[Pharmacy GRIT Article No: 20231368]

Abstract

Background: Within a public hospital chemotherapy day unit (CDU), a paper-based prescription request system is used to request medicines. However, unaccounted for prescriptions are common, with 16% of prescriptions requested not received by the patient or CDU pharmacists, leading to poor treatment outcomes.

Aim: To measure the impact of introducing an electronic prescription tracking tool (EPTT) on the proportion of prescriptions dispensed compared to prescriptions unaccounted for in a CDU, to determine the magnitude and clinical significance of the unaccounted prescriptions, and to assess the perceived user impact.

Method: This study involved three phases: a four-week audit of unaccounted for prescriptions with the paper-based system; a two-week implementation phase of the EPTT; and a four-week audit of unaccounted for prescriptions post-implementation. Proportions of unaccounted for prescriptions were compared between pre- and post- implementation phases and risk assessed against a validated Monash Health tool to evaluate clinical significance. Users in pre- and post-implementation phases were surveyed to ascertain the tool’s effectiveness.

Results: There were 108 unaccounted prescriptions across the 10-week study period with a 9.6% (95% confidence [CI]: 5.9–13.3, p < 0.001) absolute reduction in the proportion of unaccounted prescriptions and an odds ratio of 0.33 (95% CI: 0.21–0.51) following EPTT implementation. There was an absolute reduction of 9.1% (95% CI: 5.5–12.8, p < 0.001) in the number of unaccounted prescriptions potentially causing minor harm. All 24 users indicated improvements in ease-of-use and efficiency with the EPTT.

Conclusion: An EPTT reduced the proportion of unaccounted prescriptions and improved user-perceived workflow when requesting medications in a CDU.

Introduction

Chemotherapy is an important treatment option for patients diagnosed with cancer. Chemotherapy day units (CDU) provide systemic chemotherapy to treat cancer patients in the ambulatory care setting.¹ In a CDU setting, additional prescriptions may be requested, including supportive medicines, oral anti-cancer therapies, or patients’ regular medicines. Supportive medicines are commonly used before and after chemotherapy to prevent and manage treatment related adverse effects.² Examples include antiemetics to prevent chemotherapy induced nausea and vomiting, and colony-stimulating factors to reduce the risk of febrile neutropenia.² The administration of these supportive medicines is time critical, as delayed administration can lead to inadequate control of adverse effects, deterioration in health status, and early withdrawal from treatment.² Current evidence indicates that patient care is improved with timely and appropriate prescribing and supply of supportive medicines to prevent and manage cancer symptoms and treatment related toxicity by reducing treatment-related adverse effects and hospital readmission.^3,4 Similarly, timely supply of oral anti-cancer medicines can improve treatment outcomes by avoiding poor adherence, since patients will be self-administering these medicines at home.⁵

Multiple factors contribute to the delay in patients receiving their prescriptions, including multiple covering doctors in the unit, conflicting priorities, and not having an integrated e-prescribing and dispensing system.⁶ Within the study centre at Moorabbin Hospital, a paper-based system is used to request prescriptions for patients attending their chemotherapy appointment. Local incident data identified gaps in this process stemming from a lack of central system for prescribers, nurses, and pharmacists to track whether requests were completed. Additionally, high patient loads, time pressures, and multiple staff involved has led to a lack of a consistent and coordinated approach to prescription requests. Paper-based request slips can also be easily misplaced, resulting in prescriptions being requested but not prescribed and therefore patients leaving the CDU without their intended medicines. A local, unpublished audit identified up to 16% of prescriptions requested, were not supplied.

