Enhancing the safety of GP trainee prescribing: a literature review and evaluation of an educational intervention aiming to reduce errors and suboptimal prescribing
Publication Date
December 11, 2024
Creator
Abstract
Introduction
Medication errors causes significant morbidity and mortality in healthcare systems throughout the world. In England, approximately 5% of prescriptions issued in primary care contain an error, with around one in 550 prescriptions containing a serious error. GP trainees have been highlighted as a particular cohort of prescribers who may benefit from additional support and training to enhance their prescribing.
A literature review was conducted to ascertain details of interventions which had been undertaken with GP trainees in the primary care setting, in order to reduce medication error and improve prescribing safety. More than 3000 articles were found through database searches, with fifteen articles forming the review cohort after exclusion processes. Most of the interventions reviewed were small-scale and limited in scope, with many focusing on exploring the impact of a specific teaching session or clinical experience. More meaningful changes in practice were apparent with a more comprehensive approach to prescriber education, or the delivery of personalised feedback. It was clear from the literature review that there is a need for further studies exploring prescribing safety interventions for GP trainees.
The REVISiT intervention was devised (a retrospective review of prescriptions issued by GP Trainees) and was evaluated as part of this thesis. The aim of the evaluation was to work towards identifying whether a pharmacist-led review of a series of prescription items issued by GP trainees during or nearing their final year of training is an effective means of improving prescribing safety for this cohort of prescribers.
Methods
Ten GP trainees and their trainers were recruited from the Nottingham GP Vocational Training Scheme. Each of the trainees had 100 sequential prescription items reviewed by a clinical pharmacist to identify prescribing errors, suboptimal prescribing and aspects of good practice. Prescribing data were pooled, and underwent descriptive statistical analysis to report the over-all medication error and suboptimal prescribing rate of the study participants. Prescribing data were reviewed to look for patterns of errors and suboptimal prescribing.
A detailed report was prepared for each trainee, outlining the 100 prescription items that had been reviewed. GP trainees and their trainers received the individualised report as part of a feed-back meeting, led by a clinical member of the research team. Both parties were then invited to take part in semi-structured interviews to gain their feedback regarding the review process. Further semi-structured interviews were undertaken one to three months after the first interview, to ascertain if there had been any changes in practice as a result of taking part in the study. All interviews were recorded, transcribed, and analysed for emergent themes.
Results
Ten GP trainees each had approximately 100 prescription items reviewed to highlight prescribing errors, and episodes of suboptimal prescribing. A total of 1028 prescription items were reviewed, which had been issued over the course of 1291 consultations. Ninety-two prescribing errors were reported, which represented an error rate of 8.9%. A total of 360 episodes of suboptimal prescribing were identified, representing a suboptimal prescribing rate of 35%.
The most common error was that relating to the dose or strength of a medication prescribed (n=30, 32.6% of all prescribing errors). Examples of this error included prescribing an inappropriate dose of antibiotics for the indication or age of the patient, or prescribing a dose of medication that was not in keeping with the narrative within a consultation. The second most common error type observed related to providing incomplete or inadequate information on the prescription (n=24; 26.1% of all errors). Examples of this error type included the provision of non-specific dosage instructions for potent steroid creams, and antibiotic eye drops where there could be confusion regarding their use.
Participants reported a range of reasons for wanting to engage with the study, including a perceived deficit in prescribing education, and a desire to receive personalised feedback on their practice. On receiving feedback, all trainees made plans for some practice changes, ranging from acting upon knowledge deficits at the level of an individual drug or drug class, to making a commitment to more diligent use of local guidance. In the follow-up interview, some participants were able to report actual changes they had made to their prescribing practice, such as amending the length of antibiotic courses to reflect local guidance. One trainer was able to report how they had changed the way in which they were supervising a new trainee, in order to enhance prescribing safety.
Discussion
Even though the study cohort was small and self-selected, the reported error rate of 8.9% was similar to other reported error rates associated with junior doctor prescribing in different settings. The types of errors were also similar to those reported in the literature.
The way in which feedback was delivered in the REVISiT intervention was largely in keeping with best practice in postgraduate education, and therefore facilitated appropriate behaviour change. which were sometimes able to be corroborated by trainer report. Participants were, however, concerned that some of the suboptimal prescribing that had been reported to them was overly ‘picky’, and not be in keeping with ‘real world practice.’ One of the limitations of the study was that actual practice change was not able to be discerned: only reports of practice change.
Exploring the rationale behind error occurrence revealed the importance of adopting primary-care-specific prescribing knowledge, as well as appropriate consultation management techniques. Relating the findings of the study to error theory permitted exploration of the utility of the intervention beyond merely providing corrective feedback.
In common with other studies involving junior doctors, participants from the REVISiT study commented on the inadequacy of prior education in prescribing for preparing them to practice as doctors. Other systems-related issues that were highlighted included the effect of time-pressures on the quality of note keeping and prescribing, and frustrations with the way in which computer systems outworked the prescribing process.
Conclusion
Prescribing medication is a fundamental part of the care delivered to patients in primary care. Medication error continues to be responsible for significant unnecessary morbidity and mortality. Educational interventions have been shown to enhance prescribing safety in hospital and outpatient settings, but there remains a paucity of robust studies, especially in the domain of primary care, and specifically involving GP trainees.
Through the REVISiT intervention, 10 GP trainees received personalised feedback on a series of 100 prescription items. The process received positive feedback from the trainees and their trainers, and changes to practice were reported.
Although a natural follow-on from this study may be to explore the validity of the intervention through a larger scale, randomised controlled trial, this may not actually be a practical next step due to the pharmacist-resource that would have to be utilised. Findings from the REVISiT study have, however, already been used to enhance the teaching of safer prescribing, and have also formed the basis for development of a national prescribing assessment for GP trainees, which has recently been approved by the General Medical Council.
Further evaluation of the national prescribing assessment, and indeed exploring a robust prescribing evaluation for other clinicians in the primary care setting would be a useful next step with the lessons learned from the REVISiT study.
Item Type
ethesis
Thesis Type
DM
Supervisors
Subjects (LC)
Associated Schools / Departments
School of Medicine
eprints ID
77452
UoN Repository URI
Except where otherwise noted, this item's license is described as
File(s)![Thumbnail Image]()
Name
Richard Knox - 14217032 - thesis with corrections.pdf
Type
Full-text
Description
Examined. Thesis with minor corrections
Size
4.86 MB
Format
Adobe PDF
Checksum (MD5)
cc91c2d5823878a70f5c96d92f981f55