Exploring the impact of pharmacist-led feedback on prescribing behaviour: A qualitative study

Research in Social and Administrative Pharmacy xxx (2017) 1e10

Contents lists available at ScienceDirect

Research in Social and Administrative Pharmacy journal homepage: www.rsap.org

Exploring the impact of pharmacist-led feedback on prescribing behaviour: A qualitative study M. Lloyd a, *, S.D. Watmough b, S.V. O'Brien c, N. Furlong d, K. Hardy d a

Pharmacy Department, St. Helens and Knowsley Hospitals NHS Trust, Whiston, Merseyside, L35 5DR, UK Postgraduate Professional Development, Faculty of Health and Social Care, Edge Hill University, Ormskirk, L39 4QP, UK c St. Helens CCG, St. Helens Chambers, St. Helens, Merseyside, WA10 1YF, UK d St. Helens and Knowsley Hospitals NHS Trust, Whiston, Merseyside, L35 5DR, UK b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 7 April 2017 Received in revised form 7 June 2017 Accepted 17 June 2017

Background: Prescribing errors occur frequently in hospital settings. Interventions to influence prescribing behaviour are needed with feedback one potential intervention to improve prescribing practice. Doctors have reported a lack of feedback on their prescribing previously whilst the literature exploring the impact of feedback on prescribing behaviour is limited. Objectives: To explore the impact of pharmacist-led feedback on prescribing behaviour. Methods: Semi-structured interviews were conducted with doctors who had received prescribing error feedback. A topic guide was used to explore the type of error and what impact feedback was having on prescribing behaviour. All interviews were transcribed verbatim and analysed thematically using a framework approach. Results: Twenty-three prescribers were interviewed and 65 errors discussed over 38 interviews. Key themes included; affective behaviour, learning outcome, prescribing behaviour and likelihood of error recurrence. Feedback was educational whilst a range of adaptive prescribing behaviours were also reported. Prescribers were more mindful and engaged with the prescribing process whilst feedback facilitated reflection, increased self-awareness and informed self-regulation. Greater information and feedback-seeking behaviours were reported whilst prescribers also reported greater situational awareness, and that they were making fewer prescribing errors following feedback. Conclusions: Pharmacist-led feedback was perceived to positively influence prescribing behaviour. Reported changes in prescriber behaviour resonate with the non-technical skills (NTS) of prescribing with prescribers adapting their prescribing behaviour depending on the environment and prescribing conditions. A model of prescribing is proposed with NTS activated in response to error provoking conditions. These findings have implications for prescribing education to make it a more contextualised educational process. © 2017 Elsevier Inc. All rights reserved.

1. Introduction Prescribing is a complex and high risk task.1,2 Prevalence studies have reported prescribing error rates of between 2% and 15%3e6 although this figure may be higher for junior doctors in their first two years of training,7 who also prescribe the majority of medications in hospital settings.3 Prescribing errors include errors in the decision making process and the prescription itself8 and can include incorrect dosing, frequency, quantity, indication,

* Corresponding author. E-mail address: [email protected] (M. Lloyd).

interactions and contraindications, or prescribing for an incorrect patient for example. Practical prescribing training is perceived to be suboptimal by medical students and junior doctors9 with dissatisfied feedback from recent medical graduates.10 Feedback has been described as the “cornerstone of effective clinical teaching”11 and has advantages for teaching in the clinical environment where prescribers cannot leave the ward area to attend teaching for example. However, prescribers have reported receiving little feedback on their prescribing previously3,4 with absence of feedback on prescribing errors as missed learning opportunities to learn from error, and to improve prescribing practice. Barriers to delivery or receipt of feedback could include limited opportunity from work pressures,

http://dx.doi.org/10.1016/j.sapharm.2017.06.010 1551-7411/© 2017 Elsevier Inc. All rights reserved.

Please cite this article in press as: Lloyd M, et al., Exploring the impact of pharmacist-led feedback on prescribing behaviour: A qualitative study, Research in Social and Administrative Pharmacy (2017), http://dx.doi.org/10.1016/j.sapharm.2017.06.010

2

M. Lloyd et al. / Research in Social and Administrative Pharmacy xxx (2017) 1e10

shift-work or annual leave, or lack of facilitator training for the effective delivery of feedback. A recent study12 reported positive outcomes following a pharmacist-led feedback intervention and described the impact on prescribing behaviour. However, the study was limited to antibiotic prescribing whilst the qualitative study included a limited sample size. Prescribers have reported valuing feedback on their prescribing elsewhere13 although little is known of the influence of feedback on prescribing a wider range of medications that is typical of hospital prescribing practice. Prescribing errors reflect a complex interplay of active failures, latent conditions, system failures and error provoking conditions.14 Active failures are unsafe acts15 and can be classed as mistakes (knowledge or rule-based), skill-based errors (slip or lapse) or violations.16 Feedback has the potential to improve prescribing and its influence is likely to vary depending on the type of prescribing error. However, the literature exploring the impact of feedback on prescribing behaviour is limited. An understanding of this could improve patient safety, enhance value from not having to correct errors, and inform future feedback interventions and prescribing pedagogy further. Therefore, the aim of this study was to qualitatively explore the impact of pharmacist-led, prescribing error feedback on prescriber behaviour.

issues were discussed between interviews. These themes and questions were underpinned by a review of the literature, personal insight of the researcher, previous research13 and research objectives. The topic guide consisted of two phases. The first phase was based on critical incident theory20 to inform classification of the prescribing error. This was followed by questions exploring what impact feedback was having on their prescribing and why this was happening. 2.3. Setting The study was conducted in a large acute hospital in the NorthWest of England. Interviews were conducted in a private interview room away from the clinical area. 2.4. Ethical considerations Relevant hospital and University of Liverpool ethics committees approved the study prior to data collection. Interview participation was voluntary and informed consent obtained verbally and in writing prior to commencing each interview. All data were anonymised. 2.5. Sampling and recruitment

