This report was filed in the Louisiana Metoclopramide lawsuit entitled Tarver v. Wyeth.
Metoclopramide is a neuroleptic dopamine receptor antagonist, currently marketed
in the U.S. (Reglan(R)) for short-term therapy of gastroesophageal reflux and for
treatment of diabetic gastroparesis. In general, therapy with metoclopramide
tablets should not exceed 12 weeks in duration, as prolonged therapy and
cumulative dose may increase the risk of developing tardive dyskinesia. The aim of
this review is to assist in preparation for a forthcoming advisory committee
meeting on the approvability of Myzan(R), a combination of naproxen and
metoclopramide, for the acute treatment of migraine, by estimating both the use of
metoclopramide in the U.S. as well as the proportion of patients taking
metoclopramide for more than 90 days.....
Over the ten-year period from 1995 to 2004, there was a two-fold increase in the
number of outpatient prescriptions dispensed for metoclopramide in the U.S. In
1995, the number of prescriptions dispensed for metoclopramide was almost 3.5
million. This number declined to less than 2 million prescriptions dispensed
annually from 1997 to 1999, and then increased abruptly in 2000 and reached over 7
million prescriptions dispensed during 2004. The increase in dispensed
prescriptions beginning in 2000 is likely due to the withdrawal of cisapride, a
treatment for severe nighttime heartburn, from the US. market in March 2000.
Almost 90% of the metoclopramide products dispensed were in an oral tablet dosage
form.
Over one-third of the annual visits in the U.S., in which metoclopramide was
mentioned, were made by patients ages 35-64 years old. Elderly patients (aged 65
years and older) accounted for about a third of the visits in which metoclopramide
was mentioned from 1995-1998, yet their proportions declined steadily in the
subsequent years. As opposed to the elderly population, visits in which
metoclopramide was mentioned in the pediatric population (aged 0-16 years)
increased steadily throughout the ten-year study period and constitute more than
30% of these visits in 2004. Visits in which metoclopramide was mentioned in
patients aged 17- 34 years remained steady. On average, over 55% of the annual
visits in which metoclopramide was mentioned were made by females.
During the last five years, from 2000 through 2004, more than 50% of
metoclopramide uses were for gastroenterology-related diagnoses. "Esophagitis"
(ICD-9 code 530.1) was ranked the most common diagnosis for metoclopramide
mentions, accounting for over 40% of metoclopramide mentions.
United States District Court, W.D. Louisiana.
TARVER et al,
v.
WYETH INC et al.
No. 04CV02036.
June 10, 2005.
MEMORANDUM DEPARTMENT OF HEALTH AND HUMAN SERVICES PUBLIC HEALTH SERVICE FOOD AND
DRUG ADMINISTRATION CENTER FOR DRUG EVALUATION AND RESEARCH
PID#: D050036
FROM: Sigal Kaplan, Ph.D, B.Pharm Pharmacoepidemiologist Division of Surveillance,
Research and Communication Support, HFD-410
THROUGH: Gerald Dal Pan, MD., MHS. Director Division of Surveillance, Research and
Communication Support, HFD-410
TO: Judy Racoosin, MD, MPH Division of Neuropharmacological Drug Products, HFD-120
Office of New Drugs
SUBJECT: Metoclopramide: Drug Use Data Review (Myzan(R), NDA 21-645)
**This document contains proprietary data from IMS Health and Caremark which
cannot be shared outside of FDA without clearance from IMS Health and Caremark
obtained through the Office of Drug Safety.**
EXECUTIVE SUMMARY
This consult examines outpatient prescribing and drug use patterns for
metoclopramide in the U.S. as well as the duration of therapy for this drug using
longitudinal patient level data.
Metoclopramide is a neuroleptic dopamine receptor antagonist, currently marketed
in the U.S. (Reglan(R)) for short-term therapy of gastroesophageal reflux and for
treatment of diabetic gastroparesis. In general, therapy with metoclopramide
tablets should not exceed 12 weeks in duration, as prolonged therapy and
cumulative dose may increase the risk of developing tardive dyskinesia. The aim of
this review is to assist in preparation for a forthcoming advisory committee
meeting on the approvability of Myzan(R), a combination of naproxen and
metoclopramide, for the acute treatment of migraine, by estimating both the use of
metoclopramide in the U.S. as well as the proportion of patients taking
metoclopramide for more than 90 days.
The consult includes the following two sections:
I. National Utilization Patterns of Metoclopramide
Description
The first section describes the utilization patterns of metoclopramide in the U.S.
in a ten-year period from 1995-2004. This includes national estimates of physician
office visit characteristics associated with the prescription of metoclopramide,
such as patient's age, gender, indication for use and prescriber specialty
information. Data for this section were derived from cross-sectional outpatient
data from IMS Health audits (IMS Health, National Prescription Audit PlusTM [NPA
PlusTM] and IMS Health, National Disease and Therapeutic IndexTM [NDTITM]),
proprietary drug use databases licensed by the Agency.