Transitioning from a paper-based system to an electronic system such as Electronic Medical Records across Victorian hospitals has shown multiple perceived benefits including improved clinical information access, productivity, and patient safety through a reduction in prescribing errors.⁷ Moving from a paper-based to electronic-based system has also reduced nursing and pharmacist workload and increased time available to spend on patient care activities.^8,9 Benefits of electronic prescription tracking tools have been highlighted in other settings such as for the prescribing of controlled substances¹⁰ and pathology tests¹¹, however the benefits and challenges with implementing an electronic prescription tracking tool (EPTT) in a CDU has not been reported in the literature. Therefore, this study aims to determine the impact of introducing an EPTT on the number and clinical significance of prescriptions unaccounted for in the CDU, as well as assess and compare user opinions on the paper-based request system versus EPTT.

Method

Ethics Statement

This study was reviewed and approved by Monash Health Human and Research Ethics Committee (Reference no: QA/75183/MonH-2021-75183).

Study setting and inclusion criteria

This study was conducted at Monash Health, a multi-site, metropolitan teaching hospital network with over 2000 beds in Melbourne, Australia. Moorabbin Hospital, as part of the Monash Health network, provides specialty surgical and oncology services, including a CDU, which served as the site for this 10-week study from June–August 2020. The CDU operates Monday–Friday, 9am–5pm. It uses a hybrid system involving a combination of paper-based and electronic systems. Local audits indicate there are up to 150 prescriptions written for CDU patients each week.

The current workflow in the CDU at Moorabbin Hospital begins with a pharmacist or nurse requesting a Pharmaceutical Benefits Scheme (PBS) prescription from the doctor, using handwritten slips for supportive medicines, oral anti-cancer therapies, patients’ regular medicines, or other medicines deemed necessary by the pharmacy, nursing, or medical staff. Each prescription request has its own paper request slip, detailing the patient’s name, Unique Reference Number (URN), date of birth, and the prescription required (Appendix 1). Slips were physically given to the prescriber to request a prescription. The prescriber reviews requests and passes the prescriptions to the pharmacist for dispensing. Pharmacists and nurses place prescription requests for patients admitted to the CDU that day or for a future date. All types of prescription requests were included in this study including supportive medicines, oral anti-cancer therapies, patients’ regular medicines, and any other medicines as determined necessary by the pharmacy, nursing, or medical staff. Any requests made retrospectively after the patient was discharged were excluded from this study. Patients receiving chemotherapy outside of the CDU were not included in this study.

Study design

The study consisted of three phases. Phase 1 was a four-week pre-implementation phase with the paper system and development of the EPTT. It involved a pre-implementation audit of the proportion of prescriptions requested using the paper-based system but not received by the pharmacy, and a staff survey to ascertain their feedback on the system. Phase 2 consisted of two weeks of intervention implementation. Phase 3 was a four-week post-implementation phase, involving an audit of the proportion of prescriptions requested using the EPTT but not received by the pharmacy, as well as a staff survey to ascertain feedback on the EPTT (Figure 1).

Figure 1. Flowchart of study design

Phase 1

The initial four-week, pre-implementation phase involved prospective collection of all paper prescription requests to identify the number of prescriptions unaccounted for. This was any prescription request made, but not dispensed, by 5pm on the day of the patient's treatment (Monday–Friday). Paper slips were written throughout the day by CDU pharmacists and nurses, scanned to the study investigators, then provided to the doctor to request a prescription. At 5pm each day of the study period, the study investigators reviewed dispensing records to identify whether a prescription request was dispensed. Paper requests were reviewed at 5pm (Monday–Friday) and were deemed unaccounted for if there was no dispensing record by the end of the day the patient was booked in for treatment. This also accounted for prescriptions requested in advance such as the day prior to treatment (Figure 2a).

Figure 2a. How to identify a prescription unaccounted for in the pre-implementation phase (paper-based system)

A survey was distributed to nurses, pharmacists, and doctors working in the CDU to collect opinions of the paper-based system. The survey was designed to explore users’ perceptions and attitudes towards the current system, its practicality, and to identify opportunities for improvement (Appendix 2). It consisted of five questions with Likert scale responses ranging from strongly disagree to strongly agree. Open-ended questions were included for users to indicate what worked well or could be improved with the paper system. All users were invited via email to respond to an anonymous Google Forms (Alphabet Inc, Mountain View, CA, USA) survey, which was open for a period of two weeks, with reminders emailed three days prior to the closing date.

