Objective Junior doctors have the highest rates of prescribing errors, yet no study has set out to understand the differences between completely novice prescribers (Foundation year one (FY1) doctors) and those who have gained some experience (Foundation year two (FY2) doctors). The objective of this study was to uncover the causes of prescribing errors made by FY2 doctors and compare them with previously collected data of the causes of errors made by FY1 doctors.
Design Qualitative interviews, using the critical incident technique, conducted with 19 FY2 doctors on the causes of their prescribing errors and compared with interviews previously conducted with 30 FY1 doctors. Data were analysed using a constant comparison approach after categorisation of the data using Reason’s model of accident causation and the London protocol.
Results Common contributory factors in both FY1 and FY2 doctors’ prescribing errors included working on call, tiredness and complex patients. Yet, important differences were revealed in terms of application of prescribing knowledge, with FY1 doctors lacking knowledge and FY2 misplacing their knowledge. Due to the rotation of foundation doctors, both groups are faced with novel prescribing contexts, yet the previous experience that FY2 doctors gained led to misplaced confidence when caution would have been expedient.
Conclusions Differences in the contributory factors of prescribing errors should be taken into account when designing interventions to improve the prescribing of foundation doctors. Furthermore, careful consideration should be taken when inferring expertise in FY2 doctors, who are likely to prescribe in contexts in which their experience is little different to an FY1 doctor.
- education & training (see medical education & training)
- risk management
- medical education & training
- qualitative research
- quality in health care
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
- education & training (see medical education & training)
- risk management
- medical education & training
- qualitative research
- quality in health care
Prescription errors are a cause of concern for healthcare organisations around the world.1 They are a significant cause of preventable adverse events leading to patient harm, lengthened hospital stays and even death.2 In UK hospitals, as internationally, most prescribing (70%) is undertaken by the most junior grades. Those doctors who have most recently graduated (ie, in their first 2 years of practice) have been found to have the highest rates of prescribing errors3 and their volume of prescribing means many patients are affected, generating some concern as to their prescribing competence and preparedness for practice.4 Doctors in the UK who have recently graduated medical school undergo a period of training (see box 1 for further explanation) over 2 years and are termed Foundation year one (FY1) and Foundation year two (FY2) doctors depending on which year of training they are completing.
Understanding the causes of errors made by those high-rate prescribers is the first step towards developing ways to reduce their incidence. The EQUIP study reported error rates across 20 different hospitals and found that FY1 doctors and FY2 doctors had similar error rates (8.4% and 10.3%, respectively) despite FY2 doctors having greater prescribing experience. Although studies have explored the causes of junior doctors’ prescribing errors,5 6 no study has explicitly set out to understand differences between completely novice prescribers (FY1 doctors) and those who have gained some experience (FY2 doctors). There are clearly differences between these groups of doctors—FY2 doctors are one year more senior and therefore more autonomous than FY1 doctors. There is also evidence that there are differences between grades of doctors in how they approach prescribing7 or what makes them feel uncomfortable about evidence-based prescribing.8 FY2 doctors are also more firmly established as work-based learners,9 but their high error rate shows that work-based learning has failed to teach safe practice. Given how the NHS is structured, it seems implausible that FY2 trainees’ prescribing errors have exactly the same causes as those made by FY1 trainees. To scope an effective intervention, good understanding of the trainees’ prescribing errors and the differences between the most novice prescribers and those who could be deemed intermediate prescribers is essential. This study sets out to uncover the causes of prescribing errors made by FY2 doctors and compare it with our previous data of the causes of errors made by FY1 doctors.
This study was based on an analysis of primary data collected from qualitative interviews with FY2 doctors with a secondary analysis of existing qualitative data10 collected from previously conducted interviews with FY1 doctors.11 Both sets of interviews collected data on the causes of prescribing errors using the critical incident technique (CIT) via one to one in-depth interviews. Prior to the interview, doctors were asked to recall their own prescribing errors and during each interview, errors were unpacked in detail using the CIT.12 The CIT enabled the interviewer to explore prescribing error scenarios with the participant (details provided in table 1) and unpicked influences that were not necessarily apparent to the prescriber themselves. The CIT also has the advantage that it ‘does not collect opinions and estimates but obtains a record of specific behaviours’.12 Furthermore, the CIT has been applied successfully by the authors before to explore prescribing influences.8 11 13
Interviews were conducted with doctors from a variety of hospitals in North West England who were following standard foundation training (Box 1). The FY1 interviews received NHS ethics committee and NHS management approval and all participants gave written consent for future use of anonymous data. The FY2 interviews were granted University Research Ethics Committee and NHS management approval.