2. Methods 2.1. The intervention The intervention was designed to reflect principles of effective feedback.11,17 Feedback was individualised per prescriber and delivered verbally and in writing using standardised proformas by a ward-based pharmacist who worked with the prescriber.13 Where a prescribing error was identified by a different pharmacist, for example on another ward or in dispensary, details of the error would be passed to the ward-based pharmacist who worked with the prescriber, to deliver feedback. The facilitating pharmacist used open questions to encourage the prescriber to reflect on the potential risk and reason for the error. The pharmacist would then encourage the prescriber to identify key learning outcomes and actions to reduce error recurrence with pharmacists providing input where needed. The facilitating pharmacist and prescriber would then sign each feedback form with the prescriber also asked to include a reflective statement in their training portfolio. The need for facilitator training in prescribing error feedback delivery has been reported previously18 and pharmacists were trained in the delivery of constructive feedback.11,17,19 Training consisted of a lecture, interactive workshops and reflective and peer-reviewed exercises to support pharmacists in their delivery of feedback. Pharmacists audited prescribing over a five-day period and then prepared feedback reports for their prescribers on overall prescribing. This was followed with further individualised feedback for any prescribing error classified as significant or greater. Error severity was graded as minor, significant, serious or potentially lethal as defined by research elsewhere,3 and reflected that used in frequent audits within the study setting by the pharmacy department. 2.2. Data collection Semi-structured interviews were used with individual prescribers to allow in-depth exploration of the impact of feedback on their prescribing practice. A topic guide (see supplementary material) was used to explore key themes and ensure consistent

Prescribers were eligible to participate if they received feedback from a pharmacist on an individual prescribing error in the previous week. This was to ensure they had sufficient memory recall of the event. Prescribers were recruited by ward-based pharmacists following delivery of feedback on their prescribing. Where prescribers expressed an interest to participate, the researcher followed up with a face-to-face discussion at ward level before arranging a mutually convenient time to conduct the interview. All prescribers (n ¼ 24) who were approached during recruitment expressed an interest to participate. 2.6. Analysis All interviews were digitally recorded and transcribed verbatim by ML with the sole exception to anonymise person and place names. Prescribing errors were classified according to Reason's error causation model14 as a knowledge based mistake (KBM), rule based mistake (RBM), slip, lapse or violation. A KBM is due to lack of knowledge of the medication or patient, whilst a RBM is due to misapplication of knowledge.14e16 A slip is an observable action where a plan is executed incorrectly, whilst a lapse is a failure of memory to perform a particular task, and a violation is where an individual intentionally deviates from best practice.14e16 Error classification was checked by a second (SDW) and third (SVOB) author for accuracy. A thematic analysis was undertaken by the researcher (ML) following the framework approach.21 This involves five stages of familiarisation, identification of a thematic framework, indexing, charting and mapping, and finally, interpretation.21 Transcripts were initially coded line-by-line with codes grouped into similar themes to produce an initial thematic framework. Codes were derived initially from the literature review and topic guide with further codes emerging from the data that could not be predicted. Reliability was enhanced by two members (SDW and SVOB) of the research team independently reading and analysing all transcripts. The three authors met regularly to discuss emergent codes and themes. Any discrepancies were resolved by a consensus. The emergent framework was then applied to the transcripts with further refinement of codes and themes through constant comparison, an approach typical of the framework method.21

Please cite this article in press as: Lloyd M, et al., Exploring the impact of pharmacist-led feedback on prescribing behaviour: A qualitative study, Research in Social and Administrative Pharmacy (2017), http://dx.doi.org/10.1016/j.sapharm.2017.06.010

M. Lloyd et al. / Research in Social and Administrative Pharmacy xxx (2017) 1e10

The final thematic framework was then mapped against the relevant participant quotes, with text copied and pasted into the conceptual framework to allow analysis and interpretation of the data.

3

4. Error recurrence Sample quotations were chosen by ML and agreed by the research team to demonstrate the experiences and views of participants.

3. Results

3.1. Affective behaviour

A total of 23 prescribers were interviewed (Table 1) and 65 errors discussed over 38 interviews. Some prescribers participated in multiple interviews where they received feedback on different prescribing errors over a period of time. The active failures consisted of 11 KBMs, 8 RBMs, 40 slips, 5 lapses and 1 violation (Table 2). Prescribers are denoted as R1-R23 as illustrated in Table 2 below. Interviews lasted between 8 and 33 min and an average of 18 min. One prescriber (Grade CT1) could not be interviewed during the recruitment phase as the error had occurred more than a week ago. Similarly, three prescribers (R1, R6 and R10) were eligible for further interviews but were not approached as the error had occurred over a week ago and memory recall would likely be limited. The reasons for the time delay were typically the prescriber being on annual leave or working night shifts. Two prescribers (prescriber R6 and R16) declined further interviews as they had already participated in the study and didn't feel the need to participate again. The final thematic framework can be viewed in Table 3. All prescribers were overwhelmingly positive of receiving feedback and openly discussed the impact feedback was having on their practice. Themes are discussed in further detail under four key headings:

3.1.1. Assertive behaviour For mistakes, prescribers reported being more assertive in challenging hierarchy, avoiding assumptions that more senior prescribers are always correct, or in seeking further information to inform their prescribing.

