Introducing a system that could create more work and cause new types of errors seems counterintuitive. But these are some of the problems being encountered by hospital pharmacists who are implementing electronic prescribing and medicines administration (EPMA) systems as part of NHS England’s plan to go paperless.
Some hospitals have had electronic prescribing (e-prescribing) systems in place since the early 1990s. But in the past five years, there has been a real push for EPMA systems to be adopted across the board in England as the NHS strives to meet its target of going paperless by 2024 — an ambition that has already fallen behind from former health secretary Jeremy Hunt’s aim of digitising secondary care by 2020.
EPMA systems can vary widely, from simple electronic versions of the traditional paper drug chart to a fully integrated system that seamlessly connects with patient records and dispensing software.
While, on paper, e-prescribing is a laudable idea, bringing it into fruition is far from simple. Arti Punn, lead community healthcare pharmacist at Birmingham Community Healthcare NHS Foundation Trust, acknowledges that, in the long run, replacing paper drug charts with their computerised counterparts will be a good thing, but she is candid about it not being an easy switch to make.
In the early days, she says, it can mean more work, can introduce errors and demands a change in mindset about how things are done.
Punn has worked at both ends of the spectrum, from basic standalone models where neither the prescriber nor pharmacist can see the ‘full picture’, to a fully integrated system that was incredibly efficient. She says that, whatever system a trust chooses, implementing it will be a significant change to how things have worked previously and teething problems will be inevitable. “You will have to be patient; there will be problems that are really specific to the system you are using, and it can cause annoyance and frustration. But if you do it right, it will be worth it,” she says.
Source: Arti Punn, Ann Slee, Bryony Dean Franklin, Matt Elliott, Hetal Halai
Progress to date
How far hospitals have come is difficult to measure because some have EPMA systems in certain clinical areas but not in others. Ann Slee, associate chief clinical information officer (medicines) at NHSX, says around 35% of acute trusts in England are now live with EPMA, meaning that 80% or more of their inpatient prescribing is done electronically. Many more are in the process of implementing systems and, with the pace of this accelerating, Slee expects 50% of acute trusts to have EPMA systems in place “before long”.
Mental health trusts are even further behind, with just 10% having EPMA fully in place. “This is increasing rapidly with help from the EPMA funding stream and the expansion of the global digital exemplar programme,” says Slee.
Scotland and Wales have their own targets for implementing EPMA, with Scotland further ahead than Wales, which is waiting for business case to be approved (see Box 1).
However, getting a system up and running is just the start. “Implementing the system is one thing,” says Aziz Sheikh, chair of Primary Care Research and Development at the University of Edinburgh. “But we are still at the very early stages of how you actually derive benefit from these systems.”
Implementing the system is one thing, but we are still at the very early stages of how you actually derive benefit from these systems
Brian Power, lead informatics pharmacist at Wirral University Teaching Hospitals NHS Foundation Trust, agrees. “Implementation is a significant milestone, but you have to keep responding to technological and clinical changes,” he says. Wirral was one of the first sites to implement an EPMA system in the early 1990s and is now a “global digital exemplar” — 17 acute trusts recognised by NHS England for their particularly advanced use of digital technology, which are also helping to develop ideas and share learning with others.
Box 1: Progress in Scotland and Wales
The goal set out in Scotland’s eHealth strategy 2014–2017 was for all NHS boards to have “implemented some elements of electronic prescribing and medicines administration (EPMA) systems with integral clinical decision support interfaced with other clinical eHealth systems by 2020”. The Hospital Electronic Prescribing and Medicines Administration (HEPMA) implementation programme is ongoing, with the latest figures suggesting the Scottish government’s contribution will be £24m between 2016 and 2023.
Rose Marie Parr, Chief Pharmaceutical Officer for Scotland, says that, to date, three health boards (NHS Ayrshire & Arran, NHS Dumfries and Galloway and NHS Forth Valley) have fully implemented HEPMA. “NHS Lanarkshire is close to completing its implementation and NHS Lothian has started implementation this year,” she adds. “NHS Grampian, NHS Highland, NHS Tayside, NHS Orkney, NHS Shetland and NHS Western Isles are taking a regional approach, again commencing this year, and NHS Greater Glasgow and Clyde is finalising its business case with a view to implementation later [in 2019].”
