Paediatric medication error: the problem

Every day of our working lives we engage in healthcare encounters that have the potential to cause unintended harm to our patients. That harm can range from distress as a result of an insensitive remark, through misdiagnosis, to treatment errors involving procedures or medication. The National Patient Safety Agency,1 reporting on an international literature review, estimated an adverse event rate of between 2.1% and 10.8% of hospital admissions for patients aged 0–17. Fortunately, only a minority of these errors result in the most severe outcomes of death or permanent disability, but this is in large part because many common paediatric medications are tolerated well, and children are resilient in their response to therapeutic mishaps. The regular occurrence of errors points to serious vulnerabilities in the medication delivery process, and it is predictable that serious and fatal events will occur unless these processes are improved.

Drilling down on the causes of harm in healthcare settings, the most recently analysed data set from the National Reporting and Learning Service (NRLS),1 which receives incident reports for England and Wales, demonstrated that ‘medication errors’ were the most commonly reported incident type in the paediatric population (17% of incidents in children and 15% in neonates). The age group 0–4 years had the second highest percentage of medication incidents of all age groups, only exceeded by errors in those over age 75. Of course, despite being one of the world's most comprehensive databases of patient safety information, it is known that the lack of mandatory reporting makes NRLS information a gross underestimate of the true incidence of events, particularly in primary care. Moreover, within any one organisation a wide range of local and cultural issues influences both the rates of incident reporting and the types of incidents that are reported.

In 2011, the Secretary of State for Health established a group, the Children and Young People's Health Outcomes (CYPHOF) Forum, to identify the health outcomes that matter most for children and young people, and to set out the contribution that each part of the new English health system needs to make to improve those outcomes. CYPHOF recognised that medicines pose a risk to children through two main routes:

1. First, due to extensive unlicensed and/or off-label drug usage. This is historically a consequence of inadequate testing of new medicines in the paediatric population, although matters have improved considerably since 2007, when the European Union imposed a stricter framework for development and testing of new medicines for children.

2. Second, as indicated by NRLS data, harm is all too commonly caused by PME. This can occur at any point in the pathway, including formulation, prescription and administration.

Within a wide-ranging series of recommendations2 encompassing all aspects of child health, CYPHOF proposed a series of measures to improve error reporting, as well as calling for a collaborative approach to the development of interventions aimed at reducing harm from medicines usage. In response, a Paediatric Medicines Safety Board, co-chaired by the president of the Royal College of Paediatrics and Child Health and the chief pharmaceutical officer, was established to progress this work.

Our optimistic starting point in tackling any clinical problem is the hope that there will be an evidence base to guide us. Sadly, we frequently find ourselves beset by three problems. First, there is often a lack of robust evidence underpinning the issue in hand. Second, even in the face of evidence-based solutions, lack of awareness, delays and resistance to change may impede translation into clinical practice. Finally, and arguably most frustrating, is the failure to scale up or sustain locally implemented changes, with the latter issue being one of the most challenging problems for broader health service reform and modernisation.

Mindful of these limitations, the Paediatric Medicines Safety Board commissioned a systematic review from the Evidence for Policy and Practice Information and Co-ordinating Centre (EPPI-Centre), part of the Social Science Research Unit at the University College London Institute of Education. The review sought to identify any existing evidence about the nature and extent of paediatric medication error (PME) and the interventions known to improve it.3 Although the lack of mandatory reporting already alluded to made quantification of the extent of PME unreliable, there was good evidence about the nature of errors, with off-label prescribing commonly being associated with dosing errors, particularly in primary care. Perhaps not surprisingly, the evidence base on how to tackle the whole spectrum of problems that lead to PME had significant gaps and limitations. Although there was strong evidence that electronic prescribing reduces the frequency of PME, the evidence as to whether it reduces harm from PME was more equivocal. Other interventions, such as clinical decision support and education interventions, had lesser levels of evidence underpinning them. There was only promising evidence that they reduced the frequency of PME and very little evidence regarding their impact on adverse sequelae.

Despite our intuitive belief that education must improve practice and outcomes, research in this area remains very limited, generating more questions than answers. Evaluation of education generally stops at the point of asking learners how highly they rated content and if they believe it will change their practice or direct further learning. Even when learning is tested in a simulated environment, there is little research on whether the learning actually changes practice, and, more importantly, whether changed practice impacts patient outcomes. For example, in the EPPI-Centre systematic review, only six studies met quality criteria for inclusion, and the interventions were very narrow and distinct in approach, topics and target groups; three targeted administration issues by parents and carers, and three targeted prescribing, predominantly by doctors in training. The main outcomes were either knowledge or performance in a simulated situation. Much more work is needed to determine which modalities of education work best, how these methods translate to different clinical and lay groups, and whether performance in simulated conditions translates to the bedside.