1. Affective behaviour 2. Learning outcome 3. Prescribing behaviour

Table 1 Participant information for semi-structured interviews.a Prescriber Prescriber Male or Years Qualified at Number of Number of code grade female time of interview(s) interviews errors discussed R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 Total

FY1 FY1 FY1 FY2 FY1 FY2 CT2 CT2 FY1 ST6 ST4 FY1 CT1 FY1 FY1 FY1 FY1 FY1 FY1 FY2 CT1 CT1 FY1

Female Female Female Male Female Female Female Male Male Male Male Male Male Female Female Male Female Female Female Female Female Male Female

1 1 1 2 1 2 4 5 1 8 6 1 3 1 1 1 1 1 1 2 3 3 1

4 2 3 3 1 2 2 1 3 1 1 1 1 1 1 1 1 2 1 1 1 2 2 38

9 3 3 3 1 3 2 2 3 3 1 2 3 3 1 3 2 5 3 1 4 2 3 65

a Prefix FY ¼ Foundation Year trainee, CT ¼ Core medical training doctor, ST¼ Specialist Trainee, suffix 1e6 ¼ year of training programme. Foundation grade doctors are the most junior at 1e2 years post-graduation, then CT grade doctors who are typically years 3e4 post-graduation, and then ST grades who are in specialist training positions before consultant level posts.

R16: “If they say oh go with 1g tds [three times a day] then I would usually go with what they say like if they said go with this, they are the experts … that [feedback] would be like a prompt now I guess for me to say and by the way they have a reduced creatinine clearance what dose would you go for?” [KBM] For skill-based errors, prescribers reported being more assertive in managing disruptions and communicating such issues with team members. R17: “And definitely after [ward pharmacist] pointed that out then I have … like when I am doing a kardex [inpatient medication chart] re-write or something and someone interrupts me I'll say no! [puts hand up] and I did that to [consultant] the other day … he had a group of medical students and I was writing this complicated digoxin loading dose and I just said “I'm sorry, but I'm doing this otherwise I'll make a mistake” [laughter] and I won't re-write a kardex on the ward round now, I'll put it on my list and I'll do that later because I can't multi-task.” [Lapse]

3.1.2. Reflective practice For all active failures, prescribers reported reflecting on-action to determine the error causation and for-action for appropriate corrective actions to minimise error recurrence. R21: “It's good because really this is the kind of thing that you should be talking … about isn't it and you know what are you doing wrong, what can you do to address it?” [Slip] Advancing on this, where prescribers reported prescribing a similar drug or being in a similar situation, they were reflecting-inaction to become more mindful and inform their decisions with the feedback session functioning like a cue. R4: “I was thinking about what [ward pharmacist name] said and checking the dose, frequency and stuff and writing fifteen minutes before their meal.” [KBM]

3.1.3. Self-awareness Prescribers consistently reported a raised awareness of prescribing errors, their risk, and ease at which they can occur. This was challenging the self-perceptions of prescribers by highlighting and discussing prescribing errors that they may otherwise have not known about. By raising awareness of prescribing errors, prescribers can learn and adapt their prescribing behaviour accordingly. R11: “It just reminds you how easy it is to make errors because with the law of averages you must go through an average year making x

Please cite this article in press as: Lloyd M, et al., Exploring the impact of pharmacist-led feedback on prescribing behaviour: A qualitative study, Research in Social and Administrative Pharmacy (2017), http://dx.doi.org/10.1016/j.sapharm.2017.06.010

4

M. Lloyd et al. / Research in Social and Administrative Pharmacy xxx (2017) 1e10

Table 2 Overview of Active failures discussed during prescriber interviews. Error type

Mistake Skill-based error

Prescriber Grade

KBM RBM Slip Lapse

Violation Total

FY1 (n ¼ 13)

FY2 (n ¼ 3)

CT/ST (n ¼ 7)

Total (n ¼ 23)

7 (16.3%) 4 (9.3%) 29 (67.4%) 2 (4.7%) 1 (2.3%) 43

1 (16.7%) 0 3 (50.0%) 2 (33.3%) 0 6

3 (18.8%) 4 (25.0%) 8 (50.0%) 1 (6.3%) 0 16

11 (16.9%) 8 (12.3%) 40 (61.5%) 5 (7.7%) 1 (1.5%) 65

Table 3 Thematic framework. Category

Code

1. Affective behaviour

Assertive behaviour Reflective practice Self-awareness Self-regulation Emotional impact Task prioritisation Prescribing knowledge Self-detection of errors Raised situational awareness Information seeking behaviour Mindful prescribing Systematic approach to prescribing Prescribing location Specific vs. general learning Facilitator variability

2. Learning outcome

3. Prescribing process/behaviour

4. Likelihood of Error Recurrence

amount of mistakes and no one ever tells you about it and until you get an intervention such as this you don't learn.” [RBM] This self-awareness highlighted prescribing competence, limitations in practice and areas for improvement. Without feedback, there was a sense that development of prescribing competence could be limited as one prescriber articulated. R17: “I hope that the feedback continues because without it you are unconsciously incompetent and with it you are consciously incompetent”. [KBM]

3.1.4. Self-regulation By raising self-awareness, it was consistently advocated that feedback informed prescribing, allowing prescribers to regulate their prescribing practice and adapt their prescribing behaviours, facilitated by the solution focused approach. R18: “I do also think that it is useful to get feedback and pull out specific things, because it's quite difficult to make a change unless you have that feedback and know what you have done and what you need to do.” [Slip] There was a general sense amongst prescribers that prescribing can become a routine task with ongoing feedback needed to continually self-regulate and prevent bad habits resurfacing. R3: “We haven't had that many TTO [To Take Out discharge prescription] errors recently so we're improving … I have been checking more again although I don't know if that is just a rebound effect of getting criticised and then you go a wee while and you're like oh that's okay I'm alright now.” [Slip]

3.1.5. Emotional impact Feedback was unanimously welcomed and valued by all prescribers with the process providing reassurance that they were developing as prescribers. However, for mistakes, there was a sense of embarrassment, disappointment and self-criticism for deficits in their own prescribing knowledge amongst prescribers. Whilst it was acknowledged that you could not know every single prescribing caveat, the importance of learning from error was outlined whilst feedback highlighted the importance of seeking appropriate information as reported later under prescribing practice. R4: “I was glad to have got the feedback because there are going to be lots of people on simvastatin and amiodarone here [cardiology ward] potentially. I was disappointed perhaps that I hadn't known it in the first place but that was purely a knowledge base but you can't know everything and learning from it is the most important thing.” [KBM] For slips and lapses there was a sense frustration at making an avoidable error that they could have identified themselves with a second check for example. This frustration was accelerated where workload pressures or distractions were identified as key contributing factors to the error. R6: “I think … this is just a slip really so it is just a bit frustrating really that I didn't identify it and I think that the feedback is a positive thing rather than a negative thing.” [Slip]

