Longitudinal study of the manifestations and mechanisms of technology-related prescribing errors in pediatrics

J Am Med Inform Assoc. 2025 Jan 1;32(1):105-112. doi: 10.1093/jamia/ocae218.

Abstract

Objectives: To examine changes in technology-related errors (TREs), their manifestations and underlying mechanisms at 3 time points after the implementation of computerized provider order entry (CPOE) in an electronic health record; and evaluate the clinical decision support (CDS) available to mitigate the TREs at 5-years post-CPOE.

Materials and methods: Prescribing errors (n = 1315) of moderate, major, or serious potential harm identified through review of 35 322 orders at 3 time points (immediately, 1-year, and 4-years post-CPOE) were assessed to identify TREs at a tertiary pediatric hospital. TREs were coded using the Technology-Related Error Mechanism classification. TRE rates, percentage of prescribing errors that were TREs, and mechanism rates were compared over time. Each TRE was tested in the CPOE 5-years post-implementation to assess the availability of CDS to mitigate the error.

Results: TREs accounted for 32.5% (n = 428) of prescribing errors; an adjusted rate of 1.49 TREs/100 orders (95% confidence interval [CI]: 1.06, 1.92). At 1-year post-CPOE, the rate of TREs was 40% lower than immediately post (incident rate ratio [IRR]: 0.60; 95% CI: 0.41, 0.89). However, at 4-years post, the TRE rate was not significantly different to baseline (IRR: 0.80; 95% CI: 0.59, 1.08). "New workflows required by the CPOE" was the most frequent TRE mechanism at all time points. CDS was available to mitigate 32.7% of TREs.

Discussion: In a pediatric setting, TREs persisted 4-years post-CPOE with no difference in the rate compared to immediately post-CPOE.

Conclusion: Greater attention is required to address TREs to enhance the safety benefits of systems.

Keywords: electronic health records; informatics; medication errors; patient safety; user-centered design.

MeSH terms

  • Child
  • Decision Support Systems, Clinical*
  • Electronic Health Records*
  • Hospitals, Pediatric
  • Humans
  • Longitudinal Studies
  • Medical Order Entry Systems*
  • Medication Errors* / statistics & numerical data
  • Pediatrics