Similar Sounding Drug Names

Cox AR. Marriott JF. Safety First. Pharmaceutical Marketing, February 2002

Patient Safety Problems

Look-alike medication names cause up to 25 per cent of errors. The FDA reported last month that of 400 deaths, 16 per cent involved name mix-ups. Drugs with similar-sounding names have led to deaths. What can be done to prevent any more mishaps?

In 1998 the Peruvian government passed a law banning the use of children’s names considered ridiculous, offensive, or contrary to religious beliefs. Politicians felt this necessary because of a craze for naming children withbizarre names, including the Spanish words for cutthroat and circumcision, or even chemical symbols like H20. Choosing the name of a child can be difficult, but you are unlikely to endanger people’s lives with your final choice. This is unfortunately not the case when naming drugs.

Over the recent years the dangers of medication errors have become increasingly recognised in both the lay and medical press. Medication errors can occur at any step in the supply of a medicine from prescribing, dispensing and administration.

The recent Audit Commission Report, A Spoonful of Sugar: Medicines Management in NHS Hospitals, stated that 11 per cent of patients on medical wards experience an adverse event and a third of those lead to greater morbidity or death.

12 per cent of these events were related to medicines use. In addition to the harm suffered by patients, adverse events cost the NHS £1.1 billion each year. The political will to change practice in the NHS with regard to medication errors has never been higher.

Two key government documents have been produced in the last two years. The first was An Organisation with a Memory, which acknowledged the culture of no-blame and systems failure analysis.

Next came Building a Safer NHS, which described the plan to establish the National Patient Safety Agency, the need to develop a learning culture, the need for further research, and the need for improved systems to investigate adverse events. One of the key government targets is to reduce the number of serious medication errors by 40 per cent.

It is thought that look-alike (orthographically similar) and sound-alike (phonetically similar) medication names cause up to 25 per cent of all medication errors and the US Food and Drug Administration (FDA) reported last month that of the 400 deaths on their records caused by medication errors, 16 per cent involved name mix-ups.

In December 2001, a warning concerning potential confusion of the epilepsy drug Lamictal (lamotrigine) with the antifungal drug Lamisil (terbinafine) was issued by Glaxo Wellcome. On at least 22 occasions, pharmacists in the US have dispensed the wrong drug; three people suffering seizures.This example demonstrates that a poor choice of name can cause continual problems throughout the product’s lifetime.

In 1997 the Committee on Safety of Medicines warned of confusion between
Lamictal and Lamisil. Since then warnings about the confusion have been issued
by Glaxo every twelve months. Two main areas need to be looked at.

Firstly, health professionals, regulatory authorities and the pharmaceutical
industry need to work together to identify and reduce the risk of existing
look-alike and sound-alike drug names. Secondly, rigorous methods should be used
to test the safety of proposed generic and brand names for new products.

So how many look-alike or sound-alike names are there? In 2001, the United
States Pharmacopeia released Use Caution, Avoid Confusion which highlighted 750
unique drug names that had been reported to their Medication Error Reporting
scheme, with hundreds of combinations of confusing names. Examples of look-alike
and sound-like drug combinations:

Some of these may not sound-alike when read out or look-alike when in print,
but when hand written or communicated verbally they can cause confusion. One
recent case highlighted was that of Avandia (rosiglitazone) and Coumadin
(warfarin) in the United States.

Although neither look or sound similar on this page, a handwritten
prescription was issued that had the appearance of Coumadin 4mg, but was in fact
a poorly written Avandia 4mg. The prescription was queried by a vigilant
pharmacist, who corrected the error.

One might think: “Why don’t health professionals just read prescriptions
or labels more carefully, surely it’s their responsibility to get it right?”
However, health professionals are only human. Ever locked your keys in the
house, asked guests what drink they wanted then come back with coffee when they
asked for tea? The same unconscious processes that cause these errors, occur in
professional activities.

Mistaken identity

There are two main types of errors, those made when planning an intentional
act, often with inadequate knowledge, are termed mistakes. Those occurring in
the execution of an act are known as ‘slips of action’ or a ‘lapse of
memory’.

For the majority of dispensing and prescribing errors, the type of error will
be the ‘slip or lapse’ category (eg, choosing a box of chlorpropamide
from the dispensary shelf instead of chlorpromazine).

The crucial point is that these ‘slips and lapses’ are defects of an
unconscious process and will occur no matter how well you train an individual.
The only way to prevent these ‘slips and lapses’ reaching patients is by
systems solutions, control of the environment and building in safety mechanisms
in the process which can capture errors before they reach the patient.

A second independent check on dispensed medicines, or the clinical check that
a pharmacist performs on a doctor’s prescribing, are examples of existing safety
mechanisms.

Familiarity and education is of some use, but as William James, a 19th
century psychologist put it: “Habit diminishes the conscious attention with
which our acts are performed.”

Michael Cohen, president of the Institute of Safe Medical Practices in the
US, echoes this by saying: “To prevent errors we must never rely solely on
one’s memory of problem name pairs”
. He advocates that healthcare
professionals and organisations routinely monitor information from error
reporting programmes, and apply it.