Developing a pragmatic approach to improvement

One might consider, even within the limitations of the evidence, that widespread roll-out of electronic prescribing would be the most fruitful way forward. However, even with e-prescribing being used more widely within the National Health Service, there is currently a relative lack of deployment of e-prescribing in children compared with adults. Currently, only a few of the specialist children's hospitals have bespoke paediatric systems in place, and the adult-based systems in use in non-specialist hospitals are mostly unsuitable for paediatric prescribing. This leads to additional risks as doctors in training move between different electronic systems, and between electronic and written systems. Moreover, the bulk of paediatric prescribing occurs in primary care, where paediatric-specific systems are even further over the horizon. Internationally the picture is no better than in the UK, with few systems in place that address the weight-based dosing needs of children or gestation-based prescribing for neonates.

The promising evidence supporting both education and clinical decision support tools indicates that both of these approaches should be actively pursued, particularly since they are more amenable to local low-cost implementation strategies. However, it is equally clear that the evidence base is still very thin, and much more research is needed if we are to develop clinical decision support and educational interventions that are generalisable and have maximum effect. It is also encouraging that many local teams are engaged in quality improvement projects to address PME, with a grey literature suggesting good results from a wide range of measures such as clinical staff wearing warning tabards when prescribing or dispensing to discourage interruptions, pharmacist-supported ward rounds and a plethora of smartphone apps.

Taking into account both the evidence base and the front-line practicalities, the Paediatric Medicines Safety Board proposed that a pragmatic approach to improvement should be pursued, encompassing the following elements:

1. The development of a medicines safety quality improvement network, which would act as a testing bed and means of dissemination of good practice and successful interventions. This approach would have the added advantage that organisations that prove sufficiently motivated to opt in and join this network would also be committed to sharing local data on PME, thus enabling more effective evaluation of outcomes.

2. Development and/or sharing of clinical decision support tools.

3. Collation and warehousing of existing educational resources to reduce duplication of effort, followed by a gap analysis and development of any additional educational resources felt to be missing from the existing armamentarium.

A programme to support these three elements, comprising development of a quality improvement network, clinical decision support tools and collation/development of educational resources, has been supported by Department of Health funding. The programme is now being hosted by the Royal College of Paediatrics and Child Health, in collaboration with a number of other Royal Colleges, key national bodies and stakeholder groups. The programme is constructed around a web resource developed by the American Academy of Pediatrics called Pediatric Care Online.4 The American site provides a single source of quality-assured information including best practice guidelines, quick reference materials, drug information, policy information, patient handouts and other tools, and also incorporates a broader Medline search engine.

The UK version has two distinct components, the first of which, MedsIQ,5 launched in spring 2015. MedsIQ comprises a partnership between the Royal College of Paediatrics and Child Health, the Royal College of Nursing, the Royal College of General Practitioners, the Royal Pharmaceutical Society, the American Academy of Pediatrics and the Department of Health. It brings together tools and improvement projects that have been developed to specifically address the problem of harm to children through PME and is engaging with other key networks such as ‘Making it Safer Together’, a national paediatric safety collaborative. A rating system allows users to score resources that they use, so that highly rated tools will spread more effectively. As the network develops, it will be important to move from simple ratings to qualitative feedback on the various tools; this will generate a better understanding of whether tools are rated highly because of ease of initial implementation and use, or whether they generate measurable outcomes in terms of improvements in medicines safety.

The second component, Paediatric Care Online UK, involves the same partners as MedsIQ, with the addition of the Institute of Health Visiting. It launched in beta testing phase in autumn 2015, with full roll-out planned in early 2016. It provides UK quality-assured clinical decision support tools and other important resources in a format that can be cross-hyperlinked to quality-assured practice guidance; for example, the British National Formulary for Children, Public Health England's Green Book on immunisations against infectious diseases, National Institute for Health and Care Excellence guidance and a growing repository of expert-developed content.

Can online resources and quality improvement networks have a real impact on outcomes?

The philosophy of quality improvement has gained currency across the international healthcare community and is recognised in the UK as one of the essential duties of every doctor,6 embodied in our revalidation requirements. There are legion examples of excellent local innovations as a result. The prize for healthcare systems is if we can reduce duplication of effort, develop meaningful sharing of good practice and move past the widespread belief that ‘my hospital/patient group/department is different, so we cannot adopt someone else's innovation’. It is self-evident that the combination of spiralling healthcare costs and a harsh economic environment makes the sport of ‘wheel reinvention’ a luxury we can ill afford and a disservice to our patients.

It is now over a decade since Trisha Greenhalgh produced a systematic review on the barriers and enablers to the spread of good ideas.7 We still appear to be a long way from recognising the lessons within that report and following through on some of the recommendations for further research. It remains to be seen whether MedsIQ and Paediatric Care Online UK will break through these barriers and genuinely change practice and outcomes. The MedsIQ project team strongly believes that MedsIQ will only succeed if it moves beyond a website to a community of practice and learning, and to this end international links have already been developed with colleagues in Europe, America and Australia, alongside engagement in the UK. A strategy for ongoing evaluation of usage and outcomes of Paediatric Care Online UK is being built into the programme design. However, even at this early stage, a positive outcome has been the bringing together of a strong collaborative network that crosses primary and secondary care, as well as engaging professional groups across the children's healthcare workforce. Success will be dependent on the enthusiasm with which users engage with these tools and become part of a community committed to continuously improving and developing the resources, as well as their own practice.