Pharmacist-led prescription writing educational intervention to final-year medical students: A pre-post non-randomised longitudinal study

Study design

This non-randomised pre-post longitudinal study was conducted in two phases (Figure 2). The intervention was pharmacist-led multimodal education on prescription writing skills for final-year medical students. This is a non-randomised study with no control group based on ethics, fairness, and logistics of conducting the intervention over vast geographical locations to ensure that information sharing did not contaminate the collected data.

Figure 2. Design structure of two-phase non-randomised pre-post longitudinal review study.

Phase 2A

Baseline data collected in 2023 assisted in the content validity of this two-phase study, informing phase 2B. This phase acted as the comparator to phase 2B pre-test, ensuring that students’ prescribing skills are comparable at similar time points, both between calendar years and prior to additional educational interventions. Due to study logistics with the participating university, ethical approvals, and accounting for assessment burden on participants, this cohort did not act as a control group and were not assessed 5 weeks post standard placement (with no additional education interventions). A pilot study had previously published the effects with a control group.¹¹

Phase 2B

This interventional phase formed the pre-post design of this study in 2024. The principal investigator is a pharmacist who developed and led the multimodal educational intervention conducted after the pre-test. Participants attended their standard 5-week placement block prior to completing the post-test.

Phase 3

Participants who acknowledged their participation in phase 3 voluntarily provided a contact email after completing phase 2B. The follow-up test in 2025 resulted in the longitudinal design of this study.

Study setting

This study is developed and conducted in Queensland (QLD), Australia, through the Medical School Program of the JCU College of Medicine and Dentistry. Bachelor of Medicine, Bachelor of Surgery at JCU is a 6-year undergraduate course. Like other Australian medical schools, it undergoes periodic review and accreditation by the Australian Medical Council. The course includes face-to-face and case-based learning modalities with mandatory clinical placements.

Placement blocks vary in duration and expose students to a variety of specialties. Not all students complete identical placement rotations at the same time. The JCU curriculum, like many other curricula, focuses more on therapeutic choice (for example, choice of drug, dose and route) rather than the technical skill to writing a medication prescription, which is the focus of this study. In 5^th year, students receive an introduction to prescribing as per the Pharmaceutical Benefits Scheme (PBS). Students are not assessed on this information and have limited opportunity to practice writing a prescription. Any other learnings on prescription writing are opportunistic or self-directed. Anecdotally, medical students have written medication prescriptions and had an authorised prescriber countersign, but this process is not legal. This study provided a protected environment for participants to practice writing a medication prescription, which is an essential skill that does not have mandated assessment in Australia.

Participants

As the recruited medical school, convenience sampling was used to invite all JCU final-year medical students based in Townsville, Cairns, Mackay, rural and remote locations, and on their elective rotation. JCU medical students not in their final year of medical education or from other universities were excluded from this study. This study also excluded junior doctors from 2025 who were not actively involved in phase 2B during 2024. All participants were required to provide written informed consent in each phase by completing a mandatory field when commencing the test and accepting to participate.

Phase 2A of this study was conducted in 2023 with JCU final-year medical students. This was the comparator to phase 2B pre-test. Participation was voluntary and anonymous. This cohort was not included in the interventional study.

There were 169 year-6 (final-year) medical students enrolled at JCU in each year of 2023 and 2024 and invited to participate. Our randomised controlled trial pilot study produced a difference in the total mean score of eight points between groups.¹¹ However, the pilot study involved a very small sample size, so the lower bound of the 95% confidence interval for the mean difference was substantially less (approximately 3); therefore, a conservative approach requires a sample size large enough to allow a difference as small as 3 to be detected. This was based on a two-tailed test with an alpha of 5% and power of 80%. This difference in prescribing scores is clinically meaningful and relevant to practice based on the literature.^11,12 A 25% dropout rate was factored into the sample size calculation. Using G*Power version 3.1.9.4, this study required 76 participants and therefore a total of 95 students was determined to be an ideal sample size.

We previously identified final-year medical student’s preference for prescription writing education is during the final year of medical education.² Phase 2B included the JCU final-year medical student cohort of 2024 from July to August. The consenting Phase 2B cohort formed the participant group for phase 3 in 2025 as junior doctors (interns). In the Australian context, the intern year is the first postgraduate year (PGY1) after completion of medical school and is completed in accredited hospitals and health services in accredited programs with provisional registration from the national clinician registration body (Australian Health Practitioner Regulation Agency - AHPRA). Progression to general registration as a medical officer is conditional upon the successful completion of all the required rotations in an accredited intern program.

Voluntary participation in phase 2B and phase 3 was not anonymous to ensure that data could be paired to assist with the analysis. Data from phase 2B and phase 3 was coded, analysed, and presented in de-identified formats.

