pureheart

Reported Medication Errors Associated With Methotrexate

Posted 07/20/2004

Thomas J. Moore; Christopher S. Walsh; Michael R. Cohen

Abstract and Introduction Abstract

Purpose: Medication errors reported to FDA as adverse events in which
methotrexate was identified as a possible contributor were studied.
Methods: All adverse-event reports submitted to FDA between November
1997 and December 2001 indicating potential medication errors
involving methotrexate were ****yzed to determine the indication for
use, the type of error, and the point in the medication-use process
where the error occurred. Results: A total of 106 cases of reported
medication errors associated with methotrexate were identified,
including errors resulting in 25 deaths (24%) and 48 other serious
outcomes (45%). The most common types of errors involved confusion
about the once-weekly dosage schedule (30%) and other dosage errors
(22%). The most frequently involved indication for use was rheumatoid
arthritis (42%). Of the errors, 39 (37%) were attributable to the
prescriber, 21 (20%) to the patient, 20 (19%) to dispensing, and 18
(17%) to administration by a health care professional. Conclusion: A
review of medication errors involving methotrexate revealed that
errors occurred during all phases of use, often resulted from
confusion about dosage, and often caused death or other serious
adverse effects. Introduction

Methotrexate is a folic acid antagonist used in the treatment of many
medical conditions, including rheumatoid arthritis,[1] psoriasis,[2]
and ectopic pregnancy.[3] It has also been proven effective against
several forms of cancer, including acute lymphoblastic leukemia,[1]
osteosarcoma,[4] and non-Hodgkin's lymphoma.[5] Doses of methotrexate
(as the sodium salt) for rheumatoid arthritis can be as low as 2.5 mg,
while 15 g/m2 can be used
(along with leucovorin rescue therapy) for osteosarcoma—a
10,000-fold difference. Scheduling of doses also varies widely, from
cyclical protocols for cancer chemotherapy to weekly for rheumatoid
arthritis and sarcoidosis.[6,7] Adding to this complexity,
methotrexate can be administered orally, intravenously,
intramuscularly, or intrathecally.

Since folic acid plays a major role in cell division, its inhibition
is associated with major toxicity to blood cells, the oral mucosa,
hepatocytes, and lung and fetal tissue. Severe blood-cell toxicity is
usually manifested as anemia, leucopenia, or thrombocytopenia and has
been reported in up to 36% of lymphoma patients receiving
methotrexate.[7] Oral ulceration, which can occur at the lower end of
the methotrexate dosage spectrum, was seen in 11 of 29 patients
treated for rheumatoid arthritis.[8] Methotrexate-related
hepatotoxicity, which has occurred in
10% of sarcoidosis patients,[7] is not well correlated with
liver-function test findings; biopsy is the only accurate method of
assessing liver status. Pneumonitis, which is related to the
***ulative dose, was reported in a literature review to occur in 5%
of patients undergoing long-term therapy.[9] Methotrexate is such a
potent teratogen that the manufacturer recommends that both male and
female partners avoid exposure during pregnancy.[10]

Frequent serious medication errors involving methotrexate have been
reported, and some patients have died.[11-13] The errors have
involved some aspect of the dosage regimen. Three recent fatal cases
illustrate the hazards of failing to observe the weekly dosage
schedule for rheumatoid arthritis. In one of these cases, a patient
misunderstood the directions for use and took methotrexate 2.5 mg
every 12 hours for six consecutive days instead of 2.5 mg every 12
hours for three doses each week.[14] In the second case, a patient
misread the directions on a prescription bottle and took 10 mg every
morning instead of every Monday.[14] In the third case, a physician
prescribed methotrexate 15 mg daily instead of weekly, and the patient
received nine doses.[15]

Errors like these have been reported to organizations like the
Institute for Safe Medication Practices (ISMP) and the United States
Pharmacopeia, which in turn automatically notify FDA.[16] To evaluate
the overall risk posed by methotrexate, errors need to be viewed in
aggregate, along with their associated causes, to determine the most
effective prevention strategies.

We studied medication errors reported to FDA as adverse events in
which methotrexate was identified as a possible contributor.