Careful formulary selection of products to avoid drug name confusions,
developing policies on verbal or telephone orders or interactive reminders in
computer systems being mechanisms used to prevent errors are put forward as
possible solutions.

The use of hand-written prescriptions is falling and computer-generated
prescriptions are becoming more common outside of hospitals. In future the
introduction of electronic transfer of prescriptions and patient data, and the
more widespread use of bar-coding may reduce some types of error.

class=icon height=164 alt=”Slips can be made on keyboards”
src=”Bhttp://www.blacktriangle.org/keyboard.jpg” width=129>

However, they are not a panacea and hand-written prescriptions will continue
to exist for the foreseeable future and completely paperless systems in
hospitals are still some way off.

One suggestion is that prescribers write the purpose of the medication on the
prescription to allow the pharmacist to use the diagnosis as a confirmatory
check. In addition, if patients are aware of the names of the drugs they take,
both brand and generic, then they could play a role in the prevention of
medication errors.

Shelf shouters

Industry and the regulatory authorities have an important role in monitoring
and publicising potential risks of similar sounding drug names, and it can be
hoped that the National Patient Safety Agency will contribute towards
recognising signals of similar name problems.

When a similar name problem does occur, attempts should be made to provide
solutions to prevent further errors. Health professionals should be
informed and supportive measures provided. For example, Glaxo made available
‘shelf shouters’ to differentiate Lamictal from similar drugs in pharmacies.

The Committee on Safety of Medicines (CSM) has recognised the risk of similar
sounding names. Last year, a working group on labelling and packaging of
medicines recommended that manufacturers should take steps to ensure that
different fonts, text size, and colour should be used to change the shape of a
drug name to emphasise the differences between similar drug names.

Last month, the FDA made similar recommendations to manufacturers of over 30
specific drugs, requesting that label changes be made to highlight confusing
names, by using differing coloured letters for parts of drug names or using a
mix of upper and lower case lettering, for example chlorproMAZINE and
chloproPAMIDE.

Making packaging of pharmaceutical products distinctive and making the drug
name prominent on the product’s packaging, as proposed by Patientpacks.com (see
link at end of article), may also play a part in reducing the chances of similar
drug names being confused.

However, to prevent problems with similar drug names in the future, care must
be taken when naming drugs initially. This is of importance with both the
generic and brand name.

What’s in a name?

Generic names are co-ordinated by the World Health Organisation who set up
the International Non-proprietary Names (INN) system in 1951. Proposed names are
put to WHO and the agreed names are published in various languages. Companies
can make objections within four months of the list publication after which it
becomes a recommended INN.

Modern generic names are often based on a stem representing the
pharmacological activity of the drug common to the group it most closely
resembles, with a distinctive, often abstract, prefix.

Brand names are arrived at in varying ways, often involving name-creating
agencies or through brainstorming within the marketing department of a company.
The chosen name is concerned more with the development of a solid brand image
and establishing competitive advantage in the market niche, rather than safety
issues.

For example, Zofran’s name was chosen to create a soft and reassuring image.
Often similarities with the generic name are deliberately avoided, to prevent
the product being too easy to prescribe generically.

Names are also checked against trademark registers and a screen for
suitability in differing markets to avoid a cultural faux-pas. The process is
slow and expensive and changing a brand name, once decided, is an enormously
expensive process.

The use of computerised databases and automated, quantitative measures to
analyse the orthographic similarity of drug names is a step forward. One test,
the Levenshtein distance can correctly predict 91 per cent of error pairs. An
example of how this works is given by comparing Lamictal and Lamisil. To
transform Lamisil into Lamictal, the S is changed to a C, the I is changed to an
A and the T is added.

Three edit operations are required, giving a Levenshtein distance between the
two names of three. The lower the distance, the more similar the drug names are.
A threshold for Levenshtein distances has been proposed as greater than
5 therefore this test may have predicted the problems associated with
Lamisil and Lamictal.

The CSM working group on labelling and packaging of medicines recommended
that computer technology should be used, in combination with expert review, to
ensure that similar names were not approved for use in the UK.

The FDA are also pioneering pre-marketing studies of drug names. Fake
prescriptions are issued to 120 volunteer doctors, nurses and pharmacists to
read. A third of drug names have failed to pass this hurdle owing to confusion.

The pharmaceutical industry takes pride in the quality assurance and safety
of its products and this should not be marred by the creation of a confusing
drug name. The primary purpose of the drug name must be for the clear,
unambiguous identification of the drug and marketing considerations of a new
drug should take a back seat to patient safety.

Pharma should take care when naming its children. Image is not everything.

The Authors

Anthony R Cox is an adverse drug reaction pharmacist at the West Midlands
Centre for Adverse Drug Reaction Reporting and Dr John F Marriott is a senior
lecturer in pharmacy practice at the Pharmacy Practice Research Group, Aston
University