Impact of Provider-in-Triage on Imaging Utilization in Abdominal Pain Complaints

The Provider in Triage (PIT) model has shown benefit in improving Emergency Department (ED) efficiency, but the impact on advanced imaging utilization is less defined. Our objective was to investigate the association of PIT with computed tomography (CT) utilization in patients with a chief