Computerized Physician Order Entry and Pediatric Prescribing Errors: A Multicenter Database Study

To compare dosing error rates between computerized physician order entry (CPOE) and free-text electronic prescribing in hospitalized children during the same time period, given inconsistent evidence regarding the effectiveness of CPOE in pediatric settings.

We conducted a retrospective multicenter cohort study across four Israeli general hospitals between July 1, 2022, and December 31, 2023. A computerized algorithm analyzed the ten most frequently administered systemic medications and flagged potential dosing errors defined as a ≥20% deviation from weight-based recommendations. All flagged prescriptions underwent blinded independent review by a senior pediatrician with toxicology expertise and a pediatric clinical pharmacist to confirm clinical significance. Multivariable logistic regression was used to estimate adjusted odds ratios (aORs).

Of 122,869 medication orders for 63,081 children, the confirmed error rate was 3.6%. In unadjusted analyses, dosing errors were more frequent with CPOE than with manual prescribing (3.8% vs 2.8%; p<0.001). After adjustment, CPOE was not associated with a significant overall difference in error risk (aOR 0.98; 95% confidence interval [CI] 0.88-1.09). In the enteral-only analysis, CPOE was associated with a 33% higher risk of errors in enteral medications (aOR 1.33; 95% CI 1.18-1.51; p<0.001). Error risk also varied by hospital site, patient weight, care setting, and administration route.

In real-world pediatric settings, CPOE was not associated with a lower overall incidence of dosing errors, and higher error rates were observed among enteral medications. These results suggest that CPOE may introduce route-specific vulnerabilities, highlighting the need for system-specific optimization rather than assuming electronic ordering is inherently safer.