Aim

1. Assess the incidence, type and severity of prescribing errors

2. Assess the use and efficacy of a pre–existing prescribing aid

3. Reduce the number and severity profile of prescribing errors through increased use of the prescribing aid.

Method Twenty cystic fibrosis (CF) patients, admitted consecutively to a tertiary paediatric centre, were included in the study. Use of the pre-existing CF prescribing aid and prescribing errors were detected prospectively by daily pharmacist review of electronic drug charts. A doctor and a pharmacist then retrospectively assigned a severity score (A-I), using a standardised tool.¹ Ten patients were evaluated initially. Following this, paediatric CF patients were tagged on the electronic prescribing system, creating a pop-up alert to direct prescribers to the CF prescribing aid when those patients’ electronic medicine records were opened.

Results Ten patients (163 medication orders) were evaluated in phase I pre-intervention, and ten patients (157 medication orders) were evaluated in phase II post-intervention.

In total 127 prescribing errors were recorded. The most common types of prescribing error were: failing to prescribe a patient’s regular medicine (27%); prescribing the wrong formulation (19%); prescribing the wrong dose (14%).

No errors, in either phase of the study, caused patient harm. The highest severity score in this study was awarded to prescribing errors that reached the patient, did not cause harm but required intervention to preclude harm or required extra monitoring (category D).¹

Throughout both phases, 32% of error containing orders were related to medicines that were not included in the CF prescribing aid, the error rate was comparable in both phases (30%; 35%). These medicines were non-CF specific drugs therefore not appropriate to be included in the prescribing aid. These medication orders were excluded from further pre- and post- intervention analysis.

In the pre-intervention group, 100 CF medication orders were prescribed. The prescribing aid was used in 42% of orders. 43% of medication orders had at least one prescribing error. 27% of errors were of severity category D.

In the post-intervention group, 104 CF medication orders were prescribed. The prescribing aid was used in 70% of orders. 27% of medication orders had at least one prescribing error. 15% of errors were of severity category D.

Conclusion Introducing a paediatric CF alert system successfully increased prescriber utilisation of the CF prescribing aid, with subsequent reduction in error rate and severity of CF specific medicines. We are confident that this improvement reflects our intervention, although without in-depth statistical analysis we cannot attribute causality. However, errors still occurred despite the use of the prescribing aid and errors occurred in non-CF specific medicines. A continuous quality improvement approach is being adopted to explore alternative causes for error and further interventions that can be made, with focus on dovetailing inpatient and outpatient prescribing and medication recording. Further study is required in this complex high-risk patient group where polypharmacy is unavoidable.

Reference

1. National Coordinating Council for Medication Error Reporting and Prevention. NCC MERP index for categorising medication errors2001. http://www.nccmerp.org/types-medication-errors [Accessed: 30/07/2017].