The Pharmaceutical Journal Vol 265 No 7123 p756-760
November 18, 2000 Original papers

Errors in prescribing, dispensing and administration of carbamazepine: a case report and analysis

By Carina J. Mack, RGN, Sue Kuc, RGN, and Richard A. Grünewald, DPhil, MRCP

Aim o The epilepsy nurse specialists in Sheffield identified three incidents of inadvertent intoxication with carbamazepine modified release. One of the incidents is described. Because of the potentially serious implications of such intoxication, reports of similar incidents were analysed, in order to investigate the extent of the problem and its effect on patients, and address the causes.

Design o Reports of errors in prescribing, dispensing and/or administration of carbamazepine were solicited from the readers of Epilepsy Today, the newsletter of the British Epilepsy Association, by publication of a letter to the editor; responses were invited by telephone, facsimile or letter.

Subjects and Setting o The study was based in an epilepsy liaison department of a regional neurosciences centre.

Outcome measures o Type of error and its medical and social consequences.

Results o 30 replies were received from throughout the UK, all but one from patients. There were three prescribing errors, 17 dispensing errors and 10 self- administration errors. These errors had serious medical, social and psychological consequences (eg, carbamazepine intoxication, loss of seizure control, loss of driving licence).

Conclusion o Inadvertent intoxication with carbamazepine is widespread and avoidable. The error might have been prevented in 14 cases by changing the appearance of the packaging for the product, in 12 cases by changing the appearance of the tablets and in three by education of medical staff.
Patients should be warned to check the strength of carbamazepine tablets when they are dispensed and of the difference between standard and slow-release formulations.
Pharmaceutical companies should consider the possibility of confusion between different formulations of a drug when designing packaging, and understand that patient safety must come before brand image.

We recently identified three incidents where patients taking carbamazepine modified release tablets became inadvertently intoxicated with the drug as a result of errors in the dispensing or self-
administration of the drug. One such incident is reported (see Case report opposite). As such incidents were potentially serious, we analysed the contributory factors by inviting reports of similar incidents from patients with epilepsy.

Methods

We invited patients to contact us with details of similar incidents by advertising in the journal of the British Epilepsy Association, Epilepsy Today, a magazine with a circulation of about 21,500 which is read mainly by professional and lay members of the association.1

Results

Thirty-six replies were received by telephone, facsimile and letter over a period of eight weeks from publication of the invitation in June, 1999.
Three could not be followed up because of incomplete information or incorrect telephone numbers left on voicemail. Of the 33 incidents recorded, three were irrelevant to the topic of carbamazepine prescribing, dispensing, and self-administration errors. These have been excluded from the analysis. Thirty relevant incidents were analysed (20 telephone calls, seven letters and three facsimiles).
Twenty-nine responses were from patients or their relatives, and one from a clinical nurse specialist in epilepsy. All but one related to incidents in primary care. Twenty responses were from women, and 10 from men. The incidents concerned 25 adults and five children.
Communications were received from throughout the United Kingdom.

Dosage errors There were 16 overdoses of and five underdoses of modified release carbamazepine. In five cases, patients confused 200mg and 400mg modified release tablets. In three cases, standard formulation rather than modified release formulation carbamazepine was administered.
Analysis of causes of dosage errors There were three prescribing errors, 17 dispensing errors (one possibly an error by a nurse), and 10 patient self-administration errors.

Prescribing errors There were three prescribing errors:

1. General practitioner did not state "modified release" on prescription form, resulting in dispensing of standard carbamazepine.
2. General practitioner did not update computer prescription from 200mg modified release to 400mg modified release preparation.
3. General practitioner did not alter prescription from 400mg to 200mg tablets to allow patient to titrate medication in 100mg (half-tablet) increments.

