices applied correctly. Just as Medicare reimbursement systems became benchmarks for commercial insurance, the results of efforts to reform the Medicare system are likely to be our first indication of the answer the U.S. political system will choose. It is not yet clear whether Congress will enact a broad reform, or whether it will proceed with incremental changes to avoid short-term problems. A serious “train wreck” between health care spending and other federal budget priorities may have to occur before there is the political will for broad reforms. ployer-sponsored health plans. To a lesser degree, managed care approaches were adopted for some Medicaid and Medicare enrollees. By the mid-1990s, new Medicare reimbursement policies and the widespread adoption of managed care plans had substantially reduced the growth rate of health care expenditures. All published studies show that HMOs and other managed care plans have delivered health care of equal or better clinical quality to traditional indemnity insurance plans at a lower cost. Yet as the economy strengthened during the latter ’90s, concerns about overall health care costs lessened, and the public became less willing to accept restrictions on the enrollee’s choice of physician and the physician’s treatment choices. Public opinion turned against the concept of managed care as a result of backlash from both physicians and consumers. Government mandates altered the ability of managed care to deliver on the promise of higher quality at a reduced cost. Today, health care costs are again rising. The United States is spending significantly more per capita on health care than are other developed nations, with no demonstrable improvement in health outcomes. Estimates suggest that in the next 30 years, health care costs will again rise at a rate faster than that of the economy. Further, rising costs will reflect the impact of an aging population, with more than 20 percent of individuals 65 years or older by the year 2030. Thus, it appears we are again headed for a crisis. However, the environment has changed: As a result of the experience of the last 10 years, the public is now less willing to accept changes to the health care system, and the concept of managed care has negative connotations. Consequently, the crisis may potentially be even worse than that we experienced a decade ago if we are unable to find ways to control health care expenditures. Summary 5. The last time the United States experienced rapid increases in health care expenditures in the 1980s, the solution was widespread adoption of managed care by em- 6 MANAGED CARE / SUPPLEMENT References 1. 2. 3. 4. 6. Centers for Medicare and Medicaid Services (formerly HCFA). Office of the Actuary. National Health Expenditures and Selected Economic Indicators, Levels and Average Annual Percent Change: Calendar Years 1980–2001. ‹‹http://www.hcfa.gov/stats/NHE-Proj/ proj2000/tables/t1.htm.›› Accessed: Sept. 14, 2001. Organization for Economic Cooperation and Development (OECD). OECD Health Data 2001: A Comparative Analysis of 30 Countries. OECD Health Policy Unit, Paris, France. Bureau of Labor Statistics. Consumer Price Index (CPI) Annual Rates of Health Inflation. Bureau of Business and Economic Research.‹‹http://www.bber.umt.edu/ healthcare/consumerprices.htm.›› Accessed: Sept. 14, 2001. Seward JA, Henderson JW. Report on the Cost of Health Care System Mandates. Texas Association of Business and Chambers of Commerce, Baylor University (Austin, Texas, January 1999). U.S. Census. ‹‹http://www.census.gov/population/ estimates/nation.›› Accessed: Sept. 14, 2001. Fuchs VR. Who Shall Live? Health, Economics, and Social Choice. River Edge, NJ: World Scientific Publishing Co. 1998. Impact of Acid-Related Disorders In the United States MICHAEL BROWN, MD, AND RUSSELL D. YANG, MD, PHD Rush-Presbyterian-St. Luke’s Medical Center, Chicago, and Keck School of Medicine at the University of Southern California, Los Angeles A cid-related disorders, which include duodenal ulcer disease, gastric ulcer disease, Zollinger-Ellison syndrome, and gastroesophageal reflux disease (GERD), are caused by an imbalance between mucosal defense mechanisms and acid secretions. In the United States, the lifetime prevalence of peptic ulcer disease (including duodenal ulcer and gastric ulcer) is about 10 percent.1 Estimates reveal that about 44 percent of adults have reflux symptoms, from occasional heartburn to severe GERD.2 The true prevalence of acid-related disorders may be higher, as many who experience symptoms of these conditions are likely to self-medicate and not seek the advice of a physician. Gastric, duodenal, and esophageal mucosa can all be damaged by acid. In GERD, damage to the esophageal mucosa results from excessive reflux of gastric contents into the lower esophagus, principally due to abnormally frequent relaxation of the lower esophageal sphincter (LES). Reflux disease is a common, chronic disease. Esophagitis is a complication of reflux that develops when mucosal defenses that normally counteract insults on the esophagus succumb to the onslaught of acid pepsin. Esophageal erosions or ulceration develop in about half of individuals with symptomatic GERD.3 Barrett’s esophagus occurs as the lining of the distal esophagus is progressively replaced with epithelium more resistant to peptic digestion. This epithelium is more prone to malignant transformation, thus there has been increasing concern about development of esophageal cancer in individuals with long-term, severe GERD. Peptic ulcer is a lesion in the stomach or duodenum. The major forms of peptic ulcer are duodenal ulcer and Author correspondence: Michael Brown, MD Rush-Presbyterian-St. Luke’s Medical Center 1725 West Harrison, Suite 339 Chicago, IL 60612 Phone: (312) 942-5677, ext. 239; Fax: (708) 848-2901 E-mail: [email protected] gastric ulcer, both of which are chronic diseases that are often caused by Helicobacter pylori infection but also associated with the use of aspirin and nonsteroidal antiinflammatory drugs (NSAIDs). Patients with duodenal ulcer secrete more gastric acid than do healthy individuals, and their overall 24-hour intragastric pH may be lower.4 Gastric ulcer is also associated with the imbalance between reduced gastric mucosal defense factors and gastric acid.4 H. pylori, common in both conditions, is a very prevalent organism. Approximately 20 percent of the population has the organism, but only about 1 percent to 2 percent of the population will develop a peptic ulcer.1 H. pylori is found in up to 90 percent of duodenal ulcer patients and 70 percent of gastric ulcer patients.4 A gastrin-secreting tumor causes a rare disorder, Zollinger-Ellison syndrome. Excessive gastrin causes gastric acid hypersecretion and may cause diarrhea, malabsorption, and complicated ulcer disease. This syndrome has a national prevalence of 0.1 to 3.0 per million.4 Those with Zollinger-Ellison syndrome also have a high prevalence of esophagitis due to gastric hypersecretion.3,4 Pathology of acid-related disorders Figure 1 shows the spectrum of upper gastrointestinal (GI) symptoms. 5 Ulcer-like symptoms are associated with H. pylori infection, chronic use of aspirin and/ or NSAIDs, stress, ischemia, or acid hypersecretion. Dysmotility-like symptoms tend to be related to gastric or small-bowel abnormalities. GERD-like symptoms are related to abnormal presence of acid, pepsin, and bile acids in the distal esophagus, complicated by esophageal dysmotility, gastroparesis, and abnormal tissue factors in the esophagus. Due to more frequent LES relaxations, and in some cases, reduced pressure in the LES, individuals with GERD generally have more episodes of acid reflux than do healthy subjects. The GI contents are most likely to reflux when gastric volume is increased after meals, when the gastric contents are located near the gastroesophageal junction (such as when the patient is bending down), or when gastric pressure is increased (i.e., with obesity, SUPPLEMENT / MANAGED CARE 7 lowing acid and bile salts to enter. Pepsin, a proteolytic enzyme activated when acid facilitates the release of pepsinogen, destroys the tissue in the esophagus. At a pH of 1.3 to 2.3, pepsin increases the severity of mucosal damage and produces inflammation.7 If the pH is above 2.3 to 2.5, the activity of pepsin is reduced, leading to a reduction in damage (Figure 3).8 Peptic activity is markedly reduced at a pH of 4.0 and absent at a pH >5.0; pepsin is denatured and therefore out of the system at a pH >7.0. The cycle is predictable — when protection is lost, injury results; with injury, there is increased sloughing of cells, increased epithelial turnover, and a greater proportion of immature epithelium. Immature epithelium allows more back-diffusion of acid and bile salts, which then stimulate the nervous system, causing the cognition of pain. In short, the acid breakdown of tissue defenses leads to the production of symptoms, erosions, ulceration, and other complications such as bleeding and stricture formation.9 The immature epithelium found in the distal esophagus after injury