3.1.6. Task prioritisation Prescribers reported a greater prioritisation of prescribing tasks following feedback to limit their errors. Several prescribers reported commencing discharge prescriptions in advance of discharge to ease pressures on their workload later and avoidance of multi-tasking that could cause slips. This appeared to be in countenance to previous practice where prescribing was routinely undertaken at the same time as another task or during a busy ward round for example. R10: “I think that it can definitely help them I think that errors will always occur but I do think that it helps you to prioritise your TTO's, don't bunch them and be more errr … thorough, to save any further embarrassment.” [Slip] For lapses, prescribers reported acting on prescribing jobs immediately instead of leaving them to a later time and risking forgetting to change a prescription or commence a new medication for example. R2: “I should have written it down and prescribed it there and then and then at least it would have been prescribed instead of saying oh I must do that.” [Lapse]

Please cite this article in press as: Lloyd M, et al., Exploring the impact of pharmacist-led feedback on prescribing behaviour: A qualitative study, Research in Social and Administrative Pharmacy (2017), http://dx.doi.org/10.1016/j.sapharm.2017.06.010

M. Lloyd et al. / Research in Social and Administrative Pharmacy xxx (2017) 1e10

The need to prioritise prescribing created conflict at times with junior doctors in particular, struggling to prioritise tasks and workload. 3.2. Learning outcome Several learning outcomes were reported from the feedback intervention. 3.2.1. Prescribing knowledge Whilst prescribers acknowledged they could learn from any error, it was largely mistakes that were reported to improve their drug related knowledge with education from the pharmacist making the process more meaningful and memorable. This was not restricted to specific medications but also to specific situations such as dosing in pregnancy, renal impairment or checking for drug interactions. R4: “Well you get the form and the contextualisation in terms of the interaction and what it is makes it helpful and you then remember it because you can't just remember a list of numbers oh this dose with this and this dose with this. So knowing the pharmacology behind it helps you to remember and so you can then apply it to other situations and it reminds you about the inducers and inhibitors [of enzymes] situations.” [KBM] For slips, feedback improved specific drug-related knowledge when the incorrect dose or frequency was chosen for example, and the drug specific prescribing limits discussed during feedback. However, the main learning outcomes extended beyond drug specific knowledge for these errors, and are reported under other codes for example self-detection of errors, mindful prescribing, or prescribing location below, where prescribers alter their prescribing practice to limit or identify slips. 3.2.2. Self-detection of errors Typically, prescribers reported identifying more of their own errors following feedback. The types of errors that were selfdetected appeared to be skill-based slips and was informed by feedback on all error types with prescribers more careful and mindful with their prescribing, or adjusting their practice to prevent or identify their own skill based errors. R3: “And I have actually found a couple of times when I have made mistakes and like when a dose was ten but actually it was forty and stuff like that so I have picked up some of my own errors from rechecking it. So, that is worthwhile doing.” [RBM] For mistakes, it was acknowledged that it can be difficult to identify an error where you are confident it is correct. Seeking more information and feedback on their prescribing was outlined as potential solutions to these mistakes, but equally, would not be feasible for every prescribed item. R15: “So when you are sure of something then you don't look it up for example you don't look up every dose of every drug that you prescribe. For example you prescribe Tazocin and you know the dose is 4.5g so you don't look it up. You prescribe paracetamol you know the dose is 1g so you don't look it up, you know if you are sure of something then you don't look it up.” [KBM] Through raising awareness of prescribing errors, some junior doctors reported challenging senior prescribers, such as

5

consultants, and seeking further information from them to inform their prescribing, especially if a pharmacist was likely to provide feedback on an error that they make. R16: “Now I have realised that then when I am prescribing … I am looking through the drug chart and speaking to the consultant and saying what are we doing with this? How many days do you want?” [Slip]

3.2.3. Raised Situational Awareness For skill-based errors, prescribers reported having a raised awareness of the situation when they are prescribing, such as distractions and other error provoking conditions, and the need to be vigilant of and manage these causative factors. These behaviours were reported following delivery of feedback, and in particular, identification of causative factors and solutions to limit these for future practice. R17: “Distraction wise, I have an increased awareness now and perhaps better communication with the nursing staff too, just please don't interrupt me when I am prescribing.” [Lapse] This was not limited to external factors with several prescribers reporting taking forced breaks following feedback discussions when recognising that they were tired, hungry or stressed to limit the impact on their prescribing, and in particular, the risk of them making a skill-based error such as a slip or lapse. 3.3. Prescribing practice Feedback was unanimously considered to improve prescribing practice with improved knowledge as reported above whilst the impact on other prescribing skills was also notable. 3.3.1. Information seeking behaviour For mistakes, prescribers were consistent in advocating that they were seeking information at the point of prescribing to guide and inform their decisions. This included communicating with pharmacists more and where they were guided by a more senior prescriber to initiate a medication, prescribers reported seeking more information on the dose and duration. This extended beyond drug specific medicines information, to seeking technical information on renal function, and reviewing the medication chart for potential drug interactions or duplication for example. R7: “I think as well now that I will be extra cautious and if I can't find the information or get on toxbase then I'll just call the on-call pharmacist for advice. You don't want to bother them but it is very easy to make mistakes especially when I'm not confident in a specific area.” [RBM] For slips, prescribers reported asking pharmacists to check their prescribing before submitting discharge prescriptions in particular. Prescribers also reported proactively seeking and acting on written pharmacy communication in the medical notes or on the medication chart, information that they may have previously not paid attention to. R18: “Today when I was looking at the kardexes I was making sure that I was super-duper doubly looking through them and I saw that there were pharmacy comments and I do find that I now look for green pen [colour ink that pharmacists use] whereas I perhaps