Results
Over the ten-year period from 1995 to 2004, there was a two-fold increase in the
number of outpatient prescriptions dispensed for metoclopramide in the U.S. In
1995, the number of prescriptions dispensed for metoclopramide was almost 3.5
million. This number declined to less than 2 million prescriptions dispensed
annually from 1997 to 1999, and then increased abruptly in 2000 and reached over 7
million prescriptions dispensed during 2004. The increase in dispensed
prescriptions beginning in 2000 is likely due to the withdrawal of cisapride, a
treatment for severe nighttime heartburn, from the US. market in March 2000.
Almost 90% of the metoclopramide products dispensed were in an oral tablet dosage
form.
Over one-third of the annual visits in the U.S., in which metoclopramide was
mentioned, were made by patients ages 35-64 years old. Elderly patients (aged 65
years and older) accounted for about a third of the visits in which metoclopramide
was mentioned from 1995-1998, yet their proportions declined steadily in the
subsequent years. As opposed to the elderly population, visits in which
metoclopramide was mentioned in the pediatric population (aged 0-16 years)
increased steadily throughout the ten-year study period and constitute more than
30% of these visits in 2004. Visits in which metoclopramide was mentioned in
patients aged 17- 34 years remained steady. On average, over 55% of the annual
visits in which metoclopramide was mentioned were made by females.
During the last five years, from 2000 through 2004, more than 50% of
metoclopramide uses were for gastroenterology-related diagnoses. "Esophagitis"
(ICD-9 code 530.1) was ranked the most common diagnosis for metoclopramide
mentions, accounting for over 40% of metoclopramide mentions.
During 2004, internal medicine was the most frequent prescribing specialty,
accounting for over 1.2 million (22%) dispensed prescriptions for metoclopramide,
followed by family practice (18%) gastroenterology (13%), osteopathic medicine
(8%) and pediatric practice (5%).
II. Duration of Therapy for Metoclopramide
Description
The second section depicts patient characteristics and duration of therapy for
metoclopramide in a large insured population whose prescription drug benefits were
managed by Caremark in a three-year period from 2002-2004. Longitudinal data for
this section were derived from outpatient claims data from Caremark's Dimension Rx
TM, a proprietary drug use database licensed by the Agency at the time of this
study.
Methods
The study population consisted of individuals aged 1-99 years who were
continuously eligible in the Caremark pharmacy benefit manager (PBM) database and
had prescription claims for any metoclopramide products between January 1,2002 to
December 31, 2004. Claims were excluded from this analysis based on pre-determined
exclusion criteria. The final data analysis was performed on a subset of 200,907
patients with 572,205 prescription claims for metoclopramide tablets.
Claims data for this subset of the study population were used to determine an
episode of metoclopramide therapy for an individual. An episode of therapy was
defined as one or a series of consecutive prescription claims for metoclopramide
for an individual with no more than a 30-day lag between a new prescription and
the ending date of the preceding prescription. The duration of each episode of
therapy was calculated for each episode by summing up the total days' supply from
claims that constituted that episode. The duration time lag between the claims was
not included in the calculation of the duration of therapy. The duration of the
longest episode for metoclopramide was determined for each patient. Cumulative
exposure to metoclopramide during the study period was estimated for each patient
by summing the duration of each episode.
Three sensitivity analyses were performed to examine the robustness of the
assessment of duration of therapy, duration of the longest episode of therapy as
well as the cumulative exposure.
Results
Prescriptions claims for metoclopramide were almost exclusively filled by adults,
50% of which were persons aged 35-64 years. The mean age of the study population
was 58 years old (standard deviation (SD) =17.9). About two-third of the claims
were filled by females and only one-third by males. Approximately one-half of the
claims were filled in pharmacies located in the southern region in the U.S.
The average number of prescriptions for metoclopramide per patient was 2.8
(SD=4.7). The majority of the patients (66%) had only one paid prescription claim
for metoclopramide. The length of the longest episode of therapy for most patients
(87%) varied from 1 to 90 days, yet about 13% of the patients appeared to have
received prescriptions for metoclopramide for a period longer than 90 days.
For the majority of the patients in the study (68%), the length of cumulative
therapy for metoclopramide was 1 to 30 days. The frequency of patients decreased
as the total cumulative therapy for metoclopramide increased. Nonetheless, the
frequency of patients who had cumulative therapy for metoclopramide for more than
6 months exceeded 10%.
A sensitivity analysis of the duration of the longest episode and cumulative
therapy for individuals whose claims were for 100 days of supply or less reveals
that there was no significant change from the result presented above.
Conclusions
Results from this study indicate that the number of prescriptions dispensed for
metoclopramide, which were declining from 1995 to 1999, have more than tripled in
the five-year period following the withdrawal of cisapride from the U.S. market in
March 2000. In addition to this increase in the U.S. in recent years, data from an
insured population source showed that over 10% ( 10-15%) of metoclopramide users,
had paid claims suggesting use for a period longer than the three months
recommended in the label. Moreover, data suggestive of total cumulative therapy
with metoclopramide longer than 6 months was found in about 10% of the study
population.
The increase in the number of prescriptions dispensed for metoclopramide along
with the longer than recommended duration of use found in this study was
substantial. These findings suggest that even in light of metoclopramide label
restrictions on the duration of use and its well-recognized adverse events,
particularly extrapyramidal movement disorders, physicians as well as patients did
not follow the labeled recommendation for limited duration of therapy. These
results suggest that a similar label recommendation for a new combination product
that contains metoclopramide is not likely to be followed, especially if the
product contains naproxen, which is frequently used without restrictions on
duration.
