Pediatric Appendicitis Pathway and Order Panel Adoption: Imaging and Antibiotic Use in New York City Emergency Departments

Clinical decision support (CDS) tools, including order panels, aim to provide guideline-based care and reduce practice variation. Pediatric appendicitis is a common Emergency Department (ED) presentation with variability in diagnosis and treatment. The integration of an appendicitis order panel alongside a clinical guideline pathway aimed to support ED providers in diagnostic imaging and treatment selection.

We performed a retrospective pre/post analysis across 3 academic New York City EDs of encounters among patients <18 years with a visit diagnosis of appendicitis (ICD-10 K35–K37). We compared the 12 months before versus after implementation (3/2023) of an EHR order panel with embedded CDS and pathway linkage. Data were obtained from Epic Clarity and analyzed in Databricks. Secondary analyses compared post-implementation panel users versus non-users. Outcomes included abdomen/pelvis computed tomography (CT A/P) use, antibiotic selection, antibiotic-regimen variability (Shannon entropy), and time to disposition. Statistical analyses used Chi-squared and Mann-Whitney U tests.

Overall, 272 appendicitis encounters were identified (133 pre, 139 post). The panel was used in 69/139 (50%) post-implementation encounters. CT A/P use decreased from 33% pre to 17% post (p=0.004). Individual antibiotic selection proportions were unchanged; however, antibiotic-regimen entropy was 0.82 bits pre, and in the post period was 0.70 for panel users vs 1.20 for non-users. Median time to disposition decreased from 305 to 260 minutes (p=0.12).

Implementation of a pediatric appendicitis EHR order panel linked to a guideline pathway was associated with a significant reduction in CT A/P use and lower antibiotic-regimen variability (entropy) among panel users. CT reduction occurred despite 50% panel uptake, suggesting effects both through direct panel use and indirect pathway dissemination/education. Future work will test portability to other ED pathways and further validate entropy-based measures as scalable metrics of regimen variability beyond single-medication rates.