The EPTT was developed to create a common platform for CDU staff to communicate and track prescription requests (Appendix 3). The EPTT was a shared Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) spreadsheet accessible via an internal drive between pharmacy, medical, and nursing staff. It required three patient identifiers: expected treatment date, medicine, and dose required, as well as the name of the person making the request and a free text column for additional notes. A tick-box column was included for the doctor to indicate whether a prescription was completed. The spreadsheet’s initial design was based on the existing paper-based forms in consultation with CDU pharmacists, nurses, and doctors.

Phase 2

Multiple education sessions on how to use the EPTT were conducted one week prior to implementation to capture all potential users. The EPTT was then introduced and refined in response to feedback from users over a two-week period. During this time, additional columns were added for the pharmacists to indicate whether a prescription had been received, and to remind the doctor to hand the prescription to the pharmacist. The EPTT was reliant on manual data entry by the user as well as manually saving the file.

Phase 3

A four-week post-implementation phase commenced with data collection of prescriptions unaccounted while using the EPTT. Electronic prescription requests from EPTT were collected at 5pm (Monday–Friday) and were deemed unaccounted for if there was no dispensing record by the end of the day the patient was booked in for treatment to account for prescriptions requested in advance (Figure 2b). The same users were sent a survey during the post-implementation phase. This survey repeated the same questions as the pre-implementation survey, however it also asked users to document their perceptions of the EPTT.  

Figure 2b. How to identify a prescription unaccounted for in the post-implementation phase (electronic system)

The potential clinical significance and risk of adverse outcomes to patients in the event of an unaccounted prescription were independently assessed against a risk matrix by two pharmacist investigators (JN, LC) (Appendix 4).¹⁰ This method of risk classification was consistent with Monash Health’s local procedure as well as The Society of Hospital Pharmacists Australia’s Standard of Practice for Clinical Pharmacy.^12,13 The risk rating was determined by the severity of consequences if an intervention was not made, and the likelihood of the scenario occurring without intervention with four ratings: low, moderate, high, and extreme.¹⁴ Disagreements were resolved with a third investigator (BS).

Statistical analysis

The proportions of unaccounted prescriptions were compared and used to calculate the absolute difference between the pre- and post-implementation phases. Data analysis was performed using Stata statistical software (version 17, StataCorp LLC, College Station, TX, USA). The proportion of prescriptions unaccounted for pre- and post-implementation was compared using a chi-square test. An odds ratio (OR) and 95% confidence interval (CI) were calculated to evaluate the odds of the EPTT reducing the numbers of unaccounted prescriptions. A p-value of <0.05 was defined as statistically significant a priori. The risk rating classification of unaccounted prescriptions was described using descriptive statistics.

Staff perceptions survey

Survey data on staff perception were analysed by graphing the responses to each question and comparing the proportion of responses to each question between the pre- and post-implementation phases. A thematic analysis was conducted to identify common themes amongst responses to the open-ended questions. Comments were collated via the grounded theory method, followed by coding stages. Codes were highlighted in the comments and grouped into themes by two pharmacist investigators (JN, LC).

Results

Of the 488 prescriptions requested during the pre-implementation phase, 74 (15.1%) were identified as unaccounted for. Of the 611 prescriptions requested during the post-implementation phase, 34 (5.6%) were identified as unaccounted for. There was a 9.6% (95% CI: 5.9–13.3, p < 0.001) absolute reduction in the proportion of prescriptions unaccounted for between the paper and electronic system. Unaccounted prescriptions were less likely following EPTT implementation with an OR of 0.33 (95% CI: 0.21–0.51), indicating the use of the EPTT decreased the risk of an unaccounted prescription.