Dispensing errors There were 17 dispensing errors in three categories:

1. There were 14 incidents where the pharmacy had confused 200mg and 400mg carbamazepine modified release tablets. In 11 incidents the pharmacist dispensed 400mg modified release tablets instead of 200mg modified release tablets in error in a white box. In each case the pharmacist labelled the boxes incorrectly as 200mg modified release tablets. In some cases white boxes contained a mixture of 200mg and 400mg modified release tablets, and in some cases 400mg modified release tablets in the manufacturer's box were incorrectly labelled as 200mg modified release by the pharmacist. Three incidents were reported of 200mg modified release tablets being incorrectly labelled as 400mg modified release tablets and supplied in white cardboard pharmacy boxes.
2. There was one incident of confusion between standard and modified release formulation in a hospital inpatient who was given standard formulation carbamazepine in place of the same dose of modified release carbamazepine. This may have been a prescribing error, pharmacy error or nurse administration error.
3. In two incidents, patients were dispensed 200mg standard formulation carbamazepine tablets in place of 100mg standard formulation carbamazepine tablets incorrectly labelled by the pharmacist on the outside of the box.

Patient adherence errors There were 10 patient adherence errors in two categories:

1. Eight patients were dispensed a mixture of carbamazepine modified release 400mg and 200mg tablets, but the tablets were confused by the patients because of the similarity in size and shape of tablets.
2. In two cases, patients who were accustomed to receiving 200mg modified release tablets were provided with 400mg modified release tablets. Although these were labelled correctly, patients were not warned of the difference in the tablets, and took inappropriately large doses. One patient was visually impaired, and relied on counting tablets.

Identification of errors In six cases the patients noticed the error before taking the tablets. In all cases, patients noticed that white (standard formulation) carbamazepine tablets had been supplied rather than beige/orange or brown/orange modified release formulation.
In three cases 400mg modified release tablets were identified as too large when the patients placed them in their mouths.
In the remaining 24 cases, patients noticed a problem only after becoming ill.
In three incidents where patients were inadvertently prescribed standard formulation carbamazepine, patients noticed the different appearance of the tablets before taking them, and were able to avoid the consequences of the error.

Medical consequences In one case a patient had been seizure-free on carbamazepine modified release 1g daily, but since inadvertent overdose with carbamazepine had been intolerant of this dose with resulting deterioration in seizure control.
In one case a patient with epilepsy and insulin-dependent diabetes mellitus became intoxicated with carbamazepine, resulting in vomiting and deterioration in epilepsy and diabetic control and requiring hospitalisation.
One child regularly taking 2g of carbamazepine modified release daily (but with no symptoms of intoxication) would have been given 4g carbamazepine modified release daily if the dispensing error had not been noticed by the mother.

Social and psychological consequences One child with a neurological disorder resulting in epilepsy and visual impairment was accustomed to administer medication to themselves by counting tablets. Since the incident the patient lost confidence in this method of drug self-administration and consequently some independence.
One patient lost his driving licence through breakthrough seizures after inadvertent underdosing with carbamazepine modified release.
One mother was physically unable to care for her baby while intoxicated with carbamazepine.
Two patients had to take time off work because of carbamazepine intoxication.