Methods

The data for this study were extracted from the records of FDA's
Adverse Event Reporting System (AERS).[17] AERS contains reports of
medication errors and adverse events suspected to be associated with
the administration of marketed prescription drugs and biological
products (except vaccines). Reports to the system are spontaneous and
voluntary, except that drug manufacturers are required to submit
reports about those adverse events of which they become aware.[18] In
2001, FDA received 272,000 adverse-event reports.[19]

FDA publishes quarterly data extracts of all reports received by
AERS.[20] These extracts include information about patient
demographics, the drug or drugs administered, the kinds of injuries or
other adverse events that resulted, and the medical terms used to
describe the reactions that occurred. However, the original report is
withheld to prevent accidental release of personal information; in its
place, FDA reviewers select terms from the Medical Dictionary for
Regulatory Activities (MedDRA), an online source.[21,22]

From all reports entered into AERS from its inception in November 1997
through December 2001, we extracted every report naming methotrexate,
a brand name for methotrexate, or a likely spelling variant. From the
methotrexate reports, we selected reports in which methotrexate was
listed as a suspect drug (rather than being listed as given
concomitantly with other therapy) and that contained a MedDRA term
suggesting a medication error may have taken place. The dictionary
terms used to identify reports of a possible medication error were
medication error, drug maladministration, drug interaction NOS (not
otherwise stated), drug level NOS above the****utic, accidental
overdose, accidental exposure, and drug interaction potentiation.

Each report submitted to FDA permits the reporter to specify zero,
one, or more of the following adverse-event outcomes: death, birth
defect, disability, hospitalization, threat to the patient's life,
intervention to prevent harm, and "other." Some reports do not
indicate an outcome. We put each report into one of four mutually
exclusive categories: (1) death, (2) disability or birth defect, (3)
serious outcome (nonfatal), and (4) other or none stated. The
category of serious outcome includes the FDA terms hospitalization
(initial or prolonged), life-threatening, and required intervention
to prevent harm. If an event could be put into more than one category,
it was placed in the most significant one.

Because the database's abstracts do not contain enough detail for
medication errors to be assessed, the full reports for possible cases
of medication errors were requested from FDA under the Freedom of
Information Act. These full reports formed—after FDA removed
personal information—the basis of this study.

We defined a medication error as "any preventable event that may cause
or lead to inappropriate medication use or patient harm."[23] After
the elimination of duplicate reports (such as two manufacturers
submitting published case reports about the same event), each case
was reviewed to establish whether a medication error had occurred and
was placed into one of four categories. "Definite" meant the case
report contained an unambiguous direct statement by a reporting
individual indicating that a medication error had occurred. "Probable"
meant that unambiguous information indicated that a medication error
had occurred (e.g., there was a statement that methotrexate for
rheumatoid arthritis had been administered or taken daily instead of
weekly, regardless of whether the reporter ********ly stated this was
an error). "Possible" meant there was information indicating that a
medication error may have occurred but that there could be other
explanations as well. All other cases were classified as "unlikely,"
even if the patient suffered injuries that may have been related to
administration of the drug, because information indicating an error
was absent.

For each case we assessed the failure mode, or the point in the
medication-use process where the error had occurred. The failure-mode
categories were prescribing (including writing the order), dispensing,
administering, and consuming and correspond to the major actors in
the process: physician, pharmacist, nurse or caregiver, and patient.
In addition, independent of the failure mode, we identified the type
of error that had been made: confusion between daily and weekly
administration, other dosage error, wrong drug, wrong route (e.g.,
intrathecal instead of intravenous), contraindication, labeled
interaction, and other. If a report indicated that more than one type
of error may have occurred, the most significant applicable error
category was assigned. Drug interactions classified as possible
medication errors were limited to those identified ********ly in the
prescribing information.[6] Finally, we recorded the indication for
which the drug was prescribed, either using the information in the
"indication" block in the report or extracting it from the report
narrative.

The data were ****yzed with the SPlus 6.1 statistical package
(Insightful Corp., Seattle, WA). The error reports were a population
and do not have confidence intervals or standard errors.

Results

From 156 cases of adverse events identified in the AERS computer
database, we classified 106 as medication errors (Table 1). Cases
that were eliminated included those involving duplicate copies of the
same literature report submitted by different manufacturers, reports
of possible new drug interactions about which no notice appeared in
the product labeling, and adverse reactions that may have been related
to the drug but for which no specific medication error was described
or could be deduced. Patients for whom medication errors were reported
had a mean age of 54 years, and 68% of them were women

The medication errors resulted in 25 deaths (24%). Outcomes classified
as "other" included various skin abnormalities, depressed platelet
counts, and slowed wound healing.

