Article Text
PDF
Abstract
Background Medication errors are a significant problem in theUK, and have the potential to cause harm. Children are particularly vulnerable. Errors occur at different decision-making points in the medicines pathway, and prescribing errors constitute a significant but potentially modifiable step in this process, which we targeted.
Our hospital regularly audits prescribing against our local prescribing policy (the gold standard), which is based upon GMC guidance for safe prescribing. The rate of prescription error was 67% in 2010. We defined ‘prescription error’ as any deviation from the gold standard.
From August 2010, our mandatory junior doctor induction was modified to include a 2 h lecture on paediatric prescribing (delivered by pharmacist and doctor) and a paediatric prescribing test. This combination was shown to reduce prescribing errors from 67% to 52%, and reduce incorrect or absent allergy information on prescription charts from 36% to 15% in 2010 and 2012 respectively.
The problem Although effective in improving prescribing, the lecture had a large volume of information to deliver in a limited time. Administering and marking the test was time consuming and providing individual feedback was logistically challenging.
A potential solution Web-based learning modules (WBLM) are accessible, low cost and can facilitate simultaneous education and assessment.
Aim To design and pilot a WBLM to replicate the previous content of prescribing induction. The long-term aim is to continue to improve paediatric prescribing within our hospital.
Primary objective Assess feasibility of introducing a WBLM as part of paediatric induction.
Secondary objective Assess the effect of the WBLM on prescribing practice.
The intervention The WBLM was designed by paediatricians (the authors) working with a Learnpro designer. It can be accessed remotely and is PC or tablet compatible.
There are 3 main sections. The first is an introduction to paediatric prescribing, revising pharmacokinetics and local policy for safe prescribing. The second section uses clinical cases to illustrate prescribing challenges and focuses on areas where errors frequently occur. Adverse drug reactions, antibiotic monitoring, medication error reporting and discharge prescriptions are covered. The final section focuses on intravenous fluids.
All sections contain links to relevant local clinical guidelines. Summative assessment questions are incorporated into the module, requiring trainees to obtain the correct answer before progressing.
Implementation The WBLM was designed in early 2014. Usability was tested in a small group prior to a pilot in August 2014 for junior medical and surgical doctors attending hospital induction.
Measurement of Improvement The Plan-Do-Study-Act cycle (PDSA) provides a structure for iterative testing of changes to improve quality. We are currently in the third cycle following the August 2014 pilot. All post-completion evaluations were positive. Minor technical issues reported have been addressed. The small scale testing has allowed adaptation according to feedback received.
All inpatient prescription charts, excluding PICU who use electronic prescribing, were audited in November 2014 (Figure 1).
Improving trends in prescription error and allergy documentation over time
There were 78 prescription charts containing 904 individual prescriptions. 85% of ‘regular’ prescriptions were prescribed correctly. Although only 68% of total prescriptions were correct, a significant proportion of errors were in ‘once only’ and ‘PRN’ sections, commonly completed in A+E by doctors, who had not undertaken the WBLM. From 2015 the WBLM will be mandatory for all trainees rotating through the hospital, including A+E, anaesthesia and PICU.
Conclusion It is feasible to implement a WBLM as part of paediatric prescribing induction. Extending implementation to all hospital departments has the potential to be effective in improving paediatric prescribing practice. The WBLM must work in conjunction with other measures to address human and system factors in prescribing errors. These effects should be studied with ongoing PDSA cycles.