Please cite this article in press as: Lloyd M, et al., Exploring the impact of pharmacist-led feedback on prescribing behaviour: A qualitative study, Research in Social and Administrative Pharmacy (2017), http://dx.doi.org/10.1016/j.sapharm.2017.06.010

6

M. Lloyd et al. / Research in Social and Administrative Pharmacy xxx (2017) 1e10

wouldn't have been … not interested as … not disinterested but I wouldn't have really thought to look.” [Slip]

3.3.2. Mindful prescribing Prescribers reported being more conscious of their prescribing instead of it being a routine task as reported earlier. This was informed by a raised awareness of errors and the potential emotional impact from making the error. This included risk of harm and embarrassment as reported earlier, potential outcomes reported as motivating factors to change prescribing practice. R3: “I think that I'm more aware that it is being looked at and for example I'm paying more attention than if you wasn't doing it. Now I'm like if this is wrong it costs time for somebody else and if it isn't it gets done quicker and you don't end up in that position because I'm thinking well if I can get it right the first time then I don't have to get it fixed later and I don't have to have it as a black mark as such … I knew that she would spot the error and come and tell me off so I would put a bit more effort in.” [RBM] For skill-based errors in particular, prescribers reported investing more time and not rushing the task to ensure their prescription was correct first time, every time to limit avoidable errors. R18: “I try to be a lot more careful and I try to slow down with them.” [Slip]

3.3.3. Systematic prescribing Prescribers reported adopting a more systematic approach to prescribing and introducing safeguards to minimise error; practices reinforced by feedback. This included seeking appropriate information to limit mistakes, not assuming others' prescribing was correct to limit RBMs, and double checking of prescriptions to limit slips for example. R7: “It is just about being methodical so even if you are busy you keep going through those basics and making sure that you are not missing anything and double checking.” [RBM] Prescribers reported that skill-based errors would be difficult to eliminate entirely because of the working environment. However, introducing second checks into their prescribing practices could reduce these errors; errors they reported detecting earlier. Prescribers reported specific changes to their prescribing to limit specific types of error. For example, to facilitate error prevention and detection, prescribers would print off the discharge summary to check their prescribing as opposed to checking it on the computer screen. For electronic prescribing, prescribers reported typing in the full drug name as opposed to a few letters to limit selection errors. These appeared to be coping mechanisms to avoid specific errors whilst it was also suggested that the feedback was continually refining prescribing practice. R9: “Well I think that the feedback is definitely improving the way that I prescribe you know it's just like little things here and there” [Slip]

3.3.4. Prescribing location The feedback process encouraged reflection on the prescribing environment and the suitability of the working environment for prescribing was questioned. Prescribing was considered

ubiquitous, with prescribers acknowledging that they would often prescribe mid-ward round or on busy clinical stations. For skillbased errors, prescribers described increased assertive behaviour in response to distractions. Some prescribers reported refusing to complete discharges or re-write discharge prescriptions mid wardround as reported earlier. Additionally, several prescribers reported changing locations where distractions could not be mitigated. R5: “I've changed computers or I'll use this room more [a quiet doctor's office] and I'll read through them again, even read through the TTO aloud when I'm doing the TTOs.” [Lapse]

3.4. Likelihood of error recurrence The potential for feedback to reduce error recurrence was reported. 3.4.1. Specific vs generic learning Prescribers reported that they were unlikely to repeat the same mistake twice as they would be looking for that specific issue in future. This was informed by improved knowledge for mistakes, and reflection-in-action for specific errors and situations as reported earlier. R4: “It is certainly something that I will look for and have a glance to see if they are on simvastatin. I'm certainly looking for it now. Yeah I mean I'm more aware now and so will look for it [the interaction] …. But it was drug specific.” [KBM] Despite outlining the importance of seeking appropriate medicines information to reduce mistakes, prescribers acknowledged that any error reduction was likely to be limited to specific situations discussed during feedback, such as dosing in renal failure or pregnancy for example. R7: “I think that this particular case and the tools that I have been given would really just stick with the [prescribing in] pregnancy because I do know that there are renal formularies and stuff but I still don't know how to get hold of them!” [RBM] For skill-based errors, prescribers reported that the learning outcomes are generic and would not be limited to the specific error or medication. By taking more time, prioritising prescriptions and adopting a more systematic approach to their prescribing, all skillbased errors could be reduced. R12: “With these it's not a mistake or a knowledge issue, it's about the skill of prescribing” [Slip] For skill-based errors, there was a sense of futility regarding feedback amongst some prescribers as PEs were always likely to occur unless the system was changed to support prescribing, and eliminate error provoking conditions. R4: “Like the previous error that we discussed last time was more of a knowledge thing so that was why, whereas this one was just a distraction … a work-based or workload one so risk will be reduced but risk will always be there.” [Lapse]

3.4.2. Facilitator variability There appeared to be variations in facilitator approach to feedback. For mistakes, it was reported that some pharmacists

Please cite this article in press as: Lloyd M, et al., Exploring the impact of pharmacist-led feedback on prescribing behaviour: A qualitative study, Research in Social and Administrative Pharmacy (2017), http://dx.doi.org/10.1016/j.sapharm.2017.06.010