UK medical doctor training structure
Postgraduate medical training stage one—The Foundation Programme:
Foundation year one or FY1 (previously called Pre-Registration House Officer or PRHO): doctors who have recently graduated from medical school. Satisfactory completion of the FY1 year leads to full registration with the General Medical Council.
Foundation year two or FY2 (previously Senior House Officer or SHO): doctors who remain under clinical supervision but with increasing responsibility for patient care. Satisfactory completion of FY2 will lead to the award of a foundation achievement of competence document (FACD) which indicates that the foundation doctor is ready to enter a core, specialty or general practice training programme.
Postgraduate medical training stage two—Specialty Training Programmes:
Specialty Trainees (previously Registrars): doctors training in a specialist area of medicine. The area of medicine that doctors choose will determine the length of training required before becoming a fully qualified doctor. There are two types of programmes:
‘Run-through’ (ie, without any break) training programmes which range from 3 years for general practice to 5 to 7 years in other specialties.
Core and higher specialty training programmes. In these programmes, trainees undertake core training which lasts two to 3 years, depending on the specialty. This is followed by an open competition to enter a higher specialty training post.
Consultant: doctors with a Certificate of Completion of Training (CCT) which is awarded after successful completion of the above training programmes.
See www.foundationprogramme.nhs.uk/pages/home/about-the-foundation-programme and www.bma.org.uk/advice/career/studying-medicine/insiders-guide-to-medical-specialties/medical-training-pathway for more details.
Reason’s model of accident causation14 and the London protocol15 were used to categorise and present the data. Reason’s model is the most commonly used theoretical model for prescribing errors16–18 and was used to categorise each error according to the active failure (table 2). The study focused solely on non-intentional errors and not deliberate deviations from procedures, that is, violations. The London Protocol, based on Reason’s model, provides a structured framework for classifying contributory and organisational factors that arise in a healthcare setting (table 3). Data were analysed using a constant comparison approach in which active failures (ie, errors), and any associated contributory factors (as set out in the London Protocol) were examined in detail. Each contributory factor type was considered to look for differences and similarities between the FY1 and FY2 data.
For the FY1 data, classification of the types of errors was carried out independently by PL and MT. For the FY2 data, classification was carried out independently by PL and ES. Any disagreements were resolved by discussion. The potential severity of each prescribing error was discussed by the team and assigned after consensus was reached.
Interviews with all 30 FY1 doctors and 19 FY2 doctors were included in the analysis. Interviews lasted 1 hour on average (range 20–80 min). Doctors worked in a wide range of specialities (n=24) and discussed 119 prescribing errors (FY1 n=85, FY2 n=34).
Both FY1 and FY2 doctors recalled many instances in which they failed to execute a good plan. Such slips and lapses frequently entailed prescribing the incorrect dose of a medication but other slips and lapses included prescribing the wrong medication, prescribing for the wrong patient and prescribing medication to which the patient was allergic.
For both groups of doctors, rule-based mistakes commonly involved wrong dose errors, prescribing drugs that interacted with patients’ existing medication and prescribing contraindicated medication. Knowledge-based mistakes were frequently discussed by FY1 doctors and nearly always related to a lack of knowledge of drug dosage. Though not as common in FY2 accounts, their lack of experience with medications used in specialities such as rheumatology and obstetrics led to knowledge-based mistakes with doses, licensed indications and omission of medication.
A comparison of contributory factors in FY1 and FY2 prescribing errors is shown in table 4. This next section will describe those differences in greater detail.