may have increased potential for malignant transformation, i.e., progression to Barrett’s esophagus and adenocarcinoma. A reFIGURE 1 Spectrum of overlapping gastrointestinal symptoms. cently published population-based study in the New England Journal of Medicine Dysmotility-like from Sweden used personal interviews Ulcer-like Early satiety Epigastric pain on the subjects’ history of GERD. RePostprandial bloating Nocturnal searchers concluded that patients with Nausea/vomiting Fasting reflux have a fivefold increased risk of Pain unrelieved by Relieved by esophageal cancer.10 For patients with food/antacids food/antacids moderate or severe reflux symptoms ocBleeding (overt/occult) curring more than three times weekly and persisting for >20 years, the esophageal cancer risk increased significantly. There was no indication in the GERD-like report whether patients with reflux were Heartburn treated with a proton pump inhibitor Regurgitation (PPI). This preliminary study suggests Chest pain that esophageal cancer risk in those with reflux may be an issue and that control Adapted from Talley et al.5 with permission. or moderation of symptoms may be imFIGURE 2 Failed peristalsis is most prevalent in severe esophagitis. portant. Recent studies that were conducted 50 in Sweden have indicated major risk fac40 tors for the increasing incidence of esophageal adenocarcinoma in associa30 tion with reflux disease.11-13 In casecontrol studies, obesity (particularly a 20 body mass index >30 kg/m2) and gastroesophageal reflux were seen as strong 10 risk factors for developing esophageal cancer.11,12 In combination, obesity and 0 Healthy NERD Mild Severe reflux symptoms entailed highly inesophagitis esophagitis creased risk estimates.11 Another study 6 NERD = nonerosive reflux disease. Adapted from Hogan with permission. noted the relationship between the Patients (%) pregnancy, or binding clothing). The contact of this acid with the esophageal epithelium may be prolonged by abnormal esophageal motility, causing ineffective clearance. The abnormal peristalsis seen in those with GERD is typically more marked in individuals who have severe esophagitis (Figure 2).6 Patients with GERD have delayed gastric emptying, indicating that acid is not cleared from their stomach as rapidly as it is in individuals without reflux disease. The diminished clearance of acid, bile, and food from the stomach allows a greater pool for reflux and increases intraabdominal pressure. Prolonged contact of acid with the epithelium is central to the development of GERD, which develops when gastric acid overwhelms the intrinsic epithelial defenses and repair mechanisms. Unlike the stomach, which has very tight junctions between the cells and, therefore, does not allow bile salts or acid to diffuse into the mucosa and cause damage, the esophagus is lined by a squamous tissue with gap junctions between the cells. Protection against acid is conferred by overlapping of the cells. When the distal esophagus becomes damaged, however, the cells shrink, pulling apart the junctions and al- 8 MANAGED CARE / SUPPLEMENT increasing incidence of esophageal adenocarcinoma and the use of LES-relaxing medications. These medications included nitroglycerin, anticholinergics, beta-adrenergic agonists, aminophyllines, and benzodiazepines. The association was especially strong for anticholinergics.13 the concept of empiric PPI therapy in place of diagnostic endoscopy, which is a more expensive option. Endoscopy may be indicated in a patient who has dysphasia or difficulty swallowing, painful swallowing, anemia (particularly in the elderly), some weight loss, or other signs of bleeding or chest pain; other patients can be treated initially with a PPI. If the symptoms are relieved by the PPI, the individual is most likely suffering from GERD. Options in treating GERD include histamine-2 receptor antagonists (H2RAs), prokinetic agents, PPIs, or a combination of these agents. A review of the literature indicates that the available therapies for treating this disease demonstrate a wide range of effectiveness. Symptoms in acid-related disorders Treating reflux disease Studies in the literature report that reflux disease has a significant impact on quality of life as assessed by physical pain, mental well-being, and social interaction. Therefore, GERD should be treated aggressively to relieve symptoms and prevent potential long-term complications. Although patients present with a number of overlapping symptoms, physicians have found that a patient with two or more reflux-like complaints is usually suffering from reflux disease. This finding has led to Mean amplitude of esophageal contraction (mm Hg) Esophageal score Although epigastric pain is the primary symptom of peptic ulcer disease and heartburn is the primary symptom associated with GERD, patients often present with overlapping symptom complexes. Symptoms of ulcer disease and GERD may overlap with symptoms of nonulcer dyspepsia, a condition caused by abnormal upper GI motility associated with altered visceral sensation (Figure 1).5 To make the diagnosis, the physician listens to the patient’s descrip- FIGURE 3 Relationship of pH and esophageal damage in a canine model. tion of symptoms and considers the location of the pain, its character, and its rela4.5 tionship to meals. It is also useful to 4.0 determine whether eating or antacids give 3.5 relief, as is common in peptic ulcer disease. 3.0 Ulcer-like symptoms include epigastric pain (relieved by food or antacids) and bleeding, 2.5 either overt or occult. Dysmotility-like 2.0 symptoms include early satiety, postpran1.5 dial bloating, nausea or vomiting, and pain unrelieved by intake of food or antacids. 1.0 GERD-like symptoms include heartburn, 0.5 regurgitation, and chest pain. Reflux symp0.0 toms are insidious in that they come on 0 0.5 1.0 1.5 2.0 2.5 gradually, tend to persist over time, and pH force patients to change their lifestyle to accommodate the condition. There is no cor- Adapted from Goldberg et al.8 with permission. relation between the severity of heartburn and the underlying severity of the disease FIGURE 4 Number of swallows to clear bolus of food significantly greater in patients with reflux (i.e., esophagitis). Patients with GERD have compared with healthy subjects. a difficult time swallowing food. Whereas most healthy individuals will clear a bolus Healthy subjects 120 of food in one to two swallows, individuals Reflux patients with GERD will require significantly more 110 swallows (Figure 4).14 100 90 80 70 60 50 Correlation coefficient = 0.82 P<0.01 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 Number of swallows Adapted from Phaosawasdi et al.14 with permission. SUPPLEMENT / MANAGED CARE 9 The enzyme responsible for secretion of hydrogen ions from parietal cells into the gastric lumen, hydrogenpotassium adenosine triphosphatase, is called the proton pump. Compounds that bind this enzyme, known as PPIs, have a profound inhibitory effect on gastric acidity. Currently available PPIs include omeprazole, lansoprazole, pantoprazole, esomeprazole, and rabeprazole. Results of research in the last 10 to 15 years suggest that the most effective therapy for GERD is a PPI twice or three times daily and that the least effective treatment is to use over-the-counter medicines, including antacids and low-dose H2RAs. The PPIs are effective in GERD because they eliminate acid, thus preventing the release of pepsin and consequent damage to the esophageal mucosa. Effectiveness in treating acid-related disorders may be assessed through measurements of acid suppression, including the degree, the duration (percentage of a 24hour period), and the length of treatment. An increase in any or all of these parameters results in an increase in the proportion of mucosal healing at any time point.15 Curing of duodenal ulcer requires steady acid suppression throughout the day. A large study conducted with 1,000 patients found that the pH must reach 3.0 for about 16 hours each day to achieve a 90 percent healing rate.15 To heal GERD, the critical pH is 4.0. Due to rising concern about the risk of Barrett’s esophagus and esophageal cancer in those with long-term reflux disease, physicians and patients have shown less concern about harmful effects of long-term acid suppression with PPIs. Almost every algorithm for treating reflux disease suggests the use of step-up therapy. A step-up protocol recommends starting with the least expensive therapy and then modifying treatment according to the patient’s response. If the therapy is ineffective, the next more expensive therapy is tried. If that therapy fails to provide relief, the next therapy is tried; the patient continues through the list until an effective therapy is found. While such treatment initially keeps medication costs down, it ultimately leads to increases in both direct and indirect costs of caring for the patient, because therapy with less expensive options, e.g., H2RAs, is likely to fail. Data from prospective, randomized clinical trials of H2RA therapy in GERD indicate that esophagitis will heal