A spoonful of sugar…understanding medication errors

Post date: 19/04/2021 | Time to read article: 9 mins

The information within this article was correct at the time of publishing. Last updated 20/04/2021

As the GMC updates its prescribing guidance, Dr Sophie Haroon, Medicolegal Consultant at Medical Protection, looks at some common causes of claims and provides advice on avoiding the pitfalls

Numbers are a great way of getting an idea about the size of a problem. So it is without dispute that medication errors worldwide are a huge issue given they are estimated to cost $42 billion. In England, there are an estimated 237 million medication errors each year, causing or contributing to possibly 22,300 deaths annually.¹ Of note, although perhaps rather ambitious now in the light of COVID-19, the WHO in 2017 launched a global initiative to reduce medication-related harm by 50% in five years.² Prescriptions are vital for many, but all in all it kind of makes you want to put away the prescription pad.

But hang on in there for a while. There have been several good studies looking at prescribing errors and their causes.^3,4 Think of the ‘five Ts’ as a starter:

Time – never enough of it, and always in a rush, equals an increased likelihood of making an error
Tiredness – impairs the ability to think straight, so inviting error
Training – lack of knowledge around new drugs, new doses, new interactions, new computer systems – can add up to an error
Type of patient – not appreciating the differences across the age spectrum for drugs, patient compliance, patient manipulation of the doctor-patient relationship (see below) can lead to prescribing problems
Task interruption – ever started something, been interrupted, then returned only to have forgotten what you were doing? Clinicians in A&E are interrupted approximately nine times an hour when prescribing which, unsurprisingly, significantly increased errors.⁵ Undoubtedly similar frequencies exist across other clinical interfaces.

There has also been lots of advice, guidance and initiatives over the years to help with safer prescribing practice:⁶ good systems, governance, incident management, reporting, reflection, communication and training to name but a few.

New GMC guidance

“To err is human” – and errors occur. The GMC has updated their guidance on Good practice in prescribing and managing medicines and devices.⁷ It came into effect on 5 April 2021 and is essential reading.

The GMC’s guidance is clear and self-explanatory so is not explored in detail here. Instead, it is used as a springboard to look at the experience of Medical Protection in relation to prescribing error claims. Looking over the last ten years, 534 claims were related to prescribing or dispensing errors.

What causes a prescribing error claim?

In no particular order, the following are the most common topics and drugs making up a prescribing error claim:

NSAIDs – Medical Protection has seen several claims about prescribing NSAIDs (typically naproxen) without appropriate PPI cover so leading to gastritis, ulceration and occasionally upper gastrointestinal haemorrhage.
NICE guidance on this matter has been in existence since 2004, with some minor updates over the years and some more detailed addition of contraindications and monitoring in 2010.⁸ The BNF has provided guidance on when to co-prescribe a PPI since at least September 2005. Despite all this, deviations from the guidance are seen, which cannot be adequately justified – meaning many of these claims cannot be defended.

Key take home point: Readers are encouraged to review the criteria for co-prescribing a PPI. Be mindful this can happen insidiously, eg when an NSAID prescription becomes long-term (which can happen when put on repeat and no-one remembers to check if a PPI has been given), or if the patient moves through the age thresholds (so automatically changing their number of risk factors by simple senescence).
Nitrofurantoin – This type of claim usually involves a patient with recurrent UTIs who ends up on long-term nitrofurantoin. It may be initiated by secondary care – who arguably take some responsibility for initial counsel – but is usually continued in primary care – where responsibility for monitoring of pulmonary function and respiratory symptoms and liver function tests lies. A further problem arises when the patient develops respiratory problems and it is not recognised that there is a link between nitrofurantoin and the symptoms. After some time, they are diagnosed with nitrofurantoin-related pulmonary fibrosis.

Unfortunately the severity of the lung reaction and degree of resolution is directly linked to the duration of the nitrofurantoin treatment after the first clinical signs appear. Typically, the prescribers are pursued for failing to monitor the patient on nitrofurantoin or failing to link the drug to the respiratory symptoms or deranged LFTs in a timely fashion.