Intervention

Intervention and assessment criteria are specific to local state legal regulation of the Medicines and Poisons (Medicines) Regulation (MPR) 2021. Some participants may have been on placement in other locations external to QLD at the time. However, as their medical degree is based through JCU QLD, all educational materials used in the additional education modalities and assessable criteria adhered to QLD regulations.

Additional education modalities

Mokrzecki et al. phase-1 study identified that final-year medical students’ preference is a multimodal education package on the skill of prescription writing delivered by a pharmacist during standard placement blocks with hands-on experiences (workshops) and tutorials.² These preferences assisted in developing the intervention methods presented. Final-year medical students from 2023 (Phase 2A) completed only the online prescription writing test to ensure content validity during the intervention phase (phase 2B). Phase 2A testing was conducted in a whole cohort session, where participants were then offered the face-to-face education package to meet ethics criteria.

This study aimed to cover all additional education modalities by providing face-to-face education with case-based learning and hands-on experiences (roughly 1.5 h), supported by an online learning module (roughly 30 min), implemented during a standard five-week placement block (Figure 2).

Education was provided by a pharmacist regarding the skill of writing a prescription, separate from the clinical decision-making process within the medication management pathway. Education materials were supported by resources such as the MPR, PBS, and the Australian Medicines Handbook (AMH). To assess the content and remove bias from having one pharmacist educator, this intervention recruited two other pharmacists to deliver the face-to-face intervention to the participants in Cairns and Online. All trainers underwent training by SM to ensure consistency in education delivery.

The effects of this multimodal education package on prescribing skills were determined by assessing participants pre-post the intervention, and again as junior doctors (intern year).

Online learning module

The online education is an interactive learning module developed through Articulate^® by SM and launched online on LearnJCU^® (digital learning environment utilised by JCU) after students completed the pre-test. It is an overview of the face-to-face education that participants were encouraged to complete. Participants could engage with the content as many times as they wished during the five-week placement period. Access to the content closed immediately before completing the post-test.

Face-to-face education

Face-to-face education consisted of an interactive hands-on teaching model (workshop format), first reviewing the legal requirements of a prescription from the Queensland MPR (tutorial format). This guided case-based learning into training functionalities of the PBS and AMH, and how using these resources assist in writing medication prescriptions.

Hands-on opportunities

During face-to-face education, cases were used to assist participants in learning and practicing how to write various medication prescriptions. In Australia, the government and the scheduling of medicines restrict the use of medications and poisons. These restrictions determine the way a prescription needs to be written and extra legal and safety criteria that the prescriber must document. Medicines and poisons are classified under schedules (S2-10) as per the Standard for the Uniform Scheduling of Medicines and Poisons (SUSMP). For the purposes of this intervention, participants were mainly taught about S4 (Prescription Only), and S8 (Controlled Drugs) scheduling as this encompassed the general factors required when writing a prescription. Education was further extrapolated by teaching differences in PBS and non-PBS medications.

Participants also engaged with the interactive features of the online learning module to complete S4 and Monitored medicine/S8 medication prescriptions. All prescription writing cases used during the educational interventions and the pre-post and follow-up tests differed to reduce the potential of information sharing.

Assessment and data collection methods

To our knowledge, no assessment tool has been developed that has the functionality to assess the skill of writing a complete discharge or outpatient medication prescription separate from making a clinical decision or inpatient medication ordering. The assessment tool developed for this study was conceptualised by the principal investigator (SM), based on the MPR requirements, and approved by all authors (Figure 3 and Table 1). A pilot study by Mokrzecki et al.¹¹ and some of the preliminary questions of Mokrzecki et al.² assisted in the development of all test questions in each phase. Online prescription writing assessment consisted of varied preliminary questions (demographic and multiple-choice knowledge-based questions) and five cases yielding six prescriptions (Figure 3).

Figure 3. Copy of phase 2A written case information showing an example of similar Schedule four prescription cases in each phase.

Four other cases (complex S4, S4 streamline, S8 and mixed), are not displayed.

Qualtrics^® is a web-based software used by SM to roll out the online test, allowing participants to type and submit a complete medication prescription online for the case questions without prompting through an online formulary (Figure 4). Anecdotal information from phase 2A provided insight into the ability of participants to use the ‘copy and paste’ functionality of the case information into the answer boxes. As we were assessing participants knowledge on what the required components are of a prescription, to mitigate this, the case information was added as a picture into the test for phases 2B onwards.

Figure 4. Copy of Qualtrics formatting allowing students to free-type the answer to the prescribing case.

All tests had different cases used; however, each test consisted of Schedule 4 (S4), complex S4, S4 streamline, Schedule 8 (S8), and mixed (S4+S8) cases.

Each prescription was critiqued according to set criteria as per the legal MPR and PBS, which must be adhered to when prescribing in QLD. Within the 6-5-4-3 criterion, four overarching components were established (DPSP – six Drug factors, five Prescriber factors, four Safety factors, and three Patient factors), which are listed in Table 1. To our knowledge, this was the first study to develop an assessment rubric or framework to teach and critique discharge/outpatient prescriptions. This novel assessment framework did not require validation, as data collection was based on local legal regulations and guidelines of assessment rather than the establishment of a research tool. The model was also presented in this unique format in an aim to simplify teaching and learning for the participants.

After writing each prescription, participants were asked if they were confident that their prescription was legal and safe by stating ‘confident’ or ‘not confident’ (Figure 4). This self-perceived confidence rating was then re-termed “appropriately confident (or not)” based on the dispensability of the prescription(s). The dispensability of a prescription became a categorical variable and was added as another layer of assessment. This is because, numerically a prescription could have been considered a “pass”, however, if it did not contain the legal and safety requirements as per DPSP (and MPR), then a pharmacist would not be able to dispense the prescription(s) or supply the medication(s).

Qualtrics data was imported into Microsoft Excel and an appropriate statistical software system for analysis. Formatting of data coding and dispensability criteria as per the MPR were established by SM and SP. De-identified data was coded by SM, and SP independently completed random verification by cross-checking the coded data with the original de-identified data. This method was to ensure accuracy and consistency in marking. If discrepancies were identified, SM and SP together verified coded data with the original data and corrected based on the pre-agreed criteria. Where an agreement could not be reached, the remaining authors were consulted and an outcome agreed upon. SM and SP were not blinded to the data timepoint as data was collected at specific times and coding began immediately due to time constraints.

Measuring effectiveness - comparisons

Three comparator points were used for data collection. The first was the year 2023 phase 2A cohort to the 2024 phase 2B cohort pre-test data. This comparison determined if participants are comparable from 2023 to 2024 with similar baseline knowledge between years. This established the level of current education practices and the impact it is having on technical prescribing skills. The second comparator group was the 2024 phase 2B cohort pre-test to phase 2B participants’ post-test data. This tested the outcome of possible changes in prescribing knowledge and skills after the intervention.

The analysis of knowledge retention involved a comparison of phase 2B data to phase 3 data. The three time points (pre-post-longitudinal, see Figure 2) assisted in the analysis and possible justification of additional education modalities delivered during the medical degree.

Proposed data analysis

Identifiable data for phase 2B and phase 3 were collected to perform paired data analysis after reidentification. All data analyses will be presented and published in a de-identified manner.

The following datasets will be compared and analysed, where if chosen to present as a P value, <0.05 is considered statistically significant in all cases:

A new numerical variable may be created in calculating the difference in scores (between and within pre-post phase 2B and phase 3). Therefore, this may also be tested to determine whether the results are normally distributed. Further analysis will be undertaken to seek specific insights into the relative confidence in prescribing abilities in relation to dispensability.

Ethics approval

This study was approved by the JCU Human Research Ethics Committee on June 15^th, 2023 (approval number H9114). JCU Medical School provided written support for the principal investigator to conduct this study. Medical Deans Australia and New Zealand were aware of this study and provided written support in conducting phase-1 (information not provided in this paper). No further site-specific approvals were required as each study location used JCU clinical medical school facilities and resources. Research carried out for this study adhered to the Declaration of Helsinki.

Dissemination

As the principal investigator, SM completed this study for the purposes of her Doctor of Philosophy (PhD). SM is undertaking this as a PhD by publication; therefore, all results will be disseminated by publication and presentations in addition to the final thesis. The Townsville Hospital and Health Service (THHS) Study, Education, Research Trust Account (SERTA) grants assisted in providing SM funded time to complete this study and her thesis. Data sharing is not yet applicable, as confidential datasets have been generated and analysis is still being completed during the publication of this current paper (excluding phase-1, not presented).² Datasets that are generated and/or analysed will not be publicly available because of confidentiality but may be available from the corresponding author upon reasonable request.

Study status

Data collection and coding for all phases has been completed and this study has concluded. Complex data analysis is currently ongoing, and the relative results will be presented in future papers and submitted for publication when appropriate.

Consent

Participants partaking in this voluntary study provided informed consent. Additionally, before beginning each test, participants were required to provide consent to participate by completing a mandatory acknowledgment field. This process served as a form of written consent.

Participants were advised that de-identified data would be used in research publications and reports as outlined to them prior to consenting to start the voluntary test and interventions.