and symptoms will be relieved in only about 50 percent of cases, and there are no data to suggest that a significant proportion of cases will be kept in remission. In addition, tolerance occurs with all H2RAs. When acid is suppressed, histamine receptors on parietal cells are up-regulated. As the dose of the H2RA is increased, the receptors continue to be up-regulated and tolerance develops. When patients have unrelieved symptoms, there are increased costs of care beyond drug therapy. At a minimum, patients return to the physician’s office for another prescription. Costs may be increased for additional diagnostic tests (e.g., endoscopy), emergency room visits, or 10 MANAGED CARE / SUPPLEMENT hospital admissions for atypical chest pain. The H2RAs should be reserved for mild or intermittent disease and the more effective agents, the PPIs, should be the standard therapy for most patients. Summary Acid-related disorders are common conditions that negatively impact quality of life for a significant number of people nationwide. The pathology of these conditions involves an imbalance between acid secretion by gastric parietal cells and the ability of upper GI tract mucosa to defend against the effects of the acid. Therefore, therapy is targeted at elevating gastric pH. PPIs are used to control the effects of excessive acid secretion. This class of drugs has a unique mechanism of action that inhibits the final pathway to gastric acid secretion — the parietal cell proton pump. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. Sanders SW. Pathogenesis and treatment of acid peptic disorders: comparison of proton pump inhibitors with other antiulcer agents. Clin Ther. 1996;18:2–34. Scott M, Gelhot AR. Gastroesophageal reflux disease: diagnosis and management. Am Fam Physician. 1999;59:1161–1169. Hetzel D. Acid pump inhibitors: the treatment of gastroesophageal reflux. Aust Fam Physician. 1998;27:487–491. Richardson P, Hawkey CJ, Stack WA. Proton pump inhibitors: pharmacology and rationale for use in gastrointestinal disorders. Drugs. 1998;56:307–335. Talley NJ, Weaver AL, Tesmer DL, Zinsmeister AR. Lack of discriminant value of dyspepsia subgroups in patients referred for upper endoscopy. Gastroenterology. 1993;105:1378–1386. Hogan WJ. Gastroesophageal reflux disease: an update on management. J Clin Gastroenterol. 1990;12(suppl 2):S21–S28. Bell NJV, Burget D, Howden CW, Wilkinson J, Hunt RH. Appropriate acid suppression for the management of gastroesophageal reflux disease. Digestion. 1992;51(suppl 1):59–67. Goldberg HI, Dodds WJ, Gee S, Montgomery C, Zboralske F. Role of acid and pepsin in acute experimental esophagitis. Gastroenterology. 1969;56:223–230. Orlando RC. The pathogenesis of gastroesophageal reflux disease: the relationship between epithelial defense, dysmotility, and acid exposure. Am J Gastroenterol. 1997;92(suppl):3S–5S. Lagergren J, Bergstrom R, Lindgren A, Nyren O. Symptomatic gastroesophageal reflux as a risk factor for esophageal adenocarcinoma. N Engl J Med. 1999;340:825–831. Lagergren J. Increased incidence of adenocarcinoma of the esophagus and cardia: reflux and obesity are strong and independent risk factors according to the SECC study. Lakartidningen. 2000;97:1950–1953. Lagergren J, Bergstrom R, Nyren O. Association between body mass and adenocarcinoma of the esophagus and gastric cardia. Ann Intern Med. 1999;130:883–890. Lagergren J, Bergstrom R,Adami HO, Nyren O.Association between medications that relax the lower esophageal sphincter and risk for esophageal adenocarcinoma. Ann Intern Med. 2000;133:165–175. Phaosawasdi K, Malmud LS, Tolin RD, Stelzer F, Applegate G, Fisher RS. Cholinergic effects on esophageal transit and clearance. Gastroenterology. 1981;81:915–920. Burget DW, Chiverton SG, Hunt RH. Is there an optimal degree of acid suppression of healing of duodenal ulcers? a model of the relationship between ulcer healing and acid suppression. Gastroenterology. 1990;99:345–351. Comparative Pharmacology Of Proton Pump Inhibitors JOSEPH A. BARONE, PHARMD, AND JOHN R. HORN, PHARMD Department of Pharmacy Practice and Administration, Rutgers University College of Pharmacy, Piscataway, N.J., and the University of Washington School of Pharmacy, Seattle G astric acid secretion is regulated by a network of central and peripheral mechanisms. The final step in the process occurs when hydrogen-potassium adenosine triphosphatase (the proton pump) exchanges intracellular hydrogen for extracellular potassium.1,2 Parietal cells of the gut contain a number of proton pumps. The proton pumps are inactive until they migrate to and extend cysteine residues through the cell wall. Once a proton pump is active, it has the ability to pump hydrogen ions into the secretory canaliculus, creating an acidic environment. The hydrogen ions combine with chlorine to form hydrochloric acid. The formation of hydrochloric acid is dependent on a number of mediators that activate the parietal cell through effects on membrane receptors, ion channels, and signal transduction mechanisms.1 Proton pump inhibitors (PPIs) block the final step of gastric acid secretion by the parietal cell. Proton pump inhibitors The PPIs are substituted benzimidazoles. The agents that belong to this class of drugs include omeprazole, lansoprazole, pantoprazole, esomeprazole, and rabeprazole. All PPIs are effective therapies for control of the excess gastric acid secretion that is associated with acid-related disorders. Whereas all the PPIs currently on the market have been shown to be clinically useful, they do not all have the same pharmacologic and clinical properties. The pharmacokinetic properties of the PPIs are summarized in Table 1. Knowledge of the differences that exist in pharmacology and clinical safety and efficacy may add to the optimal use of these agents for the management of acid-related disorders. Activation All PPIs are prodrugs that require acid-induced activation. After administration, PPIs are absorbed systemically and then resecreted into the canalicular space. Because the PPIs are weak bases, they become ionized when they are in the acidic environment of the canalicular space. Once ionized, the PPI is trapped in the acidic medium, becoming concentrated at the site of activity. The molecule is protonated, which changes its shape, and it is converted to its active form — a sulfenamide with exposed sulfur atoms. The exposed sulfur atoms bind covalently to the sulfur atoms in the cysteine groups of the protein pump. Once the drug binds to the proton pump, the pump is unable to exchange potassium for hydrogen in the parietal cell and is thus rendered inactive (Figure 1).1 All the PPIs currently on the market share one common binding site; each also has between one and four additional binding sites. Differences in activation. While the process of activation is generally the same for each of the PPIs, the protonation and activation steps are both pH and drug dependent. The five currently available agents differ in rate of activation, which may impact their onset of TABLE 1 Comparative pharmacokinetics of available proton pump inhibitors (PPIs) Author correspondence: Joseph A. Barone, PharmD, FCCP Associate Professor and Chairman Rutgers University College of Pharmacy Department of Pharmacy Practice and Administration 160 Frelinghuysen Road Piscataway, N.J. 08854-8020 Phone: (732) 445-3285 Fax: (732) 445-2533 E-mail: [email protected] PPI Protein binding ( %) Time to peak concentration (h) 0.5–1.0 1.5 1.9 0.7–1.5 95 97 98 96.3 0.5–3.5 1.7 2–4 1.6–5.0 1.2 1.5 97 97 1.6 1.6 Bioavailability Half-life (%) (h) Omeprazole 30–40 Lansoprazole 80 Pantoprazole 77 Rabeprazole 52 Esomeprazole 20 mg 50–68 40 mg 64–89 SUPPLEMENT / MANAGED CARE 11 The times to onset of action (inhibition of gastric acid) of the various PPIs may correlate with their pKa values. For example, because of its greater reactivity, rabeprazole may have a more rapid response to first dose than other PPIs. This effect has been demonstrated both in vivo and in vitro.4,5 In an in vitro study using porcine gastric cells, which are similar to human gastric cells, rabeprazole demonstrated 100 percent inhibition of the proton pump after only 5 minutes (Figure 3).3 This inhibition was maintained throughout the 45-minute study period. Lansoprazole, the second best inhibitor of acid at the initial time point, was associated with only 65 percent inhibition. Lansoprazole provided similar inhibition to rabeprazole, but only after 45 minutes of incubation with the cells. By the end of the study, pantoprazole reached only 50 percent inhibition.3 These differences can be explained, in part, by the agents’ relative acid stability. Rabeprazole is a more reactive molecule that is rapidly converted to FIGURE 1 Proton pump inhibitor mechanism of action. the active sulfenamide derivative. In Hydrogen-potassium adenosine triphosphatase contrast, pantoprazole is much more (H+,K+-ATPase) is the proton pump. stable in an acid environment and its conversion to the active sulfenamide derivative is relatively slow.6 Parietal cell Acid secretion. The meaningful clinical end point