The proportion of low-risk unaccounted prescriptions in the pre- and post-implementation phases were 14.5% (n = 71; 488 prescriptions) and 5.4% (n = 33; 611 prescriptions) respectively with an absolute difference of 9.1% (95% CI: 5.5–12.8, p < 0.001). The proportion of moderate risk unaccounted prescriptions in the pre- and post-implementation phases were 0.6% (n = 3) and 0.2% (n = 1) respectively with an absolute difference of 0.45% (95% CI: -0.31–1.2, p = 0.33). There were no prescriptions classified under high risk causing catastrophic harm to patients in the pre- and post-implementation phases. See Appendix 5 for further details on risk matrix analysis categorised by medication groups under each risk category.

Survey results

There were 24 users invited to participate in the pre- and post-implementation survey, which had a 100% (n = 24) response rate. These included three (12.5%) doctors, five (20.8%) pharmacists and 16 (66.7%) nurses. In total, 58.4% of users either agreed or strongly agreed that the paper-based system was easy to use (Figure 3). This response increased to 83.3% when surveyed about the EPTT. Additionally, 91.7% of users agreed or strongly agreed that the EPTT was efficient, as compared to 45.9% for the paper-based system. Furthermore, 70.2% of users agreed or strongly agreed that it was easy to find out whether a prescription request was made with the EPTT, as compared to 36.8% of users for the paper-based system.

Figure 3: User survey responses on the paper-based system and electronic system

Themes identified from open-ended questions were similar between doctors, nurses, and pharmacists. It indicated that the EPTT had improved ‘workflow efficiency’, was ‘reliable’ and ‘easy to use’, however ‘technical difficulties’ were sometimes a barrier to the efficient functioning of the electronic tool.

Discussion

The implementation of an EPTT led to a 9.6% absolute reduction in prescriptions unaccounted for where unaccounted prescriptions were less likely to occur compared to the paper-based system (OR 0.33; 95% CI: 0.21–0.51). A lower number of prescriptions unaccounted for with the EPTT suggests patients were more likely to have their required medicines prescribed and dispensed. The difference in numbers of unaccounted prescriptions between pre- and post-implementation phases may be a result of improved data recording with the EPTT. While the EPTT could be saved and data retained, paper-based slips may be easily lost after written, or not provided to the investigator.

When paper systems are replaced with an electronic alternative, the electronic system is known to provide a more streamlined and straightforward approach to workflow,¹⁴ assist with prescribing accuracy,^15,16 and reduce medication errors.¹⁷ In an Australian study,¹¹ an electronic system replaced a paper request system for the ordering of pathology results. After implementation of the electronic workflow, there was a significant reduction in the time a laboratory test result was made available, as well as a reduction in incomplete requests. These results are consistent with the results of this study, where the implementation of a EPTT resulted in a reduction in the proportion of unaccounted prescriptions.

The majority of prescriptions unaccounted for were classified in the low-risk category with an absolute difference of 9.1% following the implementation of EPTT. Although this is statistically significant, this could simply reflect a statistically significant reduction in the overall number of unaccounted for prescriptions. This reduction in the number of low-risk unaccounted for prescriptions overall means a reduced chance of harm to patients, as receiving the intended medicines may improve patient adherence⁵ and reduce chemotherapy related adverse effects.²

A limited number of prescriptions unaccounted for were classified in the moderate-risk category with an absolute difference of 0.45%, not reaching statistical significance. The small number of moderate- and high-risk unaccounted for prescriptions detected may be due to staff actively resolving unaccounted prescriptions in the interests of patient safety. Staff may be more vigilant regarding moderate- and high-risk medicines and escalate requests to ensure patients are not discharged without important medicines. This may then result in a higher proportion of low-risk unaccounted for prescriptions detected. Although this data does not demonstrate statistical significance, a reduction in the number of moderate-risk unaccounted for prescriptions may still be of clinical significance in improving patient care.