Discussion

Modified release carbamazepine has a longer half-life and lower bioavailability compared with standard formulation carbamazepine.2 Peak anti-epileptic drug levels in the bloodstream are lower with the equivalent dose of modified release carbamazepine than standard formulation carbamazepine. Equal doses of standard formulation carbamazepine may thus precipitate symptoms of intoxication in patients who have been taking modified release carbamazepine. Standard formulation carbamazepine has a shorter effective half-life in the body compared with modified release carbamazepine. Thus, patients who inadvertently take standard formulation carbamazepine may experience symptoms of intoxication if they take the medicine only twice daily, or experience breakthrough seizures.
Symptoms of acute intoxication with carbamazepine include double vision, nausea, vomiting, headaches, dizziness, poor balance and sedation. These symptoms rapidly resolve on reduction of the dose, and long-term effects of intoxication are unusual.2
Carbamazepine standard and modified release formulations are now available from more than one manufacturer in the UK. Patients are therefore becoming accustomed to receiving tablets that differ in appearance and packaging from that with which they are familiar.3 None the less, some patients were able to recognise that they had been supplied with a tablet of unusual appearance, alerting them to the possibility of a prescribing or dispensing error. Patients should therefore be educated about the appearance and available formulations of carbamazepine, told to check which they have received and, if in doubt, ask for advice from their pharmacist, general practitioner or nurse specialist.
Because of the nature of the survey, the replies reflect a biased sample. Only those who saw the article in Epilepsy Today and who were sufficiently motivated to contact us were included. Many critical incidents may have remained unattributed to errors in carbamazepine dispensing, patient adherence or compliance. Some may have gone unnoticed by the patient or carer, and, where the patients took inappropriately low doses of carbamazepine, resulting poor seizure control may have been attributed to other factors. Intoxication with carbamazepine may be attributed to other causes by carers or general practitioners.
Over 50 per cent of the incidents were due to dispensing errors. It was likely that the incidents could have been prevented in 14 cases by changing the appearance of the manufacturer's box (both standard and modified release formulation), in 12 cases by changing the appearance of the tablets and in three cases by education of general practitioners or their practice staff. Currently, the principal manufacturer of standard and modified release carbamazepine (Tegretol, Novartis) packages all available tablet strengths and formulations in similar yellow and white boxes. The new European requirement for all medicines to be dispensed in their original manufacturer's packaging (usually as blister-packs) will reduce the risk of inadvertent supply of mixed tablets in the same white pharmacy box, but this will not prevent confusion between packaging which looks similar.
Packaging of drugs has already been identified as a cause of drug errors.4 Pharmacists and patients should be protected by multiple barriers to error, including checking and dispensing policies, clinical pharmacy procedures and obvious differences in packaging. A change in the appearance of the tablets would probably have prevented 12 of the incidents. This is emphasised by the observation that all three incidents where standard formulation carbamazepine (white) was supplied instead of modified release carbamazepine (beige/brown and orange) were immediately noticed by the patient before the medicine was taken. If 200mg and 400mg modified release carbamazepine tablets were markedly different in appearance, it is likely that some of the errors would have been prevented.
Three out of the 30 incidents arose from general practitioners' prescriptions. There were probably several reasons for such errors, but it is vital that prescribing physicians recognise the pharmacokinetic differences between standard and modified release carbamazepine.
Errors in carbamazepine dispensing, administration and compliance are likely to represent a significant cause of morbidity in the UK. It is impossible from this survey to estimate the magnitude of the problem nationally, but the impact of the incidents is often severe and may result in litigation.
Some of the errors are avoidable.5 Pharmacy dispensing errors may be reduced by adherence to protocols and prominent labelling of the packaging to reduce errors in checking by pharmacists. Pharmaceutical companies should be encouraged to market different drug formulations and strengths in packaging that is easily distinguishable. Education of patients by nurse specialists and doctors to appreciate the difference in shape, size and colour of the tablets may have some role, but this is limited by the widespread use of generic prescribing and the physical similarity of the 200mg and 400mg modified release preparations.

Acknowledgments We thank the British Epilepsy Association for their help in undertaking this study.

References

1. News. Epilepsy Today 1999;47:4.
2. Novartis. Tegretol standard formulation data sheet. Camberley, Surrey: Novartis Pharmaceuticals UK Ltd. 1999.
3. Crawford P, Hall WW, Chappel E et al. Generic prescribing for epilepsy. Is it safe? Seizure 1996;5: 1-5.
4. Intravenous potassium predicament. Clin J Oncol Nursing 1997;1: 45-9.
5. Bates DW. Frequency, consequences and prevention of adverse drug events. J Quality Clin Pract 1999;19:13-7.