for patients with acid-related disorders is control of Proton pump acid secretion. Although all of the inhibitor PPIs control acid secretion, differences have been noted among the five H+ currently marketed compounds in H+,K+-ATPase Canaliculus pH <4.0 both the degree of acid control and the consistency of acid control. As an Activation example, Figure 4 portrays data comInhibition paring acid secretion in 23 healthy Protonation Helicobacter pylori–negative subjects treated in a crossover trial with eiActive ther rabeprazole 20 mg, omeprazole sulfenamide 20 mg, or placebo once daily before Adapted from Sachs et al.1 with permission. breakfast.5 Intragastric acidity decreases via the buffering action of FIGURE 2 Activation rates of PPIs vary with pH. ingested food and then increases sub300 282 sequent to acid secretion. RabeprapH 1.2 pH 5.1 zole was associated with a statistically 250 significant reduction in acid production compared to omeprazole for the 200 entire 24-hour period, including 150 three postprandial periods. Rabeprazole also produced an earlier decrease 90 100 84 in intragastric acidity. In all subjects, there was a nocturnal spike of acidity 50 beginning between approximately 2.8 2 4.6 1.3 7.2 8:30 and 9:00 PM (nocturnal acid 0 Rabeprazole Omeprazole Lansoprazole Pantoprazole breakthrough). This spike was not Adapted from Kromer et al.4 with permission. quite as high with rabeprazole as with Activation half-life (min) action. The rate of acid-induced activation of an individual PPI depends on the reactivity (i.e., the pKa) of the molecule.3 The pKa of a PPI is the pH at which half the drug is protonated and half is unionized. The pKa values of the PPIs that are currently on the market range from 3.8 to 5.0. The activation rate changes depending on the pH. At an extremely low pH of about 1.2, the activation rates of the various PPIs are all very rapid and similar. Any small differences in activation rates are not clinically significant. At higher pH values, however, there are differences among the PPIs. For example, at a pH of 5.0, activation half-lives for the PPIs vary: approximately 7 minutes for rabeprazole, about 90 minutes for omeprazole and lansoprazole, and approximately 5 hours for pantoprazole (Figure 2).4 To some extent, these differences are related to the pKa of the drugs, with rabeprazole having the highest pKa –– about 5.0. 12 MANAGED CARE / SUPPLEMENT Intragastric acidity (mmol/L) Inhibition of proton pump (%) omeprazole. These data confirm the hypothesis that nounced with omeprazole, accounting for the high derabeprazole is activated quickly and therefore has a more gree of interpatient variability following administration rapid onset of effect when compared to omeprazole. of this agent. Interpatient variability is also seen with panAmong PPIs, rabeprazole has the most significant toprazole, since its metabolism is similar to that of day-1 effects. Intragastric pH in the 24 hours following omeprazole. Nevertheless, because lansoprazole and rabeprazole are less CYP2C19-dependent, they have administration of a single dose is significantly higher for lower variability ratios. rabeprazole than for lansoprazole, pantoprazole, or In the general population, individuals can be classified omeprazole (Figure 5).7 Two studies are available comas either homozygous extensive, heterozygous extensive, paring rabeprazole to esomeprazole. Warrington et al. or poor metabolizers of PPIs, based on their CYP2C19 found that rabeprazole 20 mg increased the percent of enzyme activity. Interpatient variability following time that gastric pH was >4.0 significantly more than esomeprazole administration was demonstrated in a study omeprazole 20 mg on both day 1 and day 5 (Figure 6).8 that evaluated the effect of polymorphic metabolism on Wilder-Smith et al. compared esomeprazole 40 mg, intragastric pH.11 Omeprazole (20 mg) had no effect on rabeprazole 20 mg, and placebo and found a significant effect on intragastric pH on day 5 for both drugs, with pH in homozygous extensive (i.e., very rapid) metaboesomeprazole showing a larger difference than rabeprazole.9 The median pH and the percent of time pH FIGURE 3 Intensity of proton pump inhibition after one dose. was >4.0 were markedly lower for 100 Rabeprazole 100 rabeprazole in the Wilder-Smith 100 Lansoprazole study compared to other studies, 83 Omeprazole however. 75 Pantoprazole Overall potency. Fujisaki et al. 66 demonstrated a tenfold difference in the potency of omeprazole and 50 49 47 rabeprazole on a molar basis.10 This difference was determined by an in 25 vitro study in a porcine parietal cell 20 model. The aim of the experiment was to ascertain the concentration of 0 5* or 10 min 45 min each of the PPIs required to produce 50 percent inhibition of parietal cell *Rabeprazole tested at 5 minutes. activity. With rabeprazole, 50 percent Adapted from Besancon et al.3 with permission. inhibition was seen at about 0.3 mmol/L. Fifty percent inhibition with FIGURE 4 Day-1 effects of rabeprazole and omeprazole in 23 healthy omeprazole was noted at approxiHelicobacter pylori–negative subjects. mately 3.0 mmol/L.10 Potency differences between Rabeprazole 20 mg omeprazole and rabeprazole appear 100 M M M Omeprazole 20 mg to influence the consistency of their Placebo acid-reducing effects. Consistency of 80 effect was evaluated by reviewing individual patient results from the 60 crossover study published by Williams et al.5 (Figure 7). The magnitude of suppression is much more 40 consistent with rabeprazole than it is with omeprazole. 20 Metabolism. The metabolism of each PPI is related to extensive he0 patic biotransformation with vary8 AM 12 PM 4 PM 8 PM 12 AM 4 AM 8 AM ing degrees of dependence on cyTime of day tochrome P450 (CYP) 2C19, a polymorphically distributed enzyme. M = meal. This effect seems to be very pro- Adapted from Williams et al.5 with permission. SUPPLEMENT / MANAGED CARE 13 lizers. A heterozygous extensive metabolizer (i.e., a rapid or intermediate metabolizer) experienced a significant increase in gastric pH. Finally, poor metabolizers of omeprazole — those with little hepatic CYP2C19 — experienced statistically significant increases in gastric pH when compared with placebo and had greater blood concentrations of omeprazole.11 In the U.S. population, only 4 percent of individuals are poor metabolizers. For omeprazole, the difference between extensive and poor metabolizers results in a sixfold difference in plasma concentrations. Clinically, this polymorphism may affect the dose of omeprazole required by patients. Some patients achieve adequate symptom control with 20 mg qd, while others require 60 mg or 80 mg qd. In contrast, rabeprazole has less than a twofold variability between extensive and poor metabolizers. This property is related to the primarily nonenzymatic metabolism of rabeprazole. Omeprazole also has the idiosyncracy of inhibiting its own metabolism by acting as both a substrate and an inhibitor of CYP2C19. When the area under the time–plasma concentration curve (AUC) is evaluated, a doubling of omeprazole AUC levels is observed between days 1 and 5. It appears that there is a decreased clearance and accumulation of omeprazole after repeated administration.12 Clinical effect of pharmacologic differences Gastroesophageal reflux. Healing rates in gastroesophageal reflux disease (GERD) or peptic ulcer disease are essentially the same for all the PPIs. Probably because healing rates are so good with these drugs — about 90 percent — it is difficult to see any clinically significant differences between PPIs in most patients. Differences are detectable, however, in surrogate measures of efficacy such as esophageal acid exposure in individuals with GERD. For example, after the first dose of rabeprazole, there was a sigFIGURE 5 Day 1 antisecretory effects of PPIs in 18 H. pylori–negative nificant decrease in acid exposure in subjects. patients with documented erosive esophagitis. After 7 days of treatment, these patients had acid exposure within normal limits.13 H. pylori eradication. Current H. pylori eradication regimens employ 10- to 14-day therapy. Seven-day therapy for H. pylori eradication would benefit patients, since 7-day regimens would both decrease costs and allow patients to use less medication. Studies have shown that omeprazole therapy for 7 days may lack sufficient efficacy because 3 or 4 days are needed to achieve maximal acid suppression with this PPI.14 Data from Pantoflickova et al.7 Data from European trials with rabeprazole indicate, however, that FIGURE 6 Day-1 and day-5 effects on gastric pH of rabeprazole and 7-day combination therapy with esomeprazole, 20 mg/d, in 25 healthy H. pylori–negative rabeprazole is as effective as the resubjects. sults seen with longer regimens.15,16 The superiority of rabeprazole over omeprazole in 1-week eradication regimens may be related to rabeprazole’s ability to more quickly affect acid control and also to rabeprazole’s ability to inhibit H. pylori. All PPIs affect H. pylori, and this activity is typically measured by determining change in urease