Prescribing errors, improvement science, nudge theory and the influence of human factors on medicines optimisation were some of the topics on the programme at the UKCPA residential symposium in November 2014
Laurence A Goldberg FRPharmS
Editorial consultant, HPE
Errors made by prescribing pharmacists occur at a rate of 0.3%, according to an award-winning project presented by Olga Crehan (Northumbria Healthcare NHS Foundation Trust). A study of pharmacist prescribers carried out in Northumbria Healthcare NHS Foundation Trust used the same methodology as the EQUIP study that reported a mean error rate of 8.9% among medical prescribers. The present study was the winner of the UKCPA/Pfizer Medication Safety Award.
In Northumbria all pharmacists are trained as prescribers and they use their prescribing ability to enhance and expand their daily activities rather than adopting specialist roles, explained Ms Crehan. Currently they prescribe for 40% of in-patients and account for 13% of all medication orders. Previous studies in the Trust had shown that ‘regular medication omitted from the prescription’ was by far the commonest prescribing error category. The pharmacy service had therefore set out to embed pharmacist prescribing into the patient pathway and targeted this area in particular, she said.
Ward-based clinical pharmacists examined prescriptions written by pharmacists on all wards on three hospital sites in the Trust over a period of 10 working days. A total of 1415 medication orders for 155 patients were included, 57% of which were for ‘regular medication omitted from the prescription’. Among these, four errors were identified, equivalent to a rate of 0.3%. Medication reconciliation was improved considerably when pharmacists undertook routine prescribing, commented Ms Crehan. The initiative was well-received by doctors who said that prescribing by pharmacists prevented unnecessary delays in treatment, she added.
Improvement science is concerned with the consistent improvement of safety, according to Clare Crowley (Lead Medicines Safety Pharmacist, Oxford University Hospitals NHS Trust). It is also used to describe ways of narrowing the gap between the research evidence base and routine care, and so it fits comfortably with the translation of trial benefits into clinical practice, she continued. It is important to understand that improvement science is not the same as research or audit.
Improvement science is underpinned by some basic concepts, such as systems and processes, the role of the customer and balancing supply and demand. It is important to understand the interaction between an improvement intervention and its context. Improvement measures are rarely successful if taken out of context, said Dr Crowley. Pharmacy is part of a bigger process and it is important to understand this in order for changes to be sustainable. In this context the patient is not the direct customer – it could be other departments in the hospital, she added.
There have been a number of approaches to quality improvement and these include ‘Six Sigma’, lean thinking, the five-step improvement approach, PDSA (plan, do, study, act) and run charts. No single method is the best and the choice of approach depends on local circumstances, explained Dr Crowley.
The first step is identifying the problem. It is helpful to start with a process map that describes all the components of the process. Two points to beware of are variability in the process and a mismatch between the description and what really happens, she warned. The next step is to identify the key drivers that influence the process. ‘Fish bone’ diagrams can illustrate this nicely. Driver diagrams can be difficult to construct but are powerful tools when completed, noted Dr Crowley.
Measuring – at the beginning and after making a change – is key to improvement. It is not an end in itself but a way of gathering information that will allow the change to be implemented, modified or rejected, emphasised Dr Crowley. A useful tip is to measure the minimum that will provide an answer. “Seek usefulness, not perfection”, she advised.
Variations can be caused by cyclical events, such as seasons, or by irregular events such as major accidents or supply chain failure. It is important to distinguish between the two because different approaches are needed to tackle them. For example, a pharmacy process might be designed to take account of seasonal variations but it would not make sense to design the everyday processes around a major accident scenario.
Turning to data presentation, Dr Crowley said that the choice lay between static and dynamic data, which shows how a process has changed with time (and interventions). For example, one group had used a run chart to show how successive process improvements over a seven-year period had enabled the pharmacy to provide medicines reconciliation services to more than 95% of patients. At the beginning of the project only 55% of patients had received the service. At each stage the median value had been calculated so that there was a clear, quantified improvement that could be related to specific interventions.
There is a common tendency to aim for an 80% success rate but this still means the process fails on 20% of occasions, said Dr Crowley. It is better to re-engineer the process to reach a target level of 95%, where the process fails on one occasion out of 20, she added.
Dr Crowley concluded that in the field of process improvement, there is no need to start from scratch, there is a wealth of information, networks and support available.
A ‘nudge’ is a way of influencing behaviour to achieve the desired outcome without reducing the number of options open to an individual but without offering an economic incentive, explained Ewan Maule (Deputy Chief Pharmacist, Operational Services, Northumberland and Tyne and Wear NHS Foundation Trust). A good example of nudge theory in practice is the way that flies are painted in urinals to provide a target and reduce urine splashes. Nudges can be used in electronic prescribing to prompt good behaviour. For example, when a warning is given, if accompanied by a message saying, “When given this information, 98% of your colleagues did not proceed”, the warning is more likely to be heeded.
The study of human factors is a growing speciality in healthcare and it has a major influence on the delivery of medicines optimisation, according to Tony Jamieson (Clinical Lead for Medicines, Yorkshire and Humber Academic Health Science Network). Human factors profoundly affect performance and the ability to use knowledge and skills effectively. Part of the reason for this is that the developed intellect and primitive instincts tend to compete. The human mind is “brilliant at solving complex problems” but it is also equipped with animal behaviours and built to function on autopilot.