INTRODUCTION
In May 2003, the Division of Neuropharmacological Drug Products within the Office
of New Drugs issued a Not Approvable letter for a New Drug Application (NDA) for
Myzan(R) (formerly MT100, NDA 21-645), a combination of naproxen and
metoclopramide for the acute treatment of migraine. The letter was issued because
the agency found that the sponsor did not establish the contribution of
metoclopramide to the efficacy of the product combination. Furthermore, a major
safety concern was raised due to the risk of movement disorders (parkinsonism,
akathisia and acute dystonia), well-known to be associated with metoclopramide
hydrochloride.
In late 2004, the sponsor submitted additional data in order to demonstrate that
metoclopramide seemed to provide a significant improvement in a subgroup of
migraine patients with no nausea at baseline. This effect seems counter-intuitive
since one would expect that the pharmacological effect of such an antiemetic drug
would be expressed and the drug would show effectiveness in patients with nausea.
Regardless of the unexpected effect of this antiemetic drug, the division may
agree to grant an indication for treatment of migraine in that subpopulation, if
the sponsor prospectively replicates that finding and demonstrates that the
clinical benefit outweighs the risk of tardive dyskinesia (TD) and other movement
disorders associated with metoclopramide.
However, before the sponsor conducts an additional efficacy study, an advisory
committee (AC) meeting is scheduled to determine if the risk of TD with chronic
intermittent use of metoclopramide is sufficiently high to preclude approval of
the drug for a migraine indication, even if the efficacy study to be conducted
shows an effect in the subpopulation of non-nauseated patients.
The Division of Surveillance, Research and Communication Support (DSRCS) within
the Office of Drug Safety (ODS) was consulted by the Division of
Neurophannacological Drug Products within the Office of New Drugs to assist in the
preparation for this advisory committee meeting on the approvability of Myzan(R)
for the acute treatment of migraine. DSRCS was requested to examine the use of
metoclopramide products in the U.S. and provide information on the following:
* Estimate and characterize the use of metoclopramide in the U.S. over the past 10
years
* Examine use patterns to determine whether physicians and patients follow the
labeled recommendations with regard to the restriction on duration of therapy, and
to specifically determine the proportion of patients whose pattern of paid
prescription claims suggests that they received treatment for a longer period than
the three months specified in the label.
This review describes outpatient prescribing and drug use patterns for
metoclopramide in the U.S. and examines the duration of therapy for this drug in
longitudinal patient level data. Proprietary drug use databases licensed by the
Agency were used to conduct this analysis.
BACKGROUND
Metoclopramide is a neuroleptic dopamine receptor antagonist, currently marketed
in the U.S. (Reglan(R)) for short-term therapy (4 to 12 weeks) of gastroesophageal
reflux, only when conservative treatment fails, and for treatment of diabetic
gastroparesis (2-8 weeks). It is available in three dosage forms: oral tablets,
oral solution and injection. The recommended administration regimen of the oral
tablets for the relief of symptomatic gastroesophageal reflux is 10 mg to 15 mg
metoclopramide up to four times a day, and for the relief of symptoms associated
with diabetic gastroparesis (diabetic gastric stasis) - 10 mg of metoclopramide
before each meal and at bedtime for two to eight weeks. In general, therapy with
metoclopramide tablets should not exceed 12 weeks in duration. Reglan(R)
(metoclopramide) labeling states that patients treated with metoclopramide may
develop tardive dyskinesia, a syndrome consisting of potentially irreversible,
involuntary, dyskinetic movements. The label states that the risk increases with
the duration of treatment and the cumulative dose.
A review of the medical literature supports that metoclopramide is associated with
TD as well as with other movement disorders.[FN1][FN2][FN3][FN4][FN5] Moreover, it
was suggested that metoclopramide-induced TD can be permanent.[FN6] However, this
type of side effect can only be expected to be observed when large populations are
exposed to the drug for a long time period, and not during the relatively short
drug development process, such as the one for Myzan(R) (where no case was reported
in the NDA). Therefore, the absence of cases of tardive dyskinesia in the Myzan(R)
trial is not indicative that this adverse reaction will not occur, but rather that
the trial duration was likely too short to detect such a safety problem. Indeed,
in an analysis of 67 cases of metoclopramide-associated tardive dyskinesia, the
mean length of treatment with metoclopramide before the onset of symptoms was 20
months.[FN4] In that study, the symptoms were still present 6 months or more after
discontinuation of metoclopramide in most patients on whom long-term follow-up was
provided.
FN1. Ganzini L, Casey DE, Hoffman WF, McCall AL. The prevalence of
metoclopramide-induced tardive dyskinesia and acute extrapyramidal movement
disorders. Arch Intern Med 1993 June 28;153(12):1469-75.
FN2. Jimenez-Jimenez FJ, Garcia-Ruiz PJ, Molina JA. Drug-induced movement
disorders. Drug Saf 1997 March; 16(3): 80-204.
FN3. Miller LG, Jankovic J. Metoclopramide-induced movement disorders.