production. About 0.2–0.3 µmol of rabeprazole will inhibit 50 percent of urease activity; omeprazole and lansoprazole are Data from Tejura et al.8 markedly less potent in their inhibi- 14 MANAGED CARE / SUPPLEMENT FIGURE 7 Rabeprazole vs omeprazole: individual responses after first dose. 1400 977 Integrated intragastric acidity (mmol·h/L) 1200 1000 800 640 600 400 331* 200 0 Rabeprazole 20 mg (n=23) Placebo (n=23) Omeprazole 20 mg (n=23) *P<0.001 vs. omeprazole and placebo. Adapted from Williams et al.5 with permission. tion of H. pylori.17 This pharmacologic difference may explain the clinical differences. On-demand therapy. Rapidity of onset with rabeprazole is a pharmacodynamic effect. This effect becomes a more important issue when patients choose to take their PPI on demand for symptom relief rather than when therapy is continuous and long term as in ulcer treatment. Many patients use PPIs as needed in response to symptoms, making the onset of action an important differentiator. In contrast to omeprazole and other PPIs, rabeprazole achieves a maximal effect after the first dose, making it a suitable agent for on-demand therapy.7 Summary The PPIs are the most effective therapy to suppress gastric acid secretion. These agents decrease acid secretion by inhibiting parietal cell proton pumps. From chemical and pharmacodynamic points of view, subtle differences that exist among the PPIs may influence clinical activity. References 1. 2. Sachs G, Shin JM, Briving C, Wallmark B, Hersey S. The pharmacology of the gastric acid pump: the H+,K+-ATPase. Annu Rev Pharmacol Toxicol. 1995;35:277–305. Blum RA. Lansoprazole and omeprazole in the treatment of acid peptic disorders. Am J Health Syst Pharm. 1998;55:2289–2298. 3. Besancon M, Simon A, Sachs G, Shin JM. Sites of reaction of the gastric H+,K+-ATPase with extracytoplasmic thiol reagents. J Biol Chem. 1997;272:22438–22446. 4. Kromer W, Kruger U, Huber R, Hartmann M, Steinijans VW. Differences in pH-dependent activation rates of substituted benzimidazoles and biological in vitro correlates. Pharmacology. 1998;56:57–70. 5. Williams MP, Sercombe J, Hamilton MI, Pounder RE. A placebo-controlled trial to assess the effects of 8 days of dosing with rabeprazole versus omeprazole on 24-h intragastric acidity and plasma gastrin concentrations in young healthy male subjects. Aliment Pharmacol Ther. 1998;12:1079–1089. 6. Richardson P, Hawkey CJ, Stack WA. Proton pump inhibitors: pharmacology and rationale for use in gastrointestinal disorders. Drugs. 1998;56:307–335. 7. Pantoflickova D, Dorta G, Jornod P, et al. Identification of the characteristics influencing the degree of antisecretory activity of PPIs [abstract]. Gastroenterology. 2000;118:A5895. 8. Tejura B, Boyce M, Warrington S, et al. Rabeprazole is more potent than esomeprazole in control of gastric pH in healthy volunteers [abstract]. Ninth United European Gastroenterology Week Meeting, Amsterdam, The Netherlands. October 2001. 9. Wilder-Smith C, Röhss K, Claar-Nilsson C, Rydholm H. Esomeprazole 40 mg provides more effective acid control than rabeprazole 20 mg [abstract]. Gut. 2000;47(suppl 3):A63. Abstract P.51. 10. Fujisaki H, Murakami M, Fujimoto M, et al. The activity of isolated porcine H+,K+ ATPase is inhibited by E3810 [abstract]. FASEB J. 1990;4:A473. SUPPLEMENT / MANAGED CARE 15 11. Furuta T, Ohashi K, Kosuge K, et al. CYP2C19 genotype status and effect of omeprazole on intragastric pH in humans. Clin Pharmacol Ther. 1999;65:552–561. 12. Duvauchelle T, Millerioux L, Gualano V, Evene E, Alcaide A. Comparative bioavailability study of two oral omeprazole formulations after single and repeated administrations in healthy volunteers. Clin Drug Invest. 1998;16:141–149. 13. Robinson M, Maton PN, Rodriguez S, Greenwood B, Humphries TJ. Effects of oral rabeprazole on oesophageal and gastric pH in patients with oesophageal reflux disease. Aliment Pharmacol Ther. 1997;11:973–980. 14. Bigard MA, Delchier JC, Riachi G, Thibault P, Barthelemy P. One-week triple therapy using omeprazole, amoxycillin and clarithromycin for the eradication of Helicobacter pylori in patients with non-ulcer dyspepsia: influence of 16 MANAGED CARE / SUPPLEMENT dosage of omeprazole and clarithromycin. Aliment Pharmacol Ther. 1998;12:383–388. 15. Stack WA, Knifton A, Thirlwell D, et al. Safety and efficacy of rabeprazole in combination with four antibiotics for eradication of Helicobacter pylori in patients with chronic gastritis with or without peptic ulceration. Am J Gastroenterol. 1998;93:1909–1913. 16. Miwa H, Yamada T, Sato K, et al. Efficacy of reduced dosage of rabeprazole in PPI/AC therapy for Helicobacter pylori infection: comparison of 20 and 40 mg rabeprazole with 60 mg lansoprazole. Dig Dis Sci. 2000;45:77–82. 17. Tsuchiya M, Imamura L, Park JB, Kobashi K. Helicobacter pylori urease inhibition by rabeprazole, a proton pump inhibitor. Biol Pharm Bull. 1995;18:1053–1056. Optimizing Acid-Suppression Therapy JEFFREY L. BARNETT, MD, AND MALCOLM ROBINSON, MD University of Michigan Medical Center, Ann Arbor, Mich., and University of Oklahoma College of Medicine, Oklahoma City, Okla. A cid-related disorders, including gastroesophageal reflux disease (GERD), duodenal ulcers, and gastric ulcers, are common chronic conditions that have a negative impact on patient quality of life. The pathogenesis of these multifactorial conditions involves an imbalance between acid secretion by gastric parietal cells and the ability of the upper gastrointestinal tract to defend itself against the injurious effects of the acid. Treatment for these disorders focuses on raising gastric pH. Neutralizing acid with antacids or decreasing acid secretion with histamine-2 receptor antagonists (H2RAs) was the standard of care before the introduction of proton pump inhibitors (PPIs). Data from throughout the last decade show PPIs to be the most effective therapy for long-term symptom control and the healing of acid-related diseases.1–4 Ulcers will usually heal when the pH is maintained above 3.0 or 3.5, and the esophagus will heal in patients with GERD if the pH is kept above 4.0. In addition to the degree of acid suppression, the duration of acid suppression is important. PPIs maintain intragastric pH at >4.0 significantly longer than do H2RAs or conventional doses of antacids.5 Clinically, this correlates with more rapid and complete mucosal healing (Figure 1). Five PPIs are currently on the market: omeprazole, esomeprazole, lansoprazole, pantoprazole, and rabeprazole. While the use of PPIs to control acid in GERD is probably the most common application of these drugs, PPIs are also used to control atypical GERD manifestations, including noncardiac chest pain, laryngitis, asthma, and cough associated with reflux. PPIs are also useful for treating duodenal ulcer, gastric ulcer, hypersecretory conditions like Zollinger-Ellison syndrome, and dyspepsia in patients presenting with GERD-like or ulcer-like symptoms. While all PPIs inhibit the enzyme Author correspondence: Jeffrey L. Barnett, MD University of Michigan Medical Center UH-2B-355 1500 East Medical Center Drive Ann Arbor, MI 48109 Phone: (734) 936-8644; Fax: (734) 936-7966 E-mail: [email protected] hydrogen-potassium adenosine triphosphatase (i.e., the proton pump) — not all PPIs have the same pharmacologic and clinical properties. Rates of PPI activation differ based on the reactivity of individual molecules.6 The PPIs also differ in potency (which affects consistency of acid suppression) and metabolism. Rabeprazole is the most rapidly activated of the PPIs.7 Due to its greater reactivity, rabeprazole has a more rapid onset of action and thus a more rapid inhibition of the proton pump when compared with the other PPIs.8,9 After the first dose, rabeprazole creates 88 percent of maximal acid suppression on day 1 of therapy. In contrast, day-1 acid suppression induced by omeprazole is 42 percent of maximum.10 In a study that compared the effects of rabeprazole and omeprazole on the activity of H+,K+-ATPase, which was isolated from porcine gastric mucosa, investigators found a tenfold difference in inhibition of the enzyme between omeprazole and rabeprazole, which influences the consistency of acid-reducing effects.11 The magnitude of suppression is much more consistent with rabeprazole than omeprazole.8 The metabolism of the PPIs also differs based on their degree of dependence on cytochrome P450 (CYP) 2C19, a polymorphically distributed enzyme. A high degree of dependence on CYP2C19 gives rise to more interpatient variability with omeprazole and pantoprazole than with lansoprazole and rabeprazole, which are less dependent on this enzyme. In the absence of head-to-head clinical studies that compare all the available PPIs in each of the different acid-related disorders, surrogate factors are considered in the selection of an appropriate drug for each patient. Rapid onset of symptom control is extremely important to patients presenting with symptoms of reflux, and rapid symptom control is most meaningful in the first few days of starting treatment in a newly diagnosed patient. It is also critical as more