The most frequent indications for the drugs involved in errors were
rheumatoid arthritis (44 cases [42%]), psoriasis
(13 [12%]), and cancer (9 [9%]). The many medical uses of methotrexate
were also reflected in a wide spectrum of other indications for which
at least one medication error was reported, including sarcoidosis,
polyarthitis, ectopic pregnancy, and lupus erythematosus.

Overall, 52% of the reported errors involved some form of overdose,
with the most common problem being taking the drug daily instead of
weekly (32 cases [30%]) (Table 2). The potential for confusion in
these error cases was increased by polypharmacy; 73% of the patients
were taking 2 or more drugs (mean, 3.5 drugs).

A second type of error involved problems that may occur with any
medication but produce serious adverse effects with methotrexate.
These included instances of the wrong medication being given, failing
to note a contraindication, and using the wrong route of
administration (giving methotrexate intrathecally when another route
was intended).

The third major type of error involved drug interactions identified in
the product labeling. The most common of the interactions that led to
harm involved the coadministration of another drug with antifolate
activity, especially trimethoprim– sulfamethoxazole.

The errors occurred in all four steps of the medication-use process:
prescribing (39 cases [37%]), dispensing (20 [19%]), administration
(18 [17%]), and consuming (21 [20%]). Errors in consuming evolved
primarily out of patient confusion about the weekly dosage schedule
(17 of 21 cases). The errors in prescribing were based on
contraindications (6 of 39 cases), labeled interactions (18 of 39),
dosage errors (6 of 39), and the wrong drug (2 of 39).

Discussion

Our study found that methotrexate is associated with significant
preventable medication errors that warrant careful measures to
improve patient safety. While some of the errors (those involving a
contraindication, dispensing of the wrong drug, or use of the wrong
route of administration) could happen for any drug and be hazardous,
the distinctive weekly or cyclical dosage scheme and low the****utic
index constitute the most clearly identifiable risks of methotrexate
and accounted for the largest number of preventable serious injuries
and deaths.

The methotrexate error data have several limitations that should be
considered in interpreting the study results. Because the error
reports were voluntary and spontaneous, they provided little or no
indication of the true number of errors that might be occurring
annually in the United States or worldwide. We know of no studies that
would provide a basis for estimating a reporting rate for medication
errors. Studies of all types of adverse events suggest that there is
1 report for every 10 to 20,000 occurrences.[24,25] Therefore, we
suspect that errors were underreported in our study by some unknown
amount.

The existence of an adverse-event report does not establish that the
suspected drug caused the event; it only raises the possibility of an
association. In the methotrexate error reports, the causal link was
reasonable but still not certain. The reports contained either an
******** statement that an error had occurred or specific factual
information indicating an error as defined by the study criteria.
However, the reports were voluntary and varied widely in the amount
of detail and length of follow-up.

An in-depth investigation involving medical records and interviews
with participants might conclude that some cases classified as
medication errors (especially among the 19 cases classified as
possible) might not have constituted preventable errors. However,
additional error cases not counted in this study might be identified
in a more detailed study of the cases that were eliminated as being
unlikely or among the thousands of adverse-event cases without a
specific term suggesting a possible error.

These findings underscore ISMP's safe practice recommendations for
methotrexate,[14] notably the need to talk to patients about the drug


addition, ISMP recommends dispensing, when appropriate, the weekly
dosage pack (Wyeth's Rheumatrix Dose Pack); it provides an additional
tool for avoiding confusion over the dosage schedule.

Electronic alerts in prescriberorder-entry systems and pharmacy
computer systems can signal clinicians to check for potential errors
whenever oral methotrexate is ordered to be given more than once
weekly. Serious drug interactions can be detected in the same manner.
Dispensing methotrexate in the weekly dosage pack and communicating
effectively with patients about unusual dosage regimens can reduce
injuries. Clinicians should encourage feedback to ensure that the
patient understands the weekly dosage schedule and that the medication
should not be used "as needed" for symptom control. Patients should
be given clear written instructions that name a specific day of the
week for taking the drug, that emphasize the weekly—not
daily—dosage schedule, and that explain the possible outcomes
of a dosage error. Conclusion

A review of medication errors involving methotrexate revealed that
errors occurred during all phases of use, often resulted from
confusion about dosage, and often caused death or other serious
adverse effects.


http://www.medscape.com/viewarticle/482389?src=mp