1. health and fitness

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.