patients use “ondemand” therapy to control their symptoms. A drug that will consistently control acid during the day as well as at night with once-daily administration is important to ensure patient compliance and minimize the cost of therapy. The drug should demonstrate effective healing and long-term maintenance of healing in patients with ulcers and in those with damage resulting from GERD. Finally, the drug of choice should have min- SUPPLEMENT / MANAGED CARE 17 time increases that esophageal pH remains below 4.0, so does the frequency of symptoms.12 Optimal therapy necessitates a PPI with a rapid onset of action to raise Symptom control with PPIs esophageal pH. Only about 5 percent to 10 percent of patients with Statistically significant differences in rapidity of sympGERD have erosive disease. Thus, controlling symptoms tom relief have been observed in studies comparing of reflux and improving patient quality of life become the rabeprazole to omeprazole.13 Figure 2 shows data from primary goals of therapy in this population. Symptom a study that evaluated the effects of the first 3 days of therfrequency in patients with GERD is directly related to the apy with these two drugs in patients who had varying dedegree of esophageal acid exposure. As the percentage of grees of symptom severity due to GERD. Rabeprazole was superior (P<0.036) to omeprazole. This difference continued through FIGURE 1 Healing rates of esophagitis correlate with duration of the first 7 days of therapy.13 pH control. Rabeprazole has been shown to be superior to omeprazole in 100 achieving rapid symptom relief in patients with active duodenal ulcer.14 Four weeks after initiating 80 therapy, patients treated with rabeprazole 20 mg had significantly 60 (P<0.038) less daytime ulcer pain severity; in addition, they had less ulcer pain frequency and less night40 time ulcer pain severity (Figure 3). r=0.87 Rabeprazole has also demon20 strated superior symptom improvement (P<0.05) when compared to treatment with 0 omeprazole in patients with active 2 4 6 8 10 12 14 16 18 20 22 gastric ulcer. 15 In a randomized Duration intragastric pH>4.0 (h/24 h) study of 227 patients administered either rabeprazole or omeprazole Adapted from Bell et al.5 with permission. once daily for 3 or 6 weeks, rabeprazole decreased gastric ulcer pain freFIGURE 2 Extent of symptom relief during first three treatment quency grades at week 3 and sigdays with rabeprazole (RAB) vs omeprazole (OME) nificantly at week 6 (P<0.006), and in patients with gastroesophageal reflux disease. decreased daytime pain severity grades significantly at week 3 OME 20 mg (n=105) RAB 20 mg (n=102) (P<0.023), with a reduction also ≥2 ≥3 ≥5 ≥2 ≥3 ≥5 0 evident at week 6, albeit not significantly. −0.2 Finally, rabeprazole has shown rapid relief of both daytime and −0.4 nighttime heartburn in patients −0.6 with endoscopically confirmed erosive esophagitis. 16 Patients had −0.8 moderate to severe symptoms at −1.0 baseline and used an interactive * * voice-response system to report −1.2 when they achieved mild or no symptoms. The majority of patients −1.4 Heartburn (HB) episodes — 65 percent — achieved satisfac*P≤0.036 rabeprazole vs omeprazole. tory relief of both daytime and nighttime heartburn on the first Heartburn episodes = number of moderate to severe episodes during screening. day of therapy. This contrasts with Data from Camacho et al.13 Change in HB score from baseline Patients healed (%) imal side effects — including a low potential for drug/drug interactions. 18 MANAGED CARE / SUPPLEMENT the level of relief achieved in a similar study with esomeprazole (45 percent) and omeprazole (32 percent) (See Figure 4).16,17 Healing with PPIs (P<0.001) for all time points. After 1 year of therapy, GERD continued to be well controlled. At week 52 of the studies, 94 percent to 97 percent of rabeprazole-treated patients had no relapse in daytime heartburn and 91 percent to 98 percent of rabeprazole-treated patients had no relapse in nighttime heartburn. Patients with Barrett’s esophagus are a special population who typically have more acid exposure than most patients with GERD and a risk of esophageal cancer if acid is left uncontrolled. Rabeprazole has been shown to maintain healing in this group as well, in a 1-year study evaluating maintenance of healing with rabeprazole versus placebo (P<0.006).21 Patients (%) with improvement at week 4 Reducing esophageal acid exposure is critical for healing of the esophageal damage evident in many patients with GERD.5 Studies have demonstrated a good correlation between the healing rate of esophagitis at 8 weeks and duration in hours that the intragastric pH is maintained above 4.0. Although similar findings are seen with a pH threshold of 3.0, the relationship is not as strong.5 Thus, achieving a pH >4.0 in the esophagus for as long as possible is an important goal of GERD therapy.18 A Extraesophageal indications for PPIs study comparing the control of intragastric acidity in healthy, Helicobacter pylori–negative subjects demonStudies with omeprazole have indicated the value of strated the superior decrease in gastric acidity with the PPIs in treating noncardiac chest pain; posterior rabeprazole when compared to omeprazole. During the 24-hour pe- FIGURE 3 Symptom relief at week 4 of therapy with rabeprazole riod following the first dose, the mean vs omeprazole in patients with active duodenal ulcer. percentage of time that intragastric pH remained >4.0 was 44.1 percent Rabeprazole 20 mg qd for rabeprazole and 24.7 percent for Omeprazole 20 mg qd omeprazole (P<0.001).8 Because the measurement of healing over several 95 96 100 92* months is a more crude measure, 83 multiple studies with direct compar76 80 isons of the different PPIs have, for 68 the most part, failed to show statisti60 cally significant differences in healing rates between these compounds. Es40 omeprazole has shown about a 9 percent improvement over omeprazole 20 in 8-week healing rates of erosive esophagitis in a large, multicenter, 0 double-blind trial.17 All the PPIs proUlcer pain Daytime ulcer Nighttime ulcer duce effective healing. Rabeprazole frequency pain severity pain severity has been shown to be equivalent to *P = 0.038 vs omeprazole. omeprazole in healing rates in both Data from Dekkers et al.14 erosive GERD and in healing of duoFIGURE 4 First-day complete heartburn relief with rabeprazole, denal ulcer.14,15 esomeprazole, and omeprazole. Maintenance therapy 70 65 60 Patients (%) GERD is a chronic disease. If therapy is stopped after erosive esophagitis is healed, about 70 percent of patients will experience relapse. Maintenance therapy is therefore important in this setting. All the PPIs maintain healing. Figure 5 shows the percentage of patients in remission while undergoing therapy with either placebo or rabeprazole in two separate studies. 19,20 Rabeprazole was significantly better than placebo 45 50 40 32 30 20 10 0 Rabeprazole 20 mg Esomeprazole 40 mg Omeprazole 20 mg Data from Robinson et al.16 and Richter et al.17 SUPPLEMENT / MANAGED CARE 19 and consistently. Due to its more rapid rate of activation, rabeprazole results in a faster onset of acStudy 1 (n=137) Study 2 (n=192) tion and faster symptom control Rabeprazole 20 mg qd Rabeprazole 20 mg qd than other PPIs. Studies comparPlacebo Placebo ing rabeprazole to omeprazole found statistically significant differences in the rapidity of symp100 90% Study 1 tom relief in patients with gastric 86% Study 2 ulcer, duodenal ulcer, and GERD. 80 Rapid symptom relief is impor60 tant to the majority of patients, as their symptoms have an impact on 40 their quality of life. Rapid sympStudy 1 29% Study 2 tom relief is also important in an 20 environment where patients selfmedicate on demand, depending 0 on daily symptoms. Rabeprazole 0 4 13 26 39 52 has also been shown to have a Time (weeks) more consistent suppression of P<0.001 vs placebo for all time points. acid, including at night. Optimizing therapy with PPIs necessitates Adapted from Birbara et al.19 and Caos et al.20 with permission. consideration not only of healing rates of the different available treatments but also of the rapidity and consistency of acid suppression that translate clinilaryngitis related to reflux, and asthma.22–24 Little data are cally into symptom relief. available on the newer PPIs in these indications, but studies are ongoing. Patients in endoscopic remission (%) FIGURE 5 Rabeprazole in long-term maintenance of healed erosive GERD. References Drug-drug interactions Interactions of drugs with PPIs can occur either as the drugs are being metabolized through the CYP450 system or at the absorption level, when absorption of the affected drug is dependent on intragastric pH.25,26 Omeprazole and lansoprazole have CYP450-related interactions. Due to the nature of their activity, all PPIs affect pH-dependent absorption of drugs such as digoxin and ketoconazole. Summary Acid-related disorders are caused by an imbalance between acid secretion by the gastric parietal cells and the defensive mechanisms of the gastrointestinal tract to protect against the effects of acid. Therapy for acid-related disorders focuses on the control of acidity. Data collected throughout the last decade have demonstrated that PPIs are the most effective therapy for acid-related disorders: PPIs have proven superior to H2RAs and antacids in numerous studies. Five PPIs are currently available in the United States. While all PPIs exert their effect through the same basic mechanism of action, they do not have the same pharmacologic and clinical properties. All PPIs are effective in healing and maintenance of gastric and duodenal ulcers and GERD. The PPIs differ, however, in their ability to control symptoms rapidly 20 MANAGED CARE / SUPPLEMENT 1. 