Clinical findings with a review of the literature. Arch Intern Med 1989
November;149(11):2486-92.
FN4. Sewell DD, Jeste DV. Metoclopramide-associated tardive dyskinesia. An
analysis of 67 cases. Arch Fam Med 1992 November;1(2):271-8.
FN5. Tarsy D, Indorf G. Tardive tremor due to metoclopramide. Mov Disord
2002 May;17(3):620-1.
FN6. Shaffer D, Butterfield M, Pamer C, Mackey AC. Tardive dyskinesia risks
and metoclopramide use before and after U.S. market withdrawal of cisapride.
JAm Pharm Assoc (Wash DC) 2004 November;44(6):661-5.
In addition to the increased risk of tardive dyskinesia, metoclopramide is
associated with other neurological and psychiatric symptoms. Tardive tremor due to
metoclopramide was recently described.[FN5] There is also a warning in Reglan(R)
labeling about a risk for depression,[FN7][FN8][FN9] suicidal ideation and
suicide. Finally, the label reports rare cases of neuroleptic malignant syndrome,
which is potentially fatal.
FN5. Tarsy D, Indorf G. Tardive tremor due to metoclopramide. Mov Disord
2002 May;17(3):620-1.
FN7. Lu ML, Pan JJ, Teng HW, Su KP, Shen WW. Metoclopramide-induced
supersensitivity psychosis. Ann Pharmacother 2002 September;36(9):1387-90.
FN8. Friend KD, Young RC. Late-onset major depression with delusions after
metoclopramide treatment. Am J Geriatr Psychiatry 1997;5(1):79-82.
FN9. Anfinson TJ. Akathisia, panic, agoraphobia, and major depression
following brief exposure to metoclopramide. Psychopharmacol Bull
2002;36(1):82-93.
The exact incidence of metoclopramide-induced tardive dyskinesia remains unclear,
but based on the medical literature long-term use of metoclopramide should be
avoided to prevent persistent and disabling movement disorders. Two factors
potentially mitigating the risk of tardive dyskinesia in the case of Myzan(R) are
the recommended intermittent use and the relatively low dose of metoclopramide.
Since metoclopramide-induced tardive dyskinesia is associated with prolonged use
of the drug, it was hoped that an examination of utilization patterns and duration
of use of metoclopramide products currently on the market would shed light on the
extent of off-label prescribing and use of metoclopramide, which could have
implications for a new metoclopramide-contained product. Specifically, the
duration of use of metoclopramide is important in determining cumulative drug
exposure, extent of adherence to the label recommendations and the risk of such
adverse drug events in metoclopramide users. Duration of use longer than
recommended in the label may suggest that the risk of Myzan(R) if approved, may
outweigh the benefit of the drug. Any excess risk from prolonged exposure to
metoclopramide should be avoided, especially when alternative migraine products
are available in the market.
METHODS
I. National Utilization Patterns of Metoclopramide
Outpatient use of metoclopramide was derived from several data sources. IMS
Health, National Prescription Audit PlusTM (NPA PlusTM) was used to examine
dispensed prescriptions of metoclopramide in the U.S. focusing on the outpatient
setting. IMS Health, National Disease and Therapeutic IndexTM (NDTITM) was further
used to describe the prescribing patterns of metoclopramide among physicians,
patient characteristics, and the indications for which the drug was prescribed.
The data sources for this analysis are described in detail below.
IMS HEALTH, NATIONAL PRESCRIPTION AUDIT PLUSTM (NPA PLUSTM) NPA PlusTM measures
the retail dispensing of prescriptions, or the frequency with which drugs move out
of retail pharmacies into the hands of consumers via formal prescriptions. These
retail pharmacies include chain, independent, food store, mail order, discount
houses, and mass merchandiser pharmacies, as well as nursing home (long-term care)
pharmacy providers. Information on the specialty of the prescribing physician can
also be collected, except for in the long-term care and mail order pharmacy
settings.
The number of dispensed prescriptions is obtained from a sample of approximately
22,000 pharmacies throughout the U.S. and projected nationally. The pharmacies in
the database account for approximately 40% of all pharmacy stores and represent
approximately 45% of prescription coverage in the U.S.
Data for this analysis include prescriptions dispensed for metoclopramide products
for the last ten-year period from January 1, 1995 - December 31, 2004, inclusive.
IMS HEALTH, NATIONAL DISEASE AND THERAPEUTIC INDEXTM (NDTITM) The National Disease
and Therapeutic IndexTM (NDTITM) is an ongoing survey designed and conducted by
IMS Health to provide descriptive information on the patterns and treatment of
disease encountered in office-based practice in the continental U.S. The data are
collected from a panel of roughly 2,000 - 3,000 office-based physicians who
complete and submit a survey of their practice patterns to IMS Health for two
consecutive days per quarter. These data may include profiles and trends of
diagnoses, patients, drug products mentioned and treatment patterns. The data are
projected nationally to reflect national prescribing patterns.
NDTITM uses the term drug uses for mentions of a drug in association with a
diagnosis during an office-based patient visit. This term may be duplicated by the
number of diagnosis for which the drug is mentioned. It is important to note that
a drug use does not necessarily result in prescription being generated. Rather,
the term indicates that a given drug was mentioned during an office visit.