There has been a national plan to implement a hospital e-prescribing system for Wales since 2007, with an initial target of rollout by 2010. However, a report on medicines management published in 2018 pointed to the frustration of pharmacy teams at the slow progress in implementing this, with paper prescribing still in place
The Welsh Hospital Electronic Prescribing Pharmacy and Medicines Administration project is now committed to implementation at a cost of £20m over a seven-year period. Cheryl Way, national pharmacy and medicines management lead at the NHS Wales Informatics Service, says Welsh health boards currently have electronic prescribing systems for chemotherapy, and procurement is underway for a national critical care system that includes electronic prescribing.
“An outline business case for a hospital e-prescribing system for Wales has been drafted and is going through the assurance process at present,” says Way. “Subject to approval, we plan to go out to tender for the system at the end of 2019 and award a contract in early 2021.”
One issue — and there is work going on to address this through NHS Digital and the digital exemplar programme — is that trusts work in silos, repeating the same mistakes made by others while battling away, heads down, to get their systems up and running. This lack of connectedness and sharing of experiences is hampering efforts.
“We need to breed a spirit of cooperation across the NHS. If you can do that, it will be to the benefit of everyone,” says Sheikh.
Power notes that working cooperatively would also give the UK a bigger voice with United States-based suppliers. “Trying to get the UK viewpoint heard and actioned is difficult,” he says. “If we want to make changes, we have to work with other hospitals with that system. If you get ten hospitals making the case for changing a process, the suppliers should be receptive to that.”
Safer than paper
One of the main reasons for implementing EPMA is safety. There are an estimated 237 million medication errors per year in the NHS in England, with 66 million of these potentially being clinically significant
Bryony Dean Franklin, a medicines safety and technology researcher at the UCL School of Pharmacy in London, warns that it is dangerous to assume that errors will disappear simply because a computer is involved. “People think [e-prescribing] will be safer, but that is not necessarily the case,” she explains.
People think e-prescribing will be safer, but that is not necessarily the case
Systematic reviews have suggested that, overall, the error rate might fall, but the evidence is less clear when it comes to the rate of adverse drug events or serious errors, she points out
. She adds that evidence from other countries, such as the United States, may not be applicable in the UK.
Recent studies within NHS hospitals have found that EPMA systems prevent some errors but not others (see Box 2).
Box 2: Does e-prescribing in hospitals reduce errors?
In a study that looked at at the impact of inpatient e-prescribing on errors in an English teaching hospital, published in March 2019, Bryony Dean Franklin, a medicines safety and technology researcher at the UCL School of Pharmacy, found that there were fewer errors overall in electronic versus paper prescriptions (6.0% vs. 7.8%)
. However, analysis of the type of error showed a more complex picture. Errors involving incorrect doses and illegible or incomplete orders were less common with electronic prescribing (e-prescribing), but errors of duplication, omission, incorrect drug and incorrect formulation were more common with e-prescribing. Error rates also varied depending on the ward or patient.
“So much depends on how the system is introduced and used, and how people are trained to use it,” explains Franklin. “You might get some prescribers who feel they don’t want to engage with that system. There is a lot of complexity around it.”
Another study of e-prescribing in the NHS, published in June 2019, found that it did not prevent most harmful prescribing errors from occurring in paediatric care
. Researchers who assessed seven different e-prescribing systems across 15 hospitals found that 90% of erroneous prescription orders tested, such as prescribing the wrong dose or prescribing a medicine to someone with a known allergy, were able to be prescribed. In 74% of cases, no warning was presented to the prescriber when they were prescribing the error.
These findings are borne out by the experiences of pharmacists working on the frontline. Matt Elliott, a pharmacist with responsibility for implementing an EPMA system at University Hospitals of Derby and Burton NHS Foundation Trust, says the data show “errors change from one type to another” when adopting a new system.We are not going to stop all of these errors, but we can try to eliminate the new types of risks that can be introduced with electronic systems, he explains.
Hetal Halai, a senior pharmacist with responsibility for e-prescribing at King’s College Hospital Foundation Trust, found that while some errors dropped away dramatically with the introduction of the new technology, others, such as duplication, increased. The pharmacy team’s approach to avoid this is to ensure that the system at their trust is as intuitive as possible and to manage user expectations about how ‘clever’ it is.
In contrast, some trusts have reported a large drop in error rates once EPMA systems are up and running. A presentation from Harrogate and District NHS Foundation Trust — made available as a learning tool for other trusts — reveals that the trust has seen substantial falls in allergy incidents, missed doses and medicines administration errors
Kavi Gohil, senior pharmacist and EPMA lead at St Andrew’s Healthcare — a mental health charity with 70 wards serving 850 patients across four sites — says the introduction of its e-prescribing system cut medicines errors by 50%. “Errors that were reduced included missing signatures and consent to treatment requests,” he says. “Also we saw falls in transcription errors and illegible handwriting, high dose or incorrect frequency.”
Franklin says the safety of EPMA systems will depend on the hospital and how it tweaks the system to mitigate different types of errors. “Each hospital needs to be looking at the sorts of errors they’re getting, because the issues they find will be different and some of this is down to working practices,” she adds.
Like Punn, Gohil says that, initially, the EPMA system at St Andrew’s Healthcare was more time consuming as staff got used to new processes, although ultimately it ended up being more efficient.
Franklin’s work has also shown that EPMA systems may increase the time it takes for pharmacists to do some tasks — most notably screening inpatient medicines — although other tasks can be made quicker
. She also found that there is a danger that pharmacists spend less time with patients because they do not have to visit the bed to get the drug chart — a finding seemingly at odds with the drive to increase pharmacists’ clinical roles.
“There’s not been that many studies on the impact on pharmacists’ time but, in general, [EPMA systems] change how you work and do not necessarily make things quicker,” she explains. “Often pharmacists will find themselves having to come in and out of different screens to do a task.”
In general, EPMA systems change how you work and do not necessarily make things quicker
The difficulties do not stop with implementation. Elliott says there are unique challenges to switching compared with implementing EPMA systems. His trust has had an EPMA system in place for years, but it has reached the end of its useful life and is being upgraded to a new one.
His colleague Anthony Johnson, a pharmacy technician who is leading the switch, describes it as “a blessing and a curse”, admitting that, despite gaining several benefits, some functionality will be lost.
In the current electronic system, the standards for discharge go above and beyond the minimum expected: email alerts have been set up for critical medicines that have been prescribed but not dispensed, and there is a system to alert when a medicine is unavailable and [prescription orders] can be automatically sent to satellite pharmacists. None of this functionality is currently available in the new system.
“We have a lot of efficiencies so we’re thinking about how we’re going to have to adapt to maintain those in [the new system]. These are all things that we’re trying to get the supplier to adopt, but other organisations are not at the stage that we are,” says Johnson, explaining that others might be coming to EPMA for the first time, rather than switching. “There will have to be a certain amount of expectation management,” adds Elliott.
We have a lot of efficiencies so we’re thinking about how we’re going to have to adapt to maintain those in the new system
Accommodating the needs of different departments within the hospital is another challenge when upgrading EPMA systems. Halai says her team worked hard to do this when the hospital switched from one EPMA system to another in 2016.
“We found A&E was completely different because of the high level of patient turnover,” she explains. “We had to deal with issues around retrospective prescribing because, particularly in [resuscitation], medicines need to be given urgently and are then recorded afterwards.” There may also be the occasional need for verbal orders, she adds.
Other issues that were encountered during the upgrade included different discharge processes needing to be mapped and accounted for, and different security rights being set for nurses because of a high number of patient group directions.
In general, says Halai, getting clinicians on board is essential, and the requirements for each area need to be considered separately.
“There is a difference between the potential of EPMA and the reality,” she explains. “We have realised that sometimes, less is more.”
There is a difference between the potential of EPMA and the reality; we have realised that sometimes, less is more
The number of order sets (a pre-made group of medicines for a specific circumstance) has also been renewed and reduced to ensure that those available are useful, and in line with current policies and protocols. And the pharmacy team is working to reduce the size of the ‘pick lists’ within the drug catalogue to prevent picking errors.
Halai and her colleagues’ views have changed over the years, and they are now moving away from long lists of predefined, prefilled orders to something that encourages prescribers to think a little more and enables them to take ownership of their prescribing decisions.
Looking to the future
As EPMA systems continue to evolve, some trusts are now starting to move their focus from implementation to getting the best out of their systems.
“There are big questions about how we can make better use of all of the data in the systems — for example, on antibiotic prescribing,” says Franklin. Sheikh is starting a project to do just that: his team has been given a £2.5m grant to develop software for EPMA systems to improve antimicrobial stewardship.
“It is the kind of thing we need to see happening across a whole swathe of areas,” he says. “Commercial systems will go so far, but additional benefits we want to see in the NHS might be very different from the United States, where a lot of the systems are developed.”
Commercial systems will go so far, but additional benefits we want to see in the NHS might be very different from the United States, where a lot of the systems are developed
Similar projects in this area could include polypharmacy and minimising risk around prescribing in chronic kidney disease, he adds.
“Some of these examples are particularly relevant in the hospital context, but others will require connectivity across sectors — for example, between primary and secondary care — and at the moment that’s not really working in the UK,” he says.
In Wirral, where the trust has now upgraded to a newer EPMA system, there has been a lot of work on making it operate as safely as it can, for example, limiting the ability to have multiple records open at once and getting rid of system quirks around dose timing. If Power had his way, there would also be as few alerts as possible.
“The buzzword these days is optimisation,” he says, adding: “We’re in a much better place than we were a decade ago, but there is still a lot of variation in where people are at and in the systems themselves”. He believes the focus now needs to be on how best to present information so people actually take note.
Sheikh agrees, saying that EPMA systems are not a silver bullet and what is now needed is a lot of work to finesse what they can offer. “There are definitely benefits to be had but there are new problems to be introduced particularly if workflows are not being thought through.”
Some of the more technologically advanced steps include closed-loop prescribing and administration — a system that electronically verifies the medicine throughout the whole process from prescription to patient. The digital exemplars have been working on this and it has now gone live in some sites. This is a relatively new concept in the UK, with most research coming from the United States, but one trial of the system on a London hospital ward found it did reduce some errors and increased confirmation of patient identity
. However, the closed-loop system, which incorporated automated dispensing, barcode scanning to confirm patient identity and electronic medication administration records, increased the time taken to carry out medicines-related tasks.
NHS England has also developed a tool called ‘ePRaSE’ (e-Prescribing Risk and Safety Evaluation) to make it easier for trusts to check if their systems are meeting best practice on avoiding major errors. “It is a simulation tool designed to provide trusts with a way of identifying how well they have configured their system locally to avoid or reduce common medication risks,” explains Slee. “It is designed to provide support to improve systems. We will use the results to help identity what additional learning might be helpful.”
The tool should pick up common safety issues, as well as more subtle signs of potential problems, such as a high number of alerts that clinicians may start to ignore.
Slee hopes that the wealth of information now available from the experiences of those who have already implemented EPMA will make it easier for those yet to take the plunge.
“They will hopefully benefit from having access to a lot of resources, others with experience and the learning network,” she says, adding that while local leadership is incredibly important, trusts may well have other challenges. “For example some may not be as digitally mature or as advanced as those who already have systems — they may well be starting from a different baseline. Having good Wi-Fi connectivity across an organisation, for example, is critical — the lack of it is one of the most common causes of failure.”
But, Slee says, she would hope that by 2024 all acute and mental health trusts will have EPMA systems in place that meet national standards.
 National Assembly for Wales Public Accounts Committee. Medicines management. National Assembly for Wales Commission; March 2018. Available at: https://www.assembly.wales/laid%20documents/cr-ld11478/cr-ld11478-e.pdf (accessed July 2019)
 Elliott R, Camacho E, Campbell F et al. Prevalence and economic burden of medication errors in the NHS in England. Rapid evidence synthesis and economic analysis of the prevalence and burden of medication error in the UK. Policy Research Unit in Economic Evaluation of Health and Care Interventions. Universities of Sheffield and York; February 2018. Available at: http://www.eepru.org.uk/wp-content/uploads/2018/02/eepru-report-medication-error-feb-2018.pdf (accessed July 2019)
 Ranji SR, Rennke S, Wachter RM. Computerised provider order entry combined with clinical decision support systems to improve medication safety: a narrative review. BMJ Quality & Safety 2014;23:773–780. doi: 10.1136/bmjqs-2013-002165
 Franklin BD & Puaar S. What is the impact of introducing inpatient electronic prescribing on prescribing errors? A naturalistic stepped wedge study in an English teaching hospital. Health Informatics J 2019; In press. doi: 10.1177/1460458219833112
 Alldred A, Parkes E. Reducing medication errors with ePMA: 7 years’ experience. Harrogate and District NHS Foundation Trust; July 2018. Available at: http://www.eprescribingtoolkit.com/wp-content/uploads/2018/11/Reducing-Medication-Errors-with-EpMA-.pdf (accessed July 2019).
 McLeod M, Karampatakis GD, Heyligen L et al. The impact of implementing a hospital electronic prescribing and administration system on clinical pharmacists’ activities - a mixed methods study. BMC Health Serv Res 2019;19(1):156. doi: 10.1186/s12913-019-3986-4
 Franklin BD, O’Grady K, Donyai Pet al. The impact of a closed-loop electronic prescribing and administration system on prescribing errors, administration errors and staff time: a before-and-after study. Qual Saf Health Care 2007;16(4):279–284. doi: 10.1136/qshc.2006.019497