2. 3. 4. 5. 6. 7. 8. 9. Richardson P, Hawkey CJ, Stack WA. Proton pump inhibitors: pharmacology and rationale for use in gastrointestinal disorders. Drugs. 1998;56:307–335. Berardi RR, Welage LS. Proton pump inhibitors in acid-related diseases. Am J Health Syst Pharm. 1998;55:2289-2298. Hunt RH. The relationship between the control of pH and healing and symptom relief in gastro-oesophageal reflux disease. Aliment Pharmacol Ther. 1995;9(suppl):3–7. Chiba N, De Gara CJ, Wilkinson JM, Hunt RH. Speed of healing and symptom relief in grade II to IV gastroesophageal reflux disease: a meta-analysis. Gastroenterology. 1997;112:1798–1810. Bell NJV, Burget D, Howden CW, Wilkinson J, Hunt RH. Appropriate acid suppression for the management of gastro-oesophageal reflux disease. Digestion. 1992;51(suppl 1):59–67. Besancon M, Simon A, Sachs G, Shin JM. Sites of reaction of the gastric H+,K+-ATPase with extracytoplasmic thiol reagents. J Biol Chem. 1997;272:22438–22446. Sachs G, Shin J, Briving C, Wallmark B, Hersey S. The pharmacology of the gastric acid pump: the H+,K+ ATPase. Annu Rev Pharmacol Toxicol. 1995;35:277–305. Williams MP, Sercombe J, Hamilton MI, Pounder RE. A placebo-controlled trial to assess the effects of 8 days of dosing with rabeprazole versus omeprazole on 24-h intragastric acidity and plasma gastrin concentrations in young healthy male subjects. Aliment Pharmacol Ther. 1998;12:1079–1089. Kromer W, Kruger U, Huber R, Hartmann M, Steinijans VW. Differences in pH-dependent activation rates of sub- 10. 11. 12. 13. 14. 15. 16. 17. stituted benzimidazoles and biological in vitro correlates. Pharmacology. 1998;56:57–70. Gardner JD, Sloan S, Barth A. Onset, duration and magnitude of gastric antisecretory effects of rabeprazole and omeprazole. Am J Gastroenterol. 1999;94:2608. Abstract 125. Fujisaki H, Murakami M, Fujimoto M, et al. The activity of isolated porcine H+,K+ ATPase is inhibited by E3810 [abstract]. FASEB J. 1990;4:A473. Joelsson B, Johnsson F. Heartburn—the acid test. Gut. 1989;30:1523–1525. Camacho F, Perdomo C, Jokubaitis L, Sloan S, Delchier J-C, Light M. Rabeprazole provides better heartburn relief compared to omeprazole in the first 3 and 7 days of treatment. Am J Gastroenterol. 2000;95:2434–2435. Abstract 82. Dekkers CPM, Beker JA, Thjodleifsson B, Gabryelewicz A, Bell NE, Humphries TJ, and the European Rabeprazole Study Group. Comparison of rabeprazole 20 mg versus omeprazole 20 mg in the treatment of active duodenal ulcer: a European multicentre study. Aliment Pharmacol Ther. 1999;13:179–186. Dekkers CPM, Beker JA, Thjodleifsson B, Gabryelewicz A, Bell NE, Humphries TJ, and the European Rabeprazole Study Group. Comparison of rabeprazole 20 mg vs omeprazole 20 mg in the treatment of active gastric ulcer: a European multicentre study. Aliment Pharmacol Ther. 1998;12:789–795. Robinson M, Fitzgerald S, Hegedus R, Murthy A, Jokubaitis L. Onset of symptom relief with rabeprazole: a community-based open-label assessment of patients with erosive esophagitis. Submitted for publication. Richter JE, Kahrilas PJ, Johanson J, et al. Efficacy and safety of esomeprazole compared with omeprazole in GERD pa- 18. 19. 20. 21. 22. 23. 24. 25. 26. tients with erosive esophagitis: a randomized controlled trial. Am J Gastroenterol. 2001;96:656–665. Hetzel D. Acid pump inhibitors: the treatment of gastroesophageal reflux. Aust Fam Physician. 1998;27:487–491. Birbara C, Breiter J, Perdomo C, Hahne W. Rabeprazole for the prevention of recurrent gastrooesophageal reflux disease. Rabeprazole Study Group. Eur J Gastroenterol Hepatol. 2000;12:889–897. Caos A, Moskovitz M, Dayal Y, Perdomo C, Niecestro R, Barth J. Rabeprazole for the prevention of pathologic and symptomatic relapse of erosive or ulcerative gastroesophageal reflux disease. Rabeprazole Study Group. Am J Gastroenterol. 2000;95:3081–3088. Barth J, Perdomo C, Sloan S. Rabeprazole maintains GERD healing in patients with Barrett’s esophagus. Am J Gastroenterol. 1999;94:2579. Abstract 10. Achem SR, Kolts BE, MacMath T, et al. Effects of omeprazole versus placebo in treatment of noncardiac chest pain and gastroesophageal reflux. Dig Dis Sci. 1997;42:2138–2145. Wo JM, Grist WJ, Gussack G, Delgaudio JM, Waring JP. Empiric trial of high-dose omeprazole in patients with posterior laryngitis: a prospective study. Am J Gastroenterol. 1997;92:2160–2165. Harding SM, Richter JE, Guzzo MR, Schan CA, Alexander RW, Bradley LA. Asthma and gastroesophageal reflux: acid suppressive therapy improves asthma outcome. Am J Med. 1996;100:395–405. Unge P, Andersson T. Drug interactions with proton pump inhibitors. Drug Safety. 1997;16:171–179. Humphries TJ, Nardi RV, Lazar JD, Spanyers SA. Drug-drug interaction evaluation of rabeprazole sodium: a clean/expected slate. Gut. 1996;39(suppl 3):A47. Abstract 297. SUPPLEMENT / MANAGED CARE 21 Evidence-Based Health Care: Making Health Policy and Management Decisions NIMISH B. VAKIL, MD University of Wisconsin Medical School, Milwaukee V iews about health care delivery in the United States have changed substantially during the last 30 years. In the 1970s, as health care costs began to rise, providers and insurers attempted to control costs by providing care at a lower cost. This concept was relatively easy to implement: health care should be delivered in the shortest possible time by the least expensive professional, using the cheapest possible therapies within the constraints of safety and effectiveness. In the 1980s and early 1990s, the focus shifted to providing better care. While a laudable goal, improved quality is difficult to deliver ad infinitum without consideration of costs. With the advent of managed care in the 1990s, there was an attempt to marry the principles of reduced cost and higher quality. Despite some success, cost reductions often superseded quality improvements, resulting in consumer and physician backlashes against managed care organizations. Today, the health care system has evolved to the point where providers and managed care organizations are beginning to recognize that to provide cost-effective quality care, they need to do the “right things right.” This means that appropriate therapy is delivered to patients at the appropriate time while potentially harmful therapies and therapies of uncertain or unproven value are avoided. Postgraduate training programs are incorporating evidence-based medicine in their curricula, and numerous organizations have been established to support the needs of evidence-based medicine.1 Author correspondence: Nimish B.Vakil, MD University of Wisconsin Medical School Sinai Samaritan Medical Center 945 North 12th Street P.O. Box 342 Milwaukee,WI 53233-0342 Phone: (414) 219-7762 Fax: (262) 524-0581 E-mail: [email protected] 22 MANAGED CARE / SUPPLEMENT What is evidence-based medicine? Evidence-based medicine is the conscientious, explicit, and judicious use of current best evidence from clinical care research in the management of individual patients.2 The evidence must be explicit, meaning that the source of the evidence must be obvious and scientifically justified, especially if used to develop guidelines or algorithms for patient care. Evidence-based medicine integrates three key components in an attempt to deliver the best care: (1) the individual clinical expertise of the provider, (2) the best available external clinical evidence from systematic research, and (3) patients’ values and expectations regarding their treatment. Evidence-based medicine is not “cookbook medicine,” because it requires the clinical judgment of the physician to extrapolate the published clinical evidence to the care of a specific patient having unique biology, values, and expectations. It is also not “cost-cutting medicine,” since evidence-based care may cause costs to rise as well as to fall. Critics of evidence-based medicine claim that it suppresses the clinical freedom of the clinician. In fact, the opinions, judgment, and expertise of the individual provider are critical to the practice of evidence-based medicine. The proficiency, judgment, and expertise acquired during years of medical practice enhance the application of evidence-based care. The treating physican is best qualified to assess the individual patient’s values, preferences, prognosis, and condition at any given time. Good doctors use both their own expertise and the best available external clinical evidence to determine best treatment practices. External