The problem in addressing these claims is that expert opinion has been highly divergent.^{9, 10} Some experts have opined that a reasonable body of GPs would not know about the link between this antibiotic and the rare respiratory and liver side-effects. Others have argued that the side-effects and the need to monitor for them has been in the BNF since at least edition number 63 in 2012 and that because they have been described since then, it would have been incumbent on prescribers to have double-checked. Furthermore, even the frequency of monitoring has been queried.¹⁰

Key take home point: The link between nitrofurantoin and respiratory and liver problems is increasingly being recognised despite its rarity. Expert opinion is divided but it will be increasingly difficult to justify not knowing the link and not monitoring patients on nitrofurantoin, given it has been in the BNF for years. Indeed, the GMC’s guidance states one must be familiar with the guidance in the BNF and British National Formulary for Children (BNFC) … and one should follow the advice in the BNF [para 13 and 14].

Review patents on long-term nitrofurantoin to ensure they are having at least six-monthly LFTs and respiratory symptom checks, and that they know to present asap if they develop any respiratory problems. Discontinue the drug if there is deterioration in lung function.
Other antibiotics – Staying with this type of drug, Medical Protection has seen claims relating to the use of ciprofloxacin in elderly patients leading to Achilles tendon rupture. The latter was not a direct contraindication but the patient pleaded they were not adequately informed of this possible side-effect.
Also, penicillin being given to penicillin allergic patients still occasionally arises.

Both types of claims usually arise out of a failure to adequately consent and inform patients about the medication, and check out contraindications even if they are not properly coded in the patient’s record. Patients themselves do not necessarily recognise the similarities between antibiotics – like flucloxacillin being related to penicillin.

Key take home point: Ensure you act within the GMC’s guidance when starting or continuing a medication in terms of patient consent, just like you would if this was an operation or procedure [para 34, 35, 40].

Do not rely on what is, or is not, recorded in the records. Older drug reactions may have been mis-coded and current medications may not be completely up to date – they should be – but recheck with the patient. And clearly document that you have checked for any allergies.
Controlled drugs and other medicines where additional safeguards are needed – Ever started a patient on lorazepam – low dose, short course – and before you know it, they are asking for more and a longer course? Medical Protection frequently sees claims of protracted prescribing of drugs like tramadol, benzodiazepines and now more lately pregabalin, to name but a few of this type. These can be difficult patients to manage with real physical and also psychosocial issues, where they appear to be walking a tightrope for life. After several years they plead they are now addicted, that they cannot function on or off the medication, and are having a torrid time coming off it.

Key take home point: The GMC’s guidance talks extensively to the principles of prescribing in this area [para 59-73]. For opioids in particular, the module Opioids Aware developed by a range of stakeholders is a particularly useful tool.¹¹

Make sure you have clear systems and processes for managing patients who are becoming addicted to, or are at risk of becoming addicted to, these sorts of drugs. Be wary of alleged lost prescriptions or the patient obtaining their medication from elsewhere – be it out of hours or over the internet. Patients may stick to seeing the same clinician in order to manipulate them into prescribing for them. Ensure there are supportive systems in place to help these colleagues manage such patients.

Not everyone on this sort of medication will abuse it. But keep patients on these sorts of drugs under regular medication review and watch out for escalating patterns of recurrent prescription requests and dose increases, which may indicate tolerance.
Repeat prescribing – Perhaps one of the most challenging areas in terms of preventing errors is the rewriting of drug charts or signing of repeat prescriptions – sometimes because this activity is done in haste and often because the patients are unfamiliar. Medical Protection sees claims where a medication has been prescribed for several years by numerous doctors, on repeat, with no formal review or consideration as to whether it should be continued for so long. Steroids often fall into this bracket – be it oral with no thought to bone protection, or in the form of eye drops leading to visual problems. Paradoxically, bisphosphonates also fall into this category with the absence of a ‘drug holiday’ (or at the very least a review and consideration of one, dependent on individual patient risk), leading to cases of atypical femoral fracture.