M. Lloyd et al. / Research in Social and Administrative Pharmacy xxx (2017) 1e10

discussed the detail behind a specific drug interaction, choice of dose or contraindication for example. Others appeared to take this further by delivering bespoke educational sessions or provide further examples to contextualise the learning. R4: “The feedback was very good and I was glad to have the feedback, you need to know about your errors so you can correct them. We discussed the differences and why it happened and she talked through the insulin education aid although I had had that last year as well.” [RBM] Similar findings were reported for skill based errors with some pharmacists advising prescribers of the need to ‘be more careful’. Other pharmacists appeared to be far more constructive with their feedback, identifying specific solutions to mitigate further problems or providing further examples and education to contextualise the situation. R19: “For the Seretide [an inhaler] device, [ward pharmacist] also brought up some inhalers to show me the difference. It does just remind you that they are different and what one patient can use isn't always the same as what another patient can use.” [Slip] Two prescribers suggested that where feedback was not particularly constructive, the potential impact went beyond the learning outcome and negatively influenced the rapport and working relationship with the pharmacist. R3: “With my newly qualified pharmacist she would be like oh well you've done this wrong but it's okay I'll just fix it for you so there was less pressure on me to get it right the first time. Whereas … I'm not intimated by [ward pharmacist] you know, but she is clearly very good at her job and you want to emulate that yourself.” [RBM]

4. Discussion This study was undertaken to determine the impact of feedback on prescribing behaviour. Feedback was perceived to positively influence prescribing practice at a knowledge based and technical level, although the greatest influence appears to be on the development of non-technical skills (NTS) required for safe and appropriate prescribing. NTS include the social skills of communication, team working and leadership and the cognitive skills of situational awareness and decision making.22 These skills also include error awareness and professional responsibility and are increasingly recognised as integral to safe prescribing.23,24 NTS reported as contributing to PEs include poor communication, managing workload, poor teamwork and supervision, and impaired prescriber well-being.3,25 A taxonomy of NTS required to prescribe safely has been proposed26 and includes:    

Situational Awareness Decision making Communication and team working Task management

Depending on the type of error, prescribers can use various nontechnical skills to limit the risk of a prescribing error. The adaptive behaviours reported in this study are presented in Table 4 below against the type of active failure (mistake or skill-based error) and non-technical skill. Prescribing is a complex skill1 littered with opportunities for

7

error.2,23 McLellan et al.27 advocated that the skill of prescribing is greater than the sum of its parts. They also proposed a model of prescribing as an integrated set of complex skills, involved in the expert performance of it, where the prescribing task interacts with the system and the authors agree with this unreservedly. However, NTS are not highlighted explicitly as a ‘part’ and irrespective of cognitive and technical prescribing abilities, errors are likely to occur where NTS are deficient. It is these NTS that are dynamically engaged with the system to influence the prescribing outcome. The authors propose that NTS (Fig. 1) are separate from other prescribing skills as they are latent and interact with the clinical environment, guided by situational awareness, and could be considered a hidden skill set for safe prescribing. The need and type of NTS to execute a safe and appropriate prescription will depend on the environment. This echoes the sentiments of McLellan et al.27 who propose that the level of prescribing effort or ‘cognitive engagement’ for a successful prescribing outcome will vary depending on the situation, mirroring the proposal that the level of skill required for a successful prescribing outcome, will vary depending on the working environment, social context and error provoking conditions. Feedback raises self-awareness, situational awareness and performance deficits,19,28 outcomes reported by participants in this study. By raising awareness of prescribing performance, prescriber perceptions are altered allowing calibration of their behaviour to achieve desirable prescribing standards, behaviour consistent with perceptual control theory.29 Feedback also identifies reasons for deficits in performance and solutions to address them, for example the NTS listed in Table 4. It is this that allows self-regulation of prescribing behaviour and is inextricably linked to reflective practice. Feedback encouraged reflection on, for and in-action with prescribers taking a step back to pause for thought. This reflection informs self-regulation of prescribers to control their cognitive processes. Such meta-cognitive processes have been reported to minimise cognitive error and invoke the “conscious mind”,30 with reflection facilitating recall of previous problems and limitations in practice.31 Such adaptive processes are possible from the direct effects of feedback and negotiated outcomes, but it is also possible that prescribers may adapt their behaviour in anticipation of feedback. Here, the anticipation of feedback could motivate prescribers to focus more on prescribing to avoid unfavourable feedback, unfavourable comparison to peers, or be viewed in a negative light by a colleague, a feedback ‘cost’ described elsewhere.32 In this study, participants reported changes in their prescribing behaviour following feedback and where this improves prescribing practice, there is potential to reduce error and patient harm whilst improving value and work flow. However, further quantitative research is necessary to corroborate these views. Prescribers have reported valuing and welcoming prescribing error feedback previously,13 with potential to enhance working relationships with pharmacists. Where inter-professional working is enhanced, there is potential to optimise the role and skills of pharmacists in prescribing decisions, as opposed to intercepting prescribing errors retrospectively. However, the impact on pharmacists of delivering this intervention is unknown and requires further investigation to explore its impact on pharmacists and pharmacist-prescriber relationships. It has been proposed that “there might be value in structured education to ensure uniform safety and nontechnical skill acquisition”.23 Learning to prescribe has been described as a social exercise3 and the interplay of non-technical skills with the environment and need for feedback to support prescribing endorses this. Prescribing education should be contextualised to raise awareness of error provoking conditions and application of relevant non-

Please cite this article in press as: Lloyd M, et al., Exploring the impact of pharmacist-led feedback on prescribing behaviour: A qualitative study, Research in Social and Administrative Pharmacy (2017), http://dx.doi.org/10.1016/j.sapharm.2017.06.010

8

M. Lloyd et al. / Research in Social and Administrative Pharmacy xxx (2017) 1e10

Table 4 Reported prescriber behaviours following feedback on specific error types. Error type

Non-technical skill

Reported NTS outcome of feedback

Mistakes (KBMs and RBMs)

Situational Awareness

    

Decision making

 Communication  and team working   Task management





Skill based errors Situational Awareness (slips and lapses)

    

   

 

  Decision making

  