As a result, the way we are made can inhibit performance, explained Mr Jamieson. “We work on autopilot whenever we can”, he added. All human beings are prone to thinking errors and cognitive distortions and these are more likely to occur when under pressure. Examples include jumping to conclusions, maximising or minimising (also known as catastrophisation), and personalisation.
A fundamental element of human factors work is engineering of systems for safe and efficient practice. One aspect of this is the standard operating procedure (SOP). SOPs are useful when it is necessary to work like a robot. However, human beings usually work to practised operating procedures (POPs) or trimmed operating procedures (TOPs). One of the problems with SOPs is that they stand still but the workplace and tasks evolve with time; what is really needed is to engineer the environment, training and peer support to make situations safer, commented Mr Jamieson. It is also important to carry out failure mode and effect analysis (FMEA) when new procedures are introduced, he added.
Good understanding of therapeutics will avoid unnecessary adverse reactions to medicines but it is an appreciation of human factors that will prevent medication errors, he concluded.
A scheme designed to reduce avoidable medicine-related readmissions to hospital has delivered positive results, according to Julia Blagburn (Senior Lead Clinical Pharmacist for Older People’s Medicine and Community Health, Newcastle upon Tyne Hospitals NHS Foundation Trust). The scheme, called REACH (Reconciliation, Education, Access to New Medicines Service (NMS)/ Medicines Use Review (MUR), Concordance and support needs assessment and High-risk medicine management) is a person-centred care model that has been developed in Newcastle.
A spot audit in 2012 showed that about 26% of readmissions were potentially avoidable. Moreover, local commissioners do not pay for a proportion of readmissions within 30 days of discharge from hospital and so this can be a costly item for Trusts. Although tools exist that can predict the likelihood of readmission this is still a long step away from preventing readmission, said Dr Blagburn. The scheme was built on the hypothesis that patient-centred care, that is, meeting each patient’s need for information, risk management or support with their medicines, might reduce the need for readmissions. Medicines reconciliation at each interface and access to the NMS or targeted MUR services in community pharmacies were central to the plan. Medications that were strongly associated with readmission were identified from the literature and pharmaceutical care bundles were developed by the pharmacy team.
The study compared readmission rates on wards where the REACH scheme was implement with the rates on control wards over a 12-month period. The results showed that about 50% of people eligible for MURs/NMS could not access them because they were housebound, cognitively impaired or their community pharmacy did not offer advanced services. The majority of these patients accepted a call or visit from hospital pharmacy staff instead. A pharmacy technician with a Medicines Management qualification carried out telephone follow-up interviews (using the MUR/NMS templates), typically taking about five minutes per patient; home visits took 20 minutes plus travelling time. The majority of interventions provided by the pharmacy technician related to information about medicines and adherence to medication regimens. The pharmacy technician resolved a wide variety of medicine problems after discharge and the support of a pharmacist was rarely required. Dr Blagburn emphasised that pharmacy staff on the intervention ward identified a larger number and range of practitioners to communicate with at discharge than the nursing staff identified on the control ward.
The results showed that the scheme works in practice although Dr Blagburn cautioned that more work might be needed to understand and improve its efficacy. There was an absolute reduction of 11% in the readmission rate in the intervention cohort during the study. In July 2014, an electronic referral pathway from hospital to community pharmacy was established using the Pharmoutcomes® web portal So far, some 400 patients have been referred for MUR or NMS.
Opioid drug interactions
When it comes to drugs, one size does not fit all, Andrew Dickman (Consultant Pharmacist, Palliative Care, Blackpool Teaching Hospitals NHS Foundation Trust) told the audience. The fate of a drug in the body depends not only on its pharmacokinetics, pharmacodynamics and pharmacogenetics but also pharmacometabolomics, he continued. Pharmacometabolomics relates to the effect of the patient’s metabolic state on the response to a drug. The ‘metabotype’ is a unique biochemical identity that determines how an individual will respond to treatment, he explained.
One interesting type of interaction is that with the efflux transporter, P-glycoprotein. P-glycoprotein is subject to genetic variation and also to induction and inhibition interactions. One of its main functions is to maintain the integrity of the blood–brain barrier. One situation where this is important is in the handling of loperamide. Loperamide is a potent opioid but it does not normally cause sedation or drowsiness because the action of P-glycoprotein prevents it from crossing the blood-brain barrier. If P-glycoprotein is inhibited, for example, by itraconazole or lansoprazole, then loperamide can enter the central nervous system. One such case has been reported in the literature, commented Dr Dickman.
Interactions that affect the metabolism of opioids can be clinically important. Interactions involving inhibition or induction of CYP3A4 and inhibition of CYP2D6 provide some of the most important examples. There are no inducers of CYP2D6, noted Dr Dickman. Both codeine and tramadol, being prodrugs, require CYP2D6 for conversion to their active forms. Inhibition of CYP2D6, for example by fluoxetine, abiraterone or quinine can reduce the analgesic effects of these drugs. It also leads to nausea and vomiting with tramadol, due to tramadol-mediated serotonin reuptake inhibition. CYP3A4 is important for the analgesic efficacy of oxycodone and for the metabolism of fentanyl and methadone. One case report described how when clarithromycin was prescribed to a patient stabilised on transdermal fentanyl, she rapidly deteriorated and experienced a respiratory arrest. The fentanyl patch was removed but restarted later with no problems.
In summary, Dr Dickman said than many drug interactions develop insidiously and that if a patient deteriorates the possibility of a drug–drug interaction should always be considered.
The UKCPA residential symposium was held at the Nottingham Crowne Plaza Hotel, 21–22 November, 2014