Key take home point: The GMC’s guidance is clear that prescribers are responsible for their own prescriptions even if the medication was started by someone else, that the prescriber must have adequate knowledge about the patient’s health, and must only prescribe a drug if it meets their needs [para 3, 20].

When requested to re-prescribe, consider whether the drug should be being continued, whether there are any long-term consequences to staying on it, whether there are any monitoring or co-prescription requirements (see, for example, points 1 and 2 above), and whether the patient is in the system for a formal review of the medication at some point.

The GMC’s guidance speaks extensively on the topic of repeat prescribing, emphasising the importance of the measures it states [para 98-102].
Drugs with a similar name – HRT patches are a notable feature in this category. It does not help that they sound very similar or that on a drop-down list, you are only one click away from prescribing the wrong preparation. The claims are usually about unopposed oestrogen being given to someone with an intact uterus, leading to hyperplasia and sometimes even cancer years down the line.

Key take home point: When prescribing a drug with a number of similarities or from a list where they are all grouped, double check that the one you think you are prescribing is the one you have really chosen, and the one you really mean to choose for that particular patient.
Doses – Drug dose errors are unfortunately most common in neonatology¹² where a dizzying array of dilutions and calculations often have to be made. But in keeping with the theme of everything sounding the same, antidepressants are not an uncommon medication where the wrong dose is prescribed. For depression in an adult, escitalopram starts at 10mg and can be prescribed at a maximum of 20mg; for citalopram the starting dose is 20mg rising to 40mg. Get the two confused and one ends up with a claim for serotonin toxicity syndrome.

Key take home point: There may be many drugs you are familiar with and do not need to refer to the BNF for. In reality, do not waste time keeping drug doses in your head. Better to look them up in the BNF to be doubly sure – that’s what it is there for – to carry the weight of drug names, drug doses, formulations, indications, contraindications and all other detail you could possibly need.
Handover and shared care – The old adage of “too many cooks…” can apply to any area of life, let alone the kitchen. But the fact of the matter is that patients can be seen by many clinicians in many settings. A recurrent issue is a change of medication when patients are discharged from hospital or seen in an outpatient setting.

In the first instance, anticoagulants often crop up. Not uncommonly there is a failure to change over from warfarin to a direct oral anticoagulant. The resultant injury is a stroke a few days or weeks down the line.

In the latter scenario, primary care may be asked to prescribe something by secondary care that is unclear. The patient then ends up on some hybrid course of medication that is not fit for purpose.

Key take home point: The GMC’s guidance gives clear detail about shared care arrangements, handover arrangements and others prescribing for a patient at the request of someone else [para 74-82].

Ensure there are good systems in place to respond to medication changes that might be detailed in patient discharge letters or outpatient clinic letters, and that the patient record is updated in a timely fashion to reflect the changes [para 55].

If the detail from one care organisation to another is unclear, confusing or felt to be clinically inappropriate, challenge before prescribing – as ultimately that prescription will be your responsibility.

Conclusion

“The fox knows many things, the hedgehog one big thing,” wrote the Greek poet Archilochus, which has been interpreted as distinguishing between “those who relate everything to a single central vision [hedgehogs]…and those who pursue many ends [foxes]”.¹³ Hopefully this article has enabled you to be more ‘fox-like’ in understanding some of the many ends that lead to a prescribing error – be it individual drugs, system process or people-directed (patients and colleagues). Prescribing errors have many causes but certain ones make it to the fore of clinical negligence.

Rather like a drug itself, this article carries a safety warning: although these topics and drugs show an insight into what is pursued in a clinical negligence claim about a prescribing error, they are obviously not an exhaustive account of what could be pursued now or in the future. However, it is hoped that by sharing some of the commonest errors, readers may in some way be forewarned and forearmed, and also encouraged to reflect on their prescribing practice in relation to these topics and drugs. And don’t forget, see the updated GMC’s guidance on this topic.⁷