Communication  and team working 

     Task management





Raised awareness of own limitations Information seeking behaviour Access pharmacist more Request second opinion on prescribing Information seeking at point of prescribing to inform decision more Access relevant guidelines and information resources Documenting course duration for antibiotics or steroids for example Clarifying/challenging prescribing decisions with senior colleagues Feedback seeking behaviour to inform practice Prescribe in context i.e. review all medications not just what you prescribe Check allergy status, renal function, comorbidities before prescribing Adhere to best prescribing practice Raised awareness of error provoking conditions Raised self-awareness of own errors Be ‘mindful’ of prescribing task (especially when on-call) Ensuring checking relevant bloods routinely when prescribing for potassium supplementation for example Aware of impact on patient safety and team members Prescribing mindfully and treating prescribing as a high-risk task Introduce safeguarding practices such as accuracy check of own work Control emotions when under stress or remove self from area until calmed down Force self to take a break when tired/ hungry/thirsty Avoid completing discharges immediately after a long/busy ward round Prescribing in a quiet location Slow down when aware that you are rushing Information seeking at point of prescribing to inform decision more Access relevant guidelines Change prescribing location if cannot manage distractions Asking for second checks before completing/submitting discharge prescriptions Communicating risks of distractions with colleagues (for example nursing staff) Challenge distracting/disruptive behaviour Refining working practices (i.e. jobs lists) to limit distractions Communicate workload demands with seniors Ask for jobs to be documented to prevent lapses in memory Seek more information when prescribing unfamiliar medications Changing routine of prescribing to limit previous mistakes (i.e. check allergy status first) Change location if cannot manage distractions

Table 4 (continued ) Error type

Non-technical skill

Reported NTS outcome of feedback  Develop a more systematic routine for own prescribing  Try and commence discharges in anticipation of discharge to limit impact on workload  Utilise other non-medical prescribing staff when workload high  Avoid multi-tasking  Avoid completing discharges or rewriting medication charts on a ward round  Prioritise workload to minimise risk of rushing  Slow down when aware that they are rushing  Print off electronic prescriptions to check on paper before submitting to pharmacy  Do not assume pharmacists will correct errors for you

technical skills to mitigate them. It is recognised that effective prescribing education is difficult,33 but equally prescribers have expressed feeling unprepared and lacking confidence to prescribe.3,30 The transition from undergraduate student to junior doctor is challenging34 and provision of prescribing support from pharmacists could help to prepare and develop prescribers. Agenda-led, outcome based feedback, should be used to facilitate achievement of prescriber competence35 via a mentor. This feedback should ideally commence at undergraduate level to encourage inter-professional learning and feedback seeking behaviour from medical students with pharmacists. Feedback should form part of an integrated prescribing curriculum with immediate feedback following contextualised learning that persists into a prescriber's formative years to develop the transferrable NTS required to prescribe safely. This approach could facilitate greater team-working and collaborative prescribing decisions upon graduation. Such skill acquisition could commence through simulated scenario training, inter-professional learning, or ‘pre-prescribing’36 for example with pharmacist-led feedback interlinked throughout. The feasibility of these interventions are unknown and require further investigation to understand both the short-term and longer-term effects on prescribing practice and inter-professional collaboration. 4.1. Strengths and limitations This is the first known study exploring the impact of feedback on general prescribing practice. These findings can inform why feedback interventions may improve prescribing and the need for further contextualised prescribing education. This study is not without limitations. Not all grades of prescriber were recruited although foundation and trainee grade doctors prescribe the majority of medications. Equally, the impact of prescribing behaviour following feedback on minor errors was not explored so it is unknown if the same behaviours are true for these errors. Prescribers were required to recall specifics of an individual error to inform classification. This may have been hindered by memory failure and recall but equally mitigated by only interviewing prescribers who had both made a PE, and received feedback in the previous week. 5. Conclusion Pharmacist-led prescribing error feedback is well received and

Please cite this article in press as: Lloyd M, et al., Exploring the impact of pharmacist-led feedback on prescribing behaviour: A qualitative study, Research in Social and Administrative Pharmacy (2017), http://dx.doi.org/10.1016/j.sapharm.2017.06.010

M. Lloyd et al. / Research in Social and Administrative Pharmacy xxx (2017) 1e10

9

Fig. 1. Model of prescribing illustrating integration of technical and nontechnical skills with the working environment in response to feedback.

valued by prescribers with potential to positively influence prescribing behaviour. Reported changes in prescriber behaviour resonate with the non-technical skills (NTS) of prescribing with prescribing behaviour adapting depending on the environment and prescribing conditions. These findings have implications for prescribing education to make it a more contextualised educational process. The model of feedback used in this study could be adopted by other organisations with ward based clinical pharmacy services to improve prescribing practice and patient safety. Declaration of interest The authors report no conflicts of interest. ML undertook this work as part of a PhD. Acknowledgements The authors would like to thank all prescribers for participating in this study and all pharmacists for delivering the feedback intervention. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.sapharm.2017.06.010. References 1. Aronson JK. Balanced prescribing. Br J Clin Pharmacol. 2006;62:629e632. 2. Coombes I, Mitchell C, Stowasser D. Safe medication practice tutorials: a