For this analysis, we examined annual mentions of metoclopramide during
office-based physician visits during the last ten-year period from January 1, 1995
- December 31, 2004, inclusive.
II. Duration of Treatment for Metoclopramide
Paid prescription claims for a 36-month period of time from Caremark (Dimension Rx
TM) provided a longitudinal data source for estimating the duration of therapy
with metoclopramide. The data source for this analysis is described in detail
below.
CAREMARK, DIMENSION RXTM
Caremark is one of the largest pharmacy benefit manager (PBM) companies in the US,
currently covering over 70 million lives, and processing over 545 million
prescription claims annually. FDA had access to Caremark's Dimension RxTM database
consisting of a subset of total Caremark paid claims, representing 350 million
claims per year for prescriptions filled in 57,000 pharmacies across the country.
These participant-level claims are available for 36 rolling months at the time
they are accessed on-line. Participants whose claims are processed by Caremark are
covered under various types of insurance plans, including health maintenance
organizations (HMOs), employers' self-insured health plans, selected managed care
plans, and other selected traditional health insurers. Caremark's Dimension RxTM
system includes participants from all 50 states including special populations such
as the elderly, children, and women of childbearing age. The representativeness of
those included in Caremark's Dimension RxTM system to all persons receiving
dispensed prescriptions in the U.S., however, is not known.
For this analysis, prescription claims for metoclopramide generic products in the
Caremark Dimension RXTM system were examined for all persons aged 1-99 years from
January 1, 2002 to December 31, 2004, inclusive.
Study Population
The study population consisted of individuals aged 1-99 years who were
continuously enrolled in the Caremark PBM and had prescription claims for any
metoclopramide products between January 1, 2002 to December 31, 2004. The original
dataset contained 611,951 prescription claims for metoclopramide for 209,641
individuals during the study period. This dataset was examined for data
inconsistency and accuracy. Claims were excluded from this analysis if they had
either quantity or days' supply that was either "unknown" or was assigned a value
of zero. Of the remaining claims, individuals were excluded if they had multiple
claims with a greater than two-year lag time in patient age between two
consecutive prescriptions. Also, individuals were excluded if the value of gender
assigned was not consistent across their associated claims or if it was assigned
as "unknown". Individuals for whom the drug quantity in the metoclopramide claims
was not a whole number were removed. As a result of these exclusions, less than 2%
of the original claims and the original individuals were removed. The final
dataset consisted of 601,226 prescription claims for metoclopramide for 207,688
patients. From this data set we extracted a subset of individuals that had claims
for metoclopramide tablets only. Our analysis pertains to this subset of 200,907
individuals only. The design and the exclusion criteria for the study sample are
illustrated in the diagram scheme in Figure 1.
TABLE
Source:
1995-1996 IMS Health, National Prescription Audit PlusTM book data- Dec 1996, USC
01200-33390
1997-1999 IMS NPA CD-ROM Dataview Analyzer Total Market CD-ROM
Years 2000 - 2004, extracted February 2005.
Measurements
Claims data for this subset of the study population were used to determine an
episode of metoclopramide therapy for an individual. The episode of therapy was
constructed based on an algorithm that was previously developed in DSRCS. We
defined an episode of therapy as one or a series of consecutive prescription
claims for metoclopramide for an individual with no more than a 30-day lag between
the dispensing date of a new prescription and the ending date of the preceding
prescription. The ending date of a prescription was calculated for each
prescription by adding the number of days' supply to the claim date. A successive
prescription for the same patient, that was dated more than 30 days after the
ending date of the preceding prescription, was not considered part of the previous
episode and was considered as the onset of a new episode of therapy.
The duration of each episode of therapy was calculated for each episode by summing
up the total days' supply from claims that constituted that episode. The time lag
between the claims was not taken into account in calculating duration of episode
of therapy. The duration of the longest episode for metoclopramide was determined
for each patient. Cumulative exposure to metoclopramide during the study period
was estimated for each patient by summing the duration of each episode.
The age of the patient was calculated from the date of birth and was defined as
the age at first prescription as we allowed individuals to age during the study
period or to have age difference of +-2 years from claim to claim.
We determined the region in which the prescription claims for metoclopramide were
filled by categorizing the pharmacy state information into four regions as
designated by the U.S. Census Bureau: Northeast, Midwest, South, and West.[FN10]
We created another category "other" which included Puerto Rico, Virgin Islands,
Northern Mariana Islands, and Guam.
FN10. US Census Bureau 2005;Available at: URL: http://
www.census.gov/popest/geographic/. Accessed: June 6, 2005.
Analytic Methods
All analyses were performed using SAS statistical software version 8.2 (SAS
Institute Inc, Cary, NC). The following sensitivity analyses were performed to
examine the robustness of the assessment of duration of the longest episode of
therapy as well as the cumulative exposure:
* Excluding individuals with extreme values of days'supply. In this analysis we
only included individuals that had 100 days' supply or less. The 100-day cut-off
was used as many health insurance companies do not cover the cost of prescriptions
that exceed 100 days' supply. We recalculated the duration of the longest episode
of therapy and the mean cumulative exposure for these individuals.
* Redefining episode of therapy. We used the original study population and
redefined the episode of therapy based on no more than a 10-day lag between the
dispensing date of a new prescription and the ending date of the preceding
prescription. We recalculated the proportion of individuals with an episode of
metoclopramide therapy longer than 90 days.
* Redefining the total duration of episode of therapy. In the third analysis, we
used the original study population and redefined the total duration of therapy for
an episode to include the time lag between prescription claims.
RESULTS
I. National Utilization Patterns for Metoclopramide
A. Prescriptions dispensed for Metoclopramide, 1995-2004
Over the ten-year period from 1995 to 2004, there was a two-fold increase in the
number of outpatient prescriptions dispensed for metoclopramide in the U.S.
(Figure 2). At the beginning of that period in 1995, the number of prescriptions
dispensed for metoclopramide was almost 3.5 million. It declined to less than 2
million prescriptions dispensed annually from 1997 to 1999, and then increased
abruptly in 2000 and reached over 7 million prescriptions dispensed during 2004.
The increase in dispensed prescriptions beginning in 2000 is likely due to the
withdrawal of cisapride, a treatment for severe nighttime heartburn, from the U.S.
market in March 2000.
Almost 90% of the metoclopramide products dispensed were in an oral tablet dosage
form; over 10% were in the syrup dosage form and less than 1% were injectables in
each of the one-year periods from 2000 to 2004.
TABLE
B. Visit Characteristics of Metoclopramide, 1995-2004
During the ten-year period from 1995 through 2004, NDTITM recorded an average of
1.4 million mentions of metoclopramide annually during physician-office visits
made in the U.S. (data not shown). Similar to the prescription data, visits with
mentions of metoclopramide decreased from 1.5 million visits during 1995 to 1.0
million visits during 1999, and increased back again to 1.5 million visits during
2004.
Oral Metoclopramide and Age
Over one third of the annual visits in the U.S., in which metoclopramide was
mentioned, were made by patients ages 35-64 years old during the ten year period
from 1995 to 2004 (Figure 3). While elderly patients (aged 65 years and older)
accounted for about a third of the visits in which metoclopramide was mentioned
from 1995-1998, their proportions declined steadily in the subsequent years. In
2004 they accounted for only 20% of the visits in which metoclopramide was
mentioned. The proportions of the visits in which metoclopramide was mentioned in
patients aged 17- 34 years remained steady from 1995 to 2004. As opposed to the
elderly population, in which we detected a decline in the visits in which
metoclopramide was mentioned, the proportion of these visits in the pediatric
population (aged 0-16 years) increased steadily throughout the ten-year period of
the study from 10% of all visits in which metoclopramide was mentioned in 1995 to
more than 30% in 2004.
TABLE
Oral Metoclopramide and Gender
From 1995 to 2004, on average over 55% of the annual visits in which
metoclopramide was mentioned were made by females (Figure 4).
TABLE
Oral Metoclopramide and Indication for Use
During the last five-year period from 2000 through 2004, more than 50% of
metoclopramide uses were for gastroenterology-related diagnoses (data not shown).
"Esophagitis" (ICD-9 code 530.1) was ranked the first most common diagnosis for
metoclopramide mentions, accounting for over 40% of metoclopramide mentions (Table
1). The two top indications for metoclopramide drug uses after "esophagitis" were
"post-surgical examination" (v670), with more than 6% of metoclopramide mentions
and "nausea and vomiting" (ICD-9 code 787.0), with about 5% of mentions. Other
diagnoses each accounted for less than 5% of the total metoclopramide mentions,
including "gastroparesis" (536.3 or 250.6 - diabetes with neurological
manifestations), which accounted for less than 1% of all indications mentioned in
2004. In general, diagnoses associated with mentions of metoclopramide showed
relatively little change across the years. Notably, "migraine unspecified" was a
diagnosis consistently associated with mention of metoclopramide since 2000,
although in less than 2% of visits. Given the small number of visits this
projection is based upon, it should be viewed with caution.
Table 1. Diagnoses Associated with Mentions of Metoclopramide During Office-Based
Physician Visits, IMS Health, National Disease and Therapeutic IndexTM, 2000 -
2004
TABLE
During the ten-year period from 1995-2004, the proportion of office-based visits
for the pediatric population increased (from 10% in 1995 to 30% in 2004) along
with the increase in the "esophagitis" diagnosis (from 35% in 1995 to 65% in 2004)
(data not shown). As of 2000, "esophagitis" (ICD-9 code 530.1) accounted for more
than two-thirds of all pediatric visits in which metoclopramide was mentioned
(Table 2). "Feeding problems newborn" diagnosis (ICD-9 code 779.3) was ranked the
3rd in that population in 2004, accounting for less than 5% of the total pediatric
visits in which metoclopramide was mentioned during the five-year period from
2000-2004.
In younger adults aged 17-34 years, while "esophagitis" diagnosis (ICD-9 code
530.1) was ranked the top diagnosis for metoclopramide mentions in the last
five-year period from 2000-2004, it accounted on average for only one-fifth of
metoclopramide mentions during that period. It should be noted however, that in
the five years before 2000, "esophagitis" diagnosis accounted for less than 10% of
metoclopramide mentions, with the exception of year 1996, during which
esophagitis" diagnosis accounted about 20% of metoclopramide mentions. Thus,
"esophagitis" diagnosis, as a reason for use of metoclopramide, seems to be
relatively increasing since 2000 in that age group.
In older adults aged 35-64 years as well as in the elderly (aged 65 years and
older) "esophagitis" diagnosis (ICD-9 code 530.1) was ranked the top diagnosis for
visits during 2000-2004, in which metoclopramide was mentioned, accounting for
over 30% of metoclopramide mentions each of these age groups. This diagnosis seems
to have increased compared to previous years (1995-1999), during which it
accounted for only about 20% of the metoclopramide mentions.
Table 2. Top 5 Diagnoses Associated with Mentions of Metoclopramide During
Office-Based Physician Visits by Age Groups, IMS Health, National Disease and
Therapeutic IndexTM, 2000 - 2004
TABLE
Physician Specialty
During 2004, internal medicine was the most frequent prescribing specialty,
accounting for over 1.2 million (22%) dispensed prescriptions for metoclopramide,
followed by family practice (18%), gastroenterology (13%), osteopathic medicine
(8%) and pediatric practice (5%) (Figure 5). Together, these specialties accounted
for two-thirds of the dispensed prescriptions for metoclopramide during 2004. In
general, prescribing patterns for this product showed no change across provider
specialties from year 2000-2004 (data not shown).
TABLE
I. Duration of Therapy for Metoclopramide Tablets
A. Characteristics of the Study Population
Over the three-year period from January 1, 2002 through December 31, 2004, there
were 572,205 prescription claims for metoclopramide tablets for 200,907 patients
who were continuously eligible in the Caremark PBM system during that time period.
The characteristics of the study population are shown in Table 3. The mean age of
the study population at first prescription was 58 years old (standard deviation
(SD) =17.9). Prescriptions claims for metoclopramide were almost exclusively
filled by adults and 50% of the claims were filled by persons aged 35-64 years.
About two-third of the claims were filled by females and only one-third by males.
Approximately 50% of the claims were filled in pharmacies located in the southern
region in the U.S.
Table 3. Demographic Characteristics of the Study Population with Prescription
Claim for Metoclopramide tablets in Caremark PBM, Caremark's Dimension Rx,
2002-2004
TABLE
B. Duration of Metoclopramide Use
The average number of prescriptions for metoclopramide per patient was 2.8
(SD=4.7) and ranged from one prescription claim up to a total of 78 prescription
claims per patient. Two-thirds of patients (66%) had only one paid prescription
claim for metoclopramide. The average days' supply for a prescription claim was 29
days (SD=22.7; range: 1-816 days). The number of episodes ranged from 1 to 20
episodes and about 78% of the study population had only one episode of therapy
over the three year period. The average duration of the longest episode for
metoclopramide use was 61 days (SD=143.3) and ranged from 1 day to over 2000 days,
and thus, could have gone beyond the period of observation. The length of the
longest episode of therapy for most patients (87%) varied from 1 to 90 days, yet
about 13% of the patients appeared to have received prescriptions for
metoclopramide for a period longer than 90 days (Table 4).
Table 4. Duration of the Longest Episode with Metoclopramide Tablets in Caremark
PBM, Caremark's Dimension Rx, 2002-2004
TABLE
During the three-year study period, the average cumulative therapy with
metoclopramide was 83 days (SD=179.0) and ranged from 1 day to a total of 2200
days, which was beyond the period of observation. For the majority of the patients
in the study (68%), the length of cumulative therapy for metoclopramide was 1 to
30 days (Figure 6). A cumulative therapy period for metoclopramide longer than 90
days was recorded for almost 20% of the patients. As the total cumulative therapy
for metoclopramide increased, the frequency of patients decreased; a decrease from
10% of patients treated for 31-60 days to 5% treated for up to 6 months. Yet, the
frequency of patients, who had cumulative therapy for metoclopramide for more than
6 months, was over 10%.
TABLE
Sensitivity Analyses
* Redefining episode of therapy as no more than a 10-day lag between the
dispensing date of a new prescription and the ending date of the preceding
prescription.
The average duration of the longest episode for metoclopramide use was 48 days
(SD=112.2). About 10% of the patients appeared to have received prescriptions for
metoclopramide for a period longer than 90 days.
* Excluding individuals with more than 100 days'supply.
The number of individuals with more than 100 days' supply was 562. After excluding
these individuals and all their associated claims, the average duration of the
longest episode for metoclopramide use was 60 days (SD=112.2). About 12.6% of the
patients appeared to have received prescriptions for metoclopramide for a period
longer than 90 days. The average cumulative therapy was 81 days (SD=175.3) and the
frequency of patients, who had cumulative therapy for metoclopramide for more than
6 months, was 11%.
* Redefining the total duration of episode of therapy by adding the time lag
between prescription claims.
The average duration of the longest episode for metoclopramide use was 68 days
(SD=158.4). About 15% of the patients appeared to have received prescriptions for
metoclopramide for a period longer than 90 days.
DISCUSSION
The increase in the number of prescriptions dispensed for metoclopramide
(following the withdrawal of cisapride from the market in 2000) along with the
longer than recommended duration of use found in this study was substantial. These
results suggest that even with metoclopramide label restrictions on the duration
of use and with its well-recognized adverse events, particularly extrapyramidal
movement disorders, physicians as well as patients did not follow the labeled
recommendation on the duration of therapy. These results may further suggest that
a similar label recommendation for a new combination product that contains
metoclopramide is not likely to be followed, especially if the product contains
naproxen, which is frequently used without restriction on duration of use.
Results from this study are supported by previous finding reported in the
literature. An increase in metoclopramide use during the six-month period after
the withdrawn of cisapride from the market was observed.[FN11] This increase was
mild and not proportional to the substantial decline observed in cisaprade use,
suggesting that cisapride was not available on the market and prescribers
responded soon to the market change by seeking substitutions. Nonetheless, this
market shift toward an increase in metoclopramide use has potential clinical
implications on the incidence of adverse drug reactions, especially when the risk
of adverse reactions increases with duration of treatment and cumulative dose.
Shaffer et al. has recently reported a series of 87 cases of
metoclopramide-associated tardive dyskinesia extracted from FDA's AERS database
with remarkably long duration of use (mean=753+-951 days).[FN6] Our data also
support a prolonged duration of use (more than 90 days) in over 10% of individuals
using metoclopramide, predisposing them to an increased risk of adverse reactions.
FN11. Glessner MR, Heller DA. Changes in related drug class utilization
after market withdrawal of cisapride. Am J Manag Care 2002
March;8(3):243-50.
FN6. Shaffer D, Butterfield M, Pamer C, Mackey AC. Tardive dyskinesia risks
and metoclopramide use before and after U.S. market withdrawal of cisapride.
JAm Pharm Assoc (Wash DC) 2004 November;44(6):661-5.
Our sensitivity analyses did not substantially change our estimates of duration of
the longest episode, duration of cumulative therapy, or the proportion of patients
taking metoclopramide for more than 90 days.
It is noteworthy that short duration of therapy for individuals who were taking
the drug in the beginning or at the end of the observation period could derive
from the boundaries of the study period, as patients could have started an episode
of therapy prior to the beginning of the observation period or concluded it after
the end of the observation period. On the other hand, episodes with prolonged
duration of therapy could also result from errors in the data recorded for days'
supply. It should be noted that the operational definition used to define duration
of therapy could also account for such extreme measurements. The duration of each
episode of therapy was calculated for each episode by summing up the total days'
supply from claims that constituted that episode, regardless of the filled date.
We found, however, that a small number of metoclopramide claims were filled on the
same day, which might suggest that patients were taking a higher dose rather than
using the drug for prolonged period. The number of such claims was small (388 in a
dataset of 572,205, < 0.1%) and thus unlikely to substantially change our
estimates.
Findings from this consult should be interpreted in the context of the known
limitations of the databases used. NPA PlusTM data provide an estimate of the
total number of prescriptions dispensed in the U.S. The data, however, do not
include historical demographic information, such as age and gender and do not
cover all prescriber specialties from mail order and long-term care channels. NDTI
TM data, on the other hand, provide estimates of patient demographics and
indications for use of medicinal products in the U.S. Yet due to the sampling and
data collection methodologies, the small sample size can make these data unstable,
particularly in sub-categories when use has low prevalence.
Caremark's Dimension RxTM data derive from a large PBM and may not be nationally
representative as patterns across populations may depend on prescription drug
coverage. Yet, its large sample size provides a useful description of prescription
drug use in the U.S. for a large proportion of the population with prescription
drug coverage. Nonetheless, these results should be interpreted with caution. For
example, the elderly are known to be underrepresented in databases dependent upon
prescription drug coverage.
While not comparable to each other in terms of generalizability, we found
differences in the proportion of pediatric use for metoclopramide between NDTITM
(30%) and Caremark's Dimension RxTM (<2%). Some of this difference can derive from
the fact that NDTITM captures drug mentions, which includes but is not limited to
written prescriptions during office visits, while Caremark captures prescriptions
filled. Another source that could account for this difference is the dosage form
of metoclopramide in these databases; NDTITM included all oral forms, while our
analysis of Caremark dataset was limited to metoclopramide tablets only. It is
also noteworthy that the majority of the metoclopramide mentions during pediatric
visits in NDTITM dataset were for children up to 3 years old and mentions for this
age group consisted of about one-fifth or one-fourth of all metoclopramide
mentions during office visits from 2002-2004. In the Caremark dataset, infants
under one year of age were excluded but could have accounted for a significant
proportion of metoclopramide claims excluded, as use could be higher in the first
year of life. Finally, the continuous enrollment eligibility criteria for the
entire study population could probably account for much of the difference, as
large proportions of young children would not meet the eligibility criteria and
would therefore have been excluded from the Caremark analysis.
CONCLUSION
Results from this study indicate that the number of prescriptions dispensed for
metoclopramide, which was declining from 1995 to 1999, has more than tripled in
the five-year period following the withdrawal of cisapride from the market in
March 2000. In addition to this increase in the U.S. in recent years, data from an
insured population source showed that over 10% (10- 5%) of metoclopramide users
had claims suggestive of use for a period longer than the three months recommended
in the label. Moreover, total cumulative therapy with metoclopramide longer than 6
months was found in about 10% of the study population.