In previous studies, non-adherence was common¹⁸ when prescriptions were not dispensed due to barriers at the prescribing level.¹⁹ The EPTT reduced the number of unaccounted prescriptions, leading to an improvement in availability of collection-ready medicines for patients. A larger sample size and longer study period may be required to fully evaluate the clinical significance of the EPTT.

Survey results suggested the EPTT was the preferred tool by users compared to the paper request slips. The limitation of having no centralised tracking system with the paper-based system improved with the use of the EPTT, where more users agreed or strongly agreed it was easy to find out whether a prescription request had been made (70.2% of users of the EPTT vs 36.8% of paper-based system users). With the EPTT, more users agreed or strongly agreed efficiency had improved (91.7% of users of the EPTT vs 45.9% of paper-based system users), suggesting a reduction in delays in prescription requests being submitted and written by doctors. This is consistent with an international study which has shown that an electronic system was preferred by users over paper-based records as it was found to be more effective and efficient.²⁰ Survey responses indicated some users felt ‘technical difficulties’ posed a challenge with the EPTT. Future improvements in the EPTT as suggested by survey comments involved having automatic updates rather than relying on users manually refreshing the page, EMR integration, and preventing the tool from locking when accessed by multiple users.

More prescription requests were collected during the post-implementation phase compared to the pre-implementation phase as it is less likely data points will be lost in the electronic system. Lost data points during prospective data collection of paper request slips may be due to many staff and pieces of paper involved in the process. This may mean the pre-implementation data is not a true reflection of the total number of prescriptions unaccounted for. However, the study was able to demonstrate statistical significance with the existing data, suggesting that if a larger sample size was used during the pre-implementation phase, the results of this study would be further strengthened.

The EPTT may be extended to other units across Monash Health including Monash Medical Centre and other tertiary institutions with a CDU to reduce the risk of the requested medicines not being dispensed. A EPTT may be potentially incorporated into Monash Health’s EMR Program once the chemotherapy workflow is designed.

Study Limitations

A possible limitation of the study includes the consistency of staff working in the CDU as differences in staff knowledge and experience may affect the results of the study. An experienced CDU pharmacist may quickly identify that a prescription has not been written and follow up with the prescriber prior to 5pm whereas an inexperienced pharmacist may not come to this conclusion as efficiently. The duration of the study was based on medical residents rotating through CDU to promote consistency and familiarity with the processes pre- and post-implementation. However, the restricted study period has limited the sample size. It is unclear whether the impact of the EPTT will be maintained across multiple staff rotations and a future study with a longer duration may assess this. The risk rating assumes that if a prescription is not dispensed, the patient does not receive the medicine. There may be other reasons why a prescription was not dispensed, however this investigation was out of the scope of this study. Potential reasons may include the prescription being deemed inappropriate to prescribe by the doctor, being deemed inappropriate to dispense by the pharmacist, or the patient declining supply.

In conclusion, the implementation of an EPTT in a CDU resulted in a reduction in the number of prescriptions unaccounted for, and an overall improvement in perceived workflow, ease-of-use, and efficiency for staff.

Acknowledgements

The study investigators would like to acknowledge Samanta Wood, Louise Lord, Brindha Garuda, and Marianne Jovanovic from the Monash Health Pharmacy Education Services team for the review of this manuscript and Eldho Paul from Monash University for statistical support.

