Prescribing errors in children

Prescribing errors are a well-known safety concern in pediatric patients. The aim of this thesis was to
investigate factors related to patients, drugs, and the mode of prescription, that influenced the rate of
prescribing errors in a population of 1000 pediatric patients hospitalized on general medical and surgical
wards at the University Children's Hospital Zurich. The data were collected during two periods in 2018
and 2019. In total 5022 prescriptions were analyzed.
The prevalence of prescribing errors under different circumstances of prescribing (2299 pre-CPOE
prescriptions with semi-structured order forms or handwritten prescriptions versus 2723 post-CPOE
prescriptions as electronic prescriptions with limited clinical decision support (CDS)), was investigated
in different age groups and different types of drugs. Additionally, the types of errors that occurred and
the severity of harm potentially caused by these errors were revealed. Furthermore, associations of
errors withs specific drug types or patients were investigated, with a particular emphasis on unlicensed
drugs.
In the first study Prescribing errors in children – what is the influence of a computerized physician order
entry (CPOE)? the influence of the prescribing mode was investigated. The prescriptions of 500 patients
before (2018) and after (2019) introduction of a CPOE were analyzed and prescribing errors assessed.
It was found that the post-CPOE prescriptions overall contained significantly fewer prescribing errors
(25 errors / 100 prescriptions) than pre-CPOE prescriptions (78 errors / 100 prescriptions) (p < 0.001).
Errors that actually could have led to patient harm with a severity rated as “temporary harm possible” to
“death” (further referred to as potentially harmful errors: PHE), were also reduced from 18 errors / 100
prescriptions pre-CPOE to 11 errors / 100 prescriptions post-CPOE (p < 0.001). The errors that occurred
most frequently in the pre-CPOE period were errors due to missing information. These errors were of
minor severity and were strongly decreased post-CPOE, where dosing errors were the most frequent
type of error. A statistically significant increase in medication reconciliation errors was observed after
CPOE introduction due to remaining hybrid prescriptions in certain cases. Overall, the CPOE had a
positive impact on patient safety.
In Prescribing patterns in pediatric general wards and their association with prescribing errors, the
second study, a sub-analysis of the first study was conducted in which prescribing patterns in the post-
CPOE population were revealed. Newborns were excluded from the study due to small sample size,
leading to a remaining population of 489 patients with 2693 prescriptions. Drugs for the nervous system,
drugs for the alimentary system and anti-infective drugs were the most frequently prescribed drug
classes, with paracetamol, metamizole and ibuprofen being the most frequently prescribed active
substances. Patient characteristics like age and gender as well as drug use were associated with
prescribing errors. Children between 2 and 11 years experienced higher error rates than infants under
2 years of age: 12.2 potentially harmful errors (PHE) / 100 prescriptions, vs. 8.5 PHE / 100 prescriptions
(p = 0.026). A statistically significant difference was also found for female patients as compared to male
patients, with the female patients having higher rates of PHE than the male (25.6 errors or 12.1 PHE /
100 prescriptions vs. 24.7 errors or 9.3 PHE / 100 prescriptions) (p = 0.035 for PHE), even though there
was no difference in the overall error rates. This finding needs further investigation.
The third study, Use of unlicensed drugs in a Swiss Pediatric University Hospital and associated
prescribing error rates, another sub-analysis of the first study laid a focus on unlicensed drugs in the
population of 1000 patients (pre- and post-CPOE), which accounted for a proportion of 10.8% of all
prescriptions. 34% of patients were prescribed at least one unlicensed drug. Oral liquid forms were the
most frequently prescribed drug form in unlicensed drugs. In the post-CPOE population unlicensed
drugs were more prone to prescribing errors than licensed drugs (32 errors / 100 prescriptions vs.
24 errors / 100 prescriptions, p = 0.024). Particularly extemporaneously prepared drugs had high error
rates of 36.4 errors / 100 prescriptions. Therefore, licensed drugs are favorable in terms of medication
safety.
Overall, this thesis highlights various aspects of prescribing errors in children and illustrates, that
pediatric patients are still at a high risk of experiencing a prescribing error. Ongoing efforts are necessary
to improve medication safety. These include electronic solutions like CPOE and CDS, multifaceted
approaches on healthcare professional and organizational level as well as support from the
pharmaceutical industry by licensing more suitable formulations for pediatric patients.