clinical evidence may consist of data from relevant clinical trials, knowledge about the precision and accuracy of specific diagnostic tests, or collective information about the efficacy and safety of therapeutic, rehabilitative, and preventive regimens. External clinical evidence may either validate or invalidate previously accepted standards or may suggest replacing those standards with options proven to be more accurate, more efficacious, or safer. External clinical evidence is meant to inform the physician and not to replace clinical judgment and knowledge of the patient. The randomized clinical trial has become the standard for judging the benefits of a particular intervention and provides the highest level of evidence for a treatment or intervention. Some questions may be answered without large randomized trials, and some cannot wait for trials to be conducted. In these instances, the best currently available evidence is used to support clinical decision making. Physicians trained in evidence-based medicine learn how to conduct a systematic analysis of data by reviewing published literature, evaluating study design and quality, assessing the quality of outcomes, and then using this information to make therapeutic decisions. Evidence-based health care management As the practice of managed care matures during the coming years, evidence-based medicine has the potential to become an integral part of delivering quality care at an appropriate cost. To fully realize these benefits, managed care organizations will need to: Ensure that services and procedures are supported by high-quality evidence. Evidence may be gathered through well-designed, well-executed technology assessments or through critical appraisal of provider practices. An established source of technologic and treatment assessment in health care is the Cochrane database, which evaluates both services and procedures and currently contains over 1,000 assessments of different disease states. Ensure that the mix of services and procedures provided is one that will give the greatest benefit to the treated population. This can be done using needs assessments to set priorities for allocating health care resources. Ensure that services and procedures are of sufficiently high quality to realize the potential identified in research settings. The efficacy of a new procedure may be defined as its performance in carefully controlled settings, usually in selected patients in the hands of experts. The effectiveness of a new treatment is its performance in routine clinical practice. When innovations are applied in community practice, the results may not be the same as seen in randomized trials, creating the need for ongoing audits of actual outcomes for patients in the community. On occasion, a treatment cannot be adequately transferred from the well-controlled research setting to a general community treatment setting. This problem may be resolved by educating patients and providers about the proper use of technologic innovations. When education does not resolve the issue, it may be that outcomes are poor, costs are higher, or adverse events are more frequent, suggesting that the treatment or procedure may not be as valuable as previously believed. In these cases, the services and procedures may need to be changed. Realities of evidence-based medicine Technologic innovation has the potential to deliver treatments that do more harm than good, more good than harm, and treatments of unknown effect. Therapies that do more harm than good are frequently used as a result of poor or insufficient research that is often based on anecdotal evidence or individual case reports. There are many unfortunate examples of therapies that do more harm than good: the Angelchik prosthesis was promoted as a cure for reflux disease but was later found to have many serious complications. The gastric bubble was proposed as a treatment for obesity and widely adopted in community practice until a randomized controlled trial showed that it was ineffective. It should be recognized that the highest levels of evidence are available for only a small number of medical procedures. Available studies may have enrolled too few patients, looked at the wrong patient population, or introduced unintended bias by their design. The principles of evidence-based medicine can prevent therapies that do more harm than good from ever being introduced. New therapies or procedures should not be instituted in clinical practice without good data from well-conducted trials. This clarifies the risk and benefits of the procedure and allows patients and physicians to make informed choices. In the future, managed care organizations need to be willing to invest in clinical research related to medical innovations to ensure that they have the data necessary to guide treatment for their members. Managed care organizations also need to be proactive and develop processes to keep abreast of new techniques and new procedures and evaluate the benefits of these innovations. As evidence-based medicine becomes more accepted, it will be necessary to come to terms with the reality of information overload for physicians. It has been estimated that a physician in internal medicine needs to read 19 articles per day every day to keep abreast of the relevant literature being published.3 This figure is likely to increase as more information becomes available. It has been estimated that in an average day, a general practitioner develops an average of 16 unanswered clinical questions. Searching for answers in the litera- Evidence-based medicine leads to better outcomes O utcomes of patients admitted for stroke in U.S. hospitals where evidence-based medicine had been incorporated was compared to outcomes in institutions without evidence-based medicine.1 The care delivered by evidence-based neurologists differed from standard care — resulting in a 22-percent lower likelihood of dying during the 90 days immediately following stroke than in institutions not practicing evidence-based medicine. SUPPLEMENT / MANAGED CARE 23 FIGURE 1 the question is asked, it will be necessary to develop the best evidence to answer the question. A new audit cycle will enable researchers to ask and answer questions efficiently (Figure 1). The new cycle will necessitate that providers and researchers ask a question, gather the evidence, and then set the standard. Once a standard is set, a re-audit will ensure that it is delivering the best outcome. To gather evidence, providers and insurers will either need to rely on systematic reviews performed by an independent organization or, if such an organization is not available, will have to invest in the time and resources needed to complete an in-house review of the available literature or conduct the relevant clinical trial. Evidence-based audit cycle. Figure courtesy of Nimish B.Vakil, MD ture and providing adequate tools to conduct these searches in managed care can optimize treatment. What does the future hold? Guidelines for treatment and medication in the next decade will all become evidence based. As a result, guidelines that depend solely on cost considerations will be challenged by providers, by provider organizations, and by patients who are becoming increasingly well informed. Therapeutic decision making will be based on cost-effectiveness within the evidence-based framework. Managers and managed care organizations charged with developing evidence-based guidelines will be forced to analyze research more carefully. Standard questions will include: • Is this the best type of research to answer this question? • Is the research of adequate quality? • Is the research applicable to my patient population and, specifically, to this individual patient? Researchers will change the types of questions they are trying to answer. Instead of “should we lower cholesterol in our population, and to what value?” we will be asking “do statins save lives, and should we pay for them?” Once 24 MANAGED CARE / SUPPLEMENT Summary The paradigm for health care delivery in the United States continues to evolve. Patients and physicians are beginning to reject guidelines for treatment that are solely based on cost reduction and are not evidence based. Future managed care guidelines will rely on the best external clinical evidence about the value of any given therapeutic intervention combined with individual physician experience and patient choice. Managed care organizations will need to play a greater role in the development, evaluation, and incorporation of clinical trials into their strategies. References 1. Mitchell JB, Ballard DJ, Whisnant JP, Ammering CJ, Samsa GP, Matchar DB. What role do neurologists play in determining the costs and outcomes of stroke patients? Stroke. 1996; 27:1937–1943. 2. Sackett DL, Rosenberg WMC, Muir Gray JA, Haynes RB, Richardson WS. Evidence-based medicine: what it is and what it isn’t. BMJ. 1996;312:71–72. 3. Davidoff F, Haynes B, Sackett D, Smith R. Evidence-based medicine — a new journal to help doctors identify the information they need. BMJ. 1995;310:1085–1086.
© Copyright 2024