practical approach to preparing prescribers. Clin Teach. 2007;4:128e134. 3. Dornan T, Ashcroft DM, Heathfield H, et al. An In-depth Investigation into the Causes of Prescribing Errors by Foundation Trainees in Relation to Their Medical Education. EQUIP Study. Manchester: University of Manchester; 2009. 4. Franklin BD, Reynolds M, Shebl NA, Burnett S, Jacklin A. Prescribing errors in hospital inpatients: a three centre study of their prevalence, types and causes. Postgrad Med J. 2011;87:739e745. 5. Seden K, Kirkham J, Kennedy T, et al. Prescribing error in patients admitted to nine hospitals across North west england. BMJ Open. 2013;3:1e14. 6. Ryan C, Ross S, Davey P, et al. Prevalence and causes of prescribing errors: the PRescribing outcomes for trainee doctors engaged in clinical training (PROTECT) study. PLoS One. 2014;9(1):e79802. 7. Ross S, Bond C, Rothnie H, Thomas S, Macleod MJ. What is the scale of prescribing errors committed by junior doctors? A systematic review. Br J Clin Pharmacol. 2009;67:629e640. 8. Aronson JK. Medication errors. Definition and classification. Br J Clin Pharmacol. 2009;67(6):599e604. 9. Heaton A, Webb DJ, Maxwell SR. Undergraduate preparation for prescribing: the views of 2413 UK medical students and recent graduates. Br J Clin Pharmacol. 2008;66:128e134. 10. Nazar H, Nazar M, Rothwell C, Portlock J, Chaytor A, Husband A. Teaching safe prescribing to medical students: perspectives in the UK. Adv Med Educ Pract. 2015;6:279e295. 11. Cantillon P, Sargeant J. Giving feedback in clinical settings. BMJ. 2008;337: a1961. 12. McLellan L, Dornan T, Newton P, et al. Pharmacist-led feedback workshops increase appropriate prescribing of antimicrobials. J Antimicrob Chemother. 2016;71:1415e1425. 13. Lloyd M, Watmough SD, O'Brien SV, Furlong N, Hardy K. A pilot study exploring doctor attitudes and opinions to receiving formalised prescribing error feedback from hospital pharmacists. Br J Hosp Med. 2015;76(12):713e718. 14. Tully MP, Ashcroft DM, Dornan T, Lewis PJ, Taylor D, Wass V. The causes of and factors associated with prescribing errors in hospital inpatients: a systematic review. Drug Saf. 2009;32:819e836. 15. Reason JT. Human Error. Cambridge: Cambridge University Press; 1990. 16. Reason J. Understanding adverse events: human factors. Qual Health Care. 1995;4:80e89. 17. Ramani S, Krackov SK. Twelve tips for giving feedback effectively in the clinical environment. Med Teach. 2012;34:787e791.

Please cite this article in press as: Lloyd M, et al., Exploring the impact of pharmacist-led feedback on prescribing behaviour: A qualitative study, Research in Social and Administrative Pharmacy (2017), http://dx.doi.org/10.1016/j.sapharm.2017.06.010

10

M. Lloyd et al. / Research in Social and Administrative Pharmacy xxx (2017) 1e10

18. Lloyd M, Watmough SD, O'Brien SV, Furlong N, Hardy K. Exploring attitudes and opinions of pharmacists towards delivering prescribing error feedback: a qualitative case study using focus group interviews. Res Soc Adm Pharm. 2016;12(3):461e474. 19. Lloyd M, Watmough SD, O'Brien SV, Furlong N, Hardy K. How to give and receive constructive feedback. Pharm J. 2016;296(7887). http://dx.doi.org/ 10.1211/PJ.2016.20200756. accessed Jan 2017. 20. Flanagan JC. The critical incident technique. Psychol Bull. 1954;51:327e358. 21. Ritchie J, Spencer L. In: Bryman A, Burgess RC, eds. Qualitative Data Analysis for Applied Policy Research. London: Routledge; 1994:173e194. Analyzing Qualitative Data. 22. Ross S, Patey R, Flin R. Is it time for a nontechnical skills approach to prescribing? Br J Clin Pharmacol. 2013;78:681e683. 23. Gordon M, Catchpole K, Baker P. Human factors perspective on the prescribing behavior of recent medical graduates: implications for educators. Adv Med Educ Pract. 2013;4:1e9. 24. Kirkham D, Darbyshire D, Gordon M, Agius S, Baker P. A solid grounding: prescribing skills training. Clin Teach. 2015;12:187e192. 25. Ross S, Ryan C, Duncan EM, et al. Perceived causes of prescribing errors by junior doctors in hospital inpatients: a study from the PROTECT programme. BMJ Qual Saf. 2013;22:97e102. 26. Dearden E, Mellanby E, Cameron H, Harden J. Which non-technical skills do junior doctors require to prescribe safely? A systematic review. Br J Clin Pharmacol. 2015;80(6):1303e1314. 27. McLellan L, Tully MP, Dornan T. How could undergraduate education prepare

new graduates to be safer prescribers? Br J Clin Pharmacol. 2012;74:605e613. 28. Randolph G, Esporas M, Provost L, Massie S, Bundy DG. Model for improvement-part Two: measurement and feedback for quality improvement efforts. Pediatr Clin North Am. 2009;56:779e798. 29. Ferguson J, Keyworth C, Tully MP. ‘If no-one stops me, I'll make the mistake again’: changing prescribing behaviours through feedback; A Perceptual Control Theory perspective’. Res Soc Adm Pharm. 2017;0:1e7. 30. Graber ML. Educational strategies to reduce diagnostic error: can you teach this stuff? Adv Health Sci Educ Theory Pract. 2009;14:63e69. 31. Croskerry P. Cognitive forcing strategies in clinical decision making. Ann Emerg Med. 2003;41:110e120. 32. Teunissen PW, Stapel DA, Van der Vleuten C, Scherpbier A, Boor K, Scheele F. Who wants feedback? An investigation of the variables influencing residents' feedback seeking behavior in relation to night shifts. Acad Med. 2009;84: 910e917. 33. Rothwell C, Burford B, Morrison J, et al. Junior doctors prescribing: enhancing their learning in practice. Br J Clin Pharmacol. 2012;73(2):194e202. 34. Brennan N, Corrigan O, Allard J, et al. The transition from medical student to junior doctor: today's experiences of Tomorrow's Doctors. Med Educ. 2010;44(5):449e458. 35. Lum EMC, Mitchell C, Coombes I. The competent prescriber: 12 core competencies for safe prescribing. Aust Prescr. 2013;36:13e16. 36. Smith SE, Tallentire VR, Cameron HS, Wood SM. Pre-prescribing: a safe way to learn at work? Clin Teach. 2012;9:45e49.

Please cite this article in press as: Lloyd M, et al., Exploring the impact of pharmacist-led feedback on prescribing behaviour: A qualitative study, Research in Social and Administrative Pharmacy (2017), http://dx.doi.org/10.1016/j.sapharm.2017.06.010