Appendices

Appendices 1–4

Appendix 5

References

1. Department of Health & Human Services. Guidelines for timely initiation of chemotherapy: a proposed framework for access to medical oncology and haematology cancer clinics and chemotherapy services in Victoria. Melbourne: State of Victoria; 2015. Available from https://www.health.vic.gov.au/sites/default/files/migrated/files/collections/policies-and-guidelines/c/chemotheraphy-guidelines-pdf.pdf.  
2. EviQ. Prevention of anti-cancer therapy induced nausea and vomiting (AINV). Sydney: Government of New South Wales; 2021. Available from https://www.eviq.org.au/clinical-resources/side-effect-and-toxicity-management/gastrointestinal/7-prevention-of-anti-cancer-therapy-induced-nausea#management>.
3. Hin CM, Hong CC. Oncology pharmacist’s role and impact on the multidisciplinary patient-centre practice of oncology clinic in public hospitals. Asia Pac J Health Manag 2019; 14(1): 16.
4. Sisay AE, Engldawork E, Yesuf TA, Ketema EB. Drug related problems in chemotherapy of cancer patients. J Cancer Sci Ther 2015; 7: 2.
5. Paolella GA., Boyd AD, Wirth SM, Cuellar S, Venepalli NK, Crawford SY. Adherence to oral anticancer medications: evolving interprofessional roles and pharmacist workforce considerations. Pharmacy (Basel) 2018; 6: 23.
6. Green CF, Hunter L, Jones L, Morris K. The TTO journey: how much of it is actually in pharmacy? Pharmacy Management 2015; 31(4): 16–20.
7. Department of Health & Human Services. Compendium of electronic medical record (EMR) benefit measures: digital health. Melbourne: State of Victoria; 2020. Available from https://www.health.vic.gov.au/sites/default/files/migrated/files/collections/research-and-reports/e/compendium-of-emr-benefit-measures-aug-2020.pdf  
8. Shao S-C, Chan Y-Y, Lin S-J, Li C-Y, Yang Y-H K, Chen Y-H, et al. Workload of pharmacists and the performance of pharmacy services. PLOS ONE 2020; 15: e0231482.
9. Seto R, Inoue T, Tsumura H. Clinical documentation improvement for outpatients by implementing electronic medical records. Stud Health Technol Inform 2014; 201: 102–7.
10. Freeman PR, Curran GM, Drummond KL, Martin BC, Teeter BS, Bradley K, et al. Utilization of prescription drug monitoring programs for prescribing and dispensing decisions: results from a multi-site qualitative study. Res Social Adm Pharm 2019; 15: 754–60.
11. Georgiou A, Westbrook J, Braithwaite J, Iedema R, Ray S, Forsyth R, Dimos A, et al. When requests become orders -- a formative investigation into the impact of a computerized physician order entry system on a pathology laboratory service. Int J Med Inform 2007; 76: 583–91.
12. Society of Hospital Pharmacists of Australia. Chapter 13: documenting clinical activities. J Pharm Prac Res 2013; 43 (S2): S42–S46.
13. Murray SL, Grantham K, Damle SB. Development of a generic risk matrix to manage project risks. Int J Ind Sys Eng 2011; 5: 36–41.
14. Green SD, Thomas JD. Interdisciplinary collaboration and the electronic medical record. Pediatr Nurs 2008; 34: 225–7.
15. Williams F, Boren SA. The role of the electronic medical record (EMR) in care delivery development in developing countries: a systematic review. Inform Prim Care 2008; 16: 139–45.
16. Barnsteiner JH. Medication reconciliation: transfer of medication information across settings – keeping it free from error. Am J Nurs 2005; 105 (3 Suppl): 31–6.
17. Sethuraman U, Kannikeswaran N, Murray KP, Zidan MA, Chamberlain JM. Prescription errors before and after introduction of electronic medication alert system in paediatric emergency department. Acad Emerg Med 2015; 22: 714–9.
18. Ekedahl A, Månsson N. Unclaimed prescriptions after automatic prescription transmittals to pharmacies. Pharm World Sci 2004; 26: 26–31.
19. Gadkari AS, McHorney CA. Medication nonfulfillment rates and reasons: narrative systematic review. Curr Med Res Opin 2010; 26: 683–705.
20. Msukwa MKB. User perceptions on electronic medical record system (EMR) in Malawi [Master of Public Health thesis]. Malawi: University of Malawi; 2011.