Work environment factors
Although busyness, due to a high workload, was cited by both groups of doctors as a contributory factor, the origins of their busyness highlighted distinct differences due to the environments in which doctors were prescribing. FY1 doctors discussed the pressure and busyness of ward rounds that made it difficult to for them to ‘double check’ their prescribing. FY2 doctors rotated onto clinical areas which were felt to be particularly pressurised or busy, such as Accident and Emergency or general practice. The former had particular pressures associated with not breaching the NHS target that 95% of patients should be admitted, transferred or discharged within 4 hours of arrival at Accident and Emergency:
…it was a very busy afternoon and it often is in A&E, lots of children coming in, lots of screaming noises, lots of people asking you questions on different cases and if you haven’t quite finished a case in the adult section, then they’ll be asking you if you’ve got a patient that’s almost breaching in terms of the 4 hours, they’ll be on your back. FY2-14
Working on call was a strong contributory factor in both novice and intermediate prescribers’ errors and was related to high workload, especially working across multiple wards. This FY2 felt that, in hindsight, he might have prescribed differently if he had been able to assess the patient’s response, as he would do during the routine day:
If I hadn’t been on call—if I hadn’t happened to have been having a couple of admissions pending in A&E that I needed to get in—then I would have stayed with her and assessed her response to treatment more accurately in real time; I didn’t want to be off the ward really but I sort of had to because I had other pressing things to do. FY2-11
A lack of support or supervision was a common theme from FY1 doctors’ interviews. These newly qualified doctors recalled how they prescribed with ‘basically no support at all’ from more senior doctors and they were expected to care for large number of patients, particularly while working overnight. Supervision was not frequently discussed by FY2 doctors, but was highlighted as a problem during overnight shifts. This FY2 doctor felt that this lack of support led her to mix up two medication charts consequently prescribing medication to the wrong children:
Paediatrics at night at the time was very, very hectic, there wasn’t very much supervision, it was just myself as an FY2 and one registrar covering all of paediatrics in a big hospital… FY2-16
The perception of pressure from the wider healthcare team, particularly consultants and nursing staff, was described by FY1 doctors who felt that this added pressure may have contributed to their errors as they rushed to complete the prescribing task. A doctor who knew the correct dose of enoxaparin prescribed it erroneously at twice the required dose, when he was an FY1, as he felt under pressure from nurses to prescribe quickly:
It wasn’t a knowledge thing, I knew that it was 1.5 per kilogram, but, again, it was just the external factors of everything else going on, having two Kardex’s in front of me, having a list of things here, trying to prioritise on an acute medical ward …there was no doctors putting pressure on me, it was just the nursing staff then. FY2-14
FY1 doctors’ accounts demonstrated the influence of others in their errors. Often doctors did as they were told by senior doctors, trusted nurses’ erroneous advice and unquestionably accepted patient’s accounts of their regular medication:
….So the Consultant was saying, We’ll write her up for Co-codamol, and then write her up for, I think it was something codeine based, or something like PRN, which I did and then after the ward round …the Pharmacist on the ward who’s absolutely lovely, came up to me and said, [Dr’s Name], erm, I've noticed a few, that you’ve written co-codamol and codeine, PRN. Erm, but it does say on the front of the chart that she’s allergic to codeine… a lot of the times you do those things because you're just in the middle of the ward round, the Consultant says something, you don't even question it. FY1-16
FY2 doctors compared the ‘solitude’ of prescribing in general practice with the team based hospital environment. Many FY1 doctors talked of the support they had from pharmacists which helped to prevent errors from reaching the patients. FY2 doctors, on the other hand, believed that their mistakes were less likely to be picked up as they now worked in departments in which there was no pharmacist cover. This FY2 prescribed an overdose of antibiotic as they had not taken into account the patient’s renal failure:
…on the ward usually I've got a pharmacist. In A&E it's different, we're not having any pharmacists or any…. on the ward we ask the pharmacists so frequently, especially in those patients with renal failure… FY2-3
Individual (staff) factors
Unsurprisingly, a lack of knowledge regarding a particular medication was most frequently described by FY1 doctors. This is discussed in detail in a separate paper.11 For FY2s, a basic lack of knowledge was less common and their errors usually related to applying the wrong rule in a new context:
I used to give clarithromycin all the time when I was doing respiratory medicine. It was very routine, that. I’d prescribe it and check if they’re on a statin, cross it off. But now I don’t know if it’s a contextual thing—I just don’t think to check if they are on it [a statin] FY2-8
FY2 doctors, despite having greater experience of prescribing in general, found themselves presented with scenarios that they had not encountered before as they were assigned to new, specialist areas and general practice as part of their training rotation. The wide range of patients, conditions and medications encountered in general practice was particularly difficult and increased their scope for error:
In hospital, you see the same three dozen conditions for most of your patients. GP is hundreds of tiny things that you see scattered through your week, it’s so varied. And it’s such a cliché, but every patient is different…and there are so many ways of making a mistake. FY2-8
As doctors accrued more knowledge, their confidence in prescribing increased. This confidence was associated with a reduced sense of risk. Sometimes they perceived they were dealing with a ‘simple decision’ or a familiar scenario so they did not check their prescribing. The danger of occasionally being unconsciously incompetent was even described by one of the FY2 doctors:
I think it’s really that crossover bit between not having a blind clue, checking everything, to actually knowing what you’re doing. It’s that little bit in the middle that you think you know what you’re doing but actually you probably don’t. And I think that’s where problems arise. FY2-9
Tiredness and fatigue caused by the physical demands of working long days and overnight were discussed by both FY1 and FY2 doctors. This FY1 doctor prescribed penicillin to a patient who was penicillin-allergic and felt that her tiredness attributed to the error:
…it was, like, a 4 day hot block where you do 13 hours shifts every day and it was at the end of it and I think I was just tired and stressed from all levels, and people were harassing you so you just forget about things and you don't be as careful you should be. FY1-19
Acutely and severely ill patients were associated with errors made by both groups of doctors. Such patients would have complex needs and normal prescribing rules may not be appropriate. One FY2 prescribed the wrong volume of fluids, because the patient was very poorly and this was not immediately obvious with her experience:
I had a patient who was—this was on-call—having an intra-abdominal bleed, but she was a very frail patient—elderly patient— and with advanced malignancy as well; and essentially she didn’t respond in the way someone would normally.FY2-11
However, unlike FY1 doctors who perceived a pressure to prescribe from the wider team,11 FY2 doctors often described how they felt indirectly pressured by patients because of the settings where they worked leading them to prescribe in haste:
When the department is very busy the waiting room, obviously, fills up and you do feel this pressure that, you know, when you call in the next patient and you look and there’s 10 more faces there, people get angrier and angrier because they’re waiting longer and longer, so there is a pressure to see patients, to do things slightly quicker and perhaps that’s one of the causes that errors slip through the net… FY2-14
Task and technology factors
The different settings in which FY1 and FY2 doctors commonly prescribe meant that they experienced different challenges. FY1 doctors grappled with drug charts that differed between hospitals. Such charts could have small boxes or folded over sections which could lead to omission of required medication:
I've missed off as-required medications before because they're on a separate page of the chart. So, I've gone through the regular medications, closed the chart and I’ve sent it off to pharmacy. FY1-15
FY2 doctors were confronted with learning to prescribe using a general practice electronic prescribing system and selection errors occurred as they were getting familiar with the system. Furthermore, FY2 doctors felt that because prescription writing is the last task within a consultation, it was a step that they sometimes rushed:
…when you get to the end of the consultation, you’re really thinking, ‘Oh, wow, my patient has been waiting three or 4 min and it’s only my second one,’ and you know, just get out the door…. And if you get to the end of the appointment…because that’s kind of the last thing you do, is print up a prescription. When you’ve got to that point, you just want to get them out the door again. FY2-8
Organisational and management factors
This contributory factor was a latent factor for all doctors in relation to the organisation of teams and workload overnight. Also, for FY2 doctors working in a highly specialised area, they experienced expectations that they were the ‘experts’. This FY2 had only been working in rheumatology for 2 weeks when he was faced with a novel prescribing decision:
I think, because we’re the specialist team, especially with rheumatology, it’s quite specialised and most people wouldn’t really question what you say, because you come in as a specialist, so no one really would know much difference….So she was, kind of, asking my opinion as a rheumatologist, which was a little bit daunting as I’m 2 weeks in. FY2-12
The ability to delegate particular tasks was hampered in particular specialties in which more junior FY1 doctors were not part of the team:
…ordinarily I would have delegated down to the FY1 some of the other stuff, but it just happened to be that the way this hospital is set up is that the admissions are infectious-diseases admissions, which FY1s don’t get involved in, so I had to do those even though it would have been better to delegate them and stay with the female. FY2-11
This is the first in-depth qualitative study to explore the differences and similarities in the causes of FY1 and FY2 doctors’ prescribing errors. The study found that there were common contributory factors in both the novice FY1 and intermediate FY2 prescribing errors that could be targeted for intervention. For instance, working on call, tiredness and fatigue and dealing with acute and complex patients appeared problematic for all foundation doctors. These factors have been identified in previous studies5 6 19 20 and interventions to target these factors are therefore likely to impact on the quality of prescribing across both grades of doctor. However, our analysis demonstrated some important differences that might allow for more targeted intervention and also go some way to explain why, despite their intermediate status of prescriber, the FY2 prescribing error rate is similar to that of an FY1 doctor.3
Previous research has identified the importance of work environment factors in junior doctors’ errors5 and the impact that this can have on checking practices. Our study highlights different sources of workload pressure experienced by these groups of doctors. Recently, The Royal College of Physicians produced guidance on how to support junior doctors to prescribe safely.21 In particular, it recommends the use of pauses on ward rounds, to ensure adequate time to prescribe before moving to the next patient. The lack of time given to prescribing on a ward round can also promote the belief that prescribing is unimportant. Tackling this attitude may improve overall safety culture. Both A&E and general practice traditionally lack pharmacist support, but recently there has been a drive to increase pharmacists in these areas as a defence against errors reaching the patient.22 23
Both FY1 and FY2 doctors regularly change clinical specialties as part of their training, with both grades being placed in specialities in which they were completely novice. The foundation curriculum states that doctors should, by the end of their FY1 year ‘prescribe safely’ and by the end of the FY2 year, they should ‘prescribe safely in differing environments’.24 Therefore, there are challenges when prescribing during both years, not just the FY1 year, that need to be overcome in order to complete the foundation training.
Prescribing in areas in which they were wholly unfamiliar was related to the occurrence of rule based mistakes due to erroneous transfer of knowledge from previous areas of practice. The challenge of transferring procedural knowledge has been described previously with FY1s.20 FY2s were more senior and felt there were expectations placed on them that they knew what they are doing. This expectation may be partly due to their full registration status (see Box 1) and the increased responsibility that FY2 doctors are expected to have in patient care. Unsurprisingly, patients who were very poorly with complex needs were associated with error, as has been reported elsewhere.6 Such patients are usually on multiple medications and both grades of doctor found such situations to be a contributory factor in some errors. A study from 2016 explored the training provision for junior doctors;25 they identified the neglect of FY2 doctors and an assumption of FY2 ability. Our study provides more evidence to support this.
There was also evidence that FY2 doctors could have misplaced confidence in their prescribing ability, which could lead them to omit a check on their prescribing. Overconfidence was described by Ryan et al in their questionnaire study of FY1 and FY2 doctors.6 Our study found that, while increased confidence in prescribing went hand in hand with increased experience, this was problematic when this confidence was misplaced, such as in a new setting where doctors were unaware of the complexity of certain prescribing decisions. Misplaced confidence could also be perpetuated by greater expectations placed on them by the wider team. It would be prudent for supervisors to provide support and oversight into FY2 doctors’ prescribing, as is expected with FY1 prescribing,21 to ensure safe prescribing continues to be engendered into the FY2 year.
FY2 doctors were exposed to patient pressure for the first time, due to practising in areas such as A&E, general practice and outpatient clinics,. Patient pressure has been shown to negatively influence prescribing decisions26 and, together with doctors rushing to complete task, introduced scope for error. Explicit training and support to manage these perceived pressures should be provided by more experienced prescribers throughout the foundation training programme.
Task and technology factors were reported by FY1 and FY2 doctors, as their interaction with both paper-based and electronic systems for prescribing were associated with errors. FY1 doctors described difficulties with medication charts but FY2 doctors instead described difficulties with the electronic systems for prescribing in general practice. This confirms some of the problems of electronic prescribing systems found previously27 leading to recent advice for junior doctors using electronic systems being provided by the Royal College of Physicians.21
Underpinning many contributory factors were wider organisational decisions often relating to the management of workload, and the structure of training and teams. Both FY1 and FY2 doctors could be adversely affected by the staffing schedule and resultant workloads. Tiredness and fatigue, especially relating to the physical demands of working on-call overnight, was a factor for both FY1 and FY2 doctors’ prescribing errors and has been commonly reported elsewhere.28 The environment in which junior doctors prescribe is complex, dynamic and often laden with pressure; high workload and a fluctuating and varying support network are common factors cited in junior doctors’ prescribing errors.5 6 19 29 Focusing on how doctors can optimise their performance under such pressure would be prudent and has been identified as perhaps more important than efforts to increase prescribing knowledge.30 Perhaps focus should turn to the strengthening of non-technical skills (ie, cognitive, social and personal resource skills)31 across the foundation years as there may be limitations on the impact that knowledge based educational interventions can have on prescribing in practice.32 However, nuances in the nature of the complexity for FY1 and FY2 doctors, as described above, should be taken into account so as to potentiate the success of improvement approaches.
As this study used self-report as a mechanism to unpick prescribing errors, there is a risk that participants responded to questions in way that they perceived as being socially acceptable (social desirability bias) or that their accounts lacked accuracy or detail (recall bias). Participants could have also exhibited hindsight bias, over simplifying the cause of errors and exaggerating their ability to have predicted their outcome. However, by having experienced researchers applying the CIT and encouraging participants to ‘understand how their decisions and actions made sense at the time’33 helped to minimise these inherent biases.
This study was based on secondary analysis of data from FY1 interviews that were collected some time ago. However, more recent studies with FY1s corroborate our findings and the Royal Collge ofPhysicians guidance was published only very recently, so we have little reason to expect much change.
Despite the extra year prescribing experience that doctors in their second year of foundation training possess, their experiences are not dissimilar to those who have just graduated medical school. Similar to FY1 doctors, FY2 doctors are expected to prescribe in new specialities and areas in which they have no prior experience. Therefore, it is logical that the prescribing error rate is similar across both groups. Furthermore, our findings demonstrate that FY2 doctors can possess a misplaced confidence in their prescribing decisions due their previous, although limited experience. Such overconfidence can result in rule-based mistakes that are exacerbated by a wider organisational expectation that FY2 doctors are more senior and therefore more experienced. We recommend that greater care should be taken when placing expectations on FY2 doctors as more experienced and that both foundation years should be perceived as similar level in terms of prescribing support that is offered and provided.
There are common contributory factors in both Foundation year one (FY1) and year two (FY2) doctors’ prescribing errors such as working on call, tiredness and fatigue and complex patients.
Differences were revealed in terms of application of prescribing knowledge, with FY1 doctors lacking knowledge and FY2 mis-applying their existing knowledge.
Such differences should be taken into account when designing interventions to improve the prescribing of junior doctors.
Care should be taken when inferring prescribing expertise in FY2 doctors so that appropriate prescribing support can be provided.
Current research questions
What are the differences in the causes of prescribing errors made by completely novice prescribers (FY1 doctors) and those who have gained some experience (FY2 doctors)?
What influence might these differences and similarities in the causes of foundation doctors’ prescribing errors have for the design of future prescribing interventions?
What is already known on the subject
Doctors in their first and second year of practice (FY1 and FY2 doctors) have the highest rates of prescribing errors and undertake the majority of prescribing within hospitals.
Understanding the causes of errors is the first step in designing interventions for their prevention.
Although we understand some of the causes of prescribing errors, the differences between completely novice prescribers (FY1 doctors) and those who have gained some experience (FY2 doctors) are yet to be explored.
Reason’s model of accident causation and the London Protocol are well established approaches to exploring the contributory factors in errors, allowing for the design of future interventions.
Thanks to all the doctors who participated in the interviews.
Contributors PJL and MPT designed the study. PJL and ES conducted the interviews. Analysis was carried out by PJL, ES and MPT. PJL wrote the first draft. MPT made revisions on subsequent drafts. All authors approved the final version.
Funding Funding was provided for the primary data collection of FY2 data by Manchester Foundation Trust Research for Patient Benefit Scheme.
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval Ethical approval was gained for the primary data collection of FY2 data from the University of Manchester Research Ethics Committee.
Provenance and peer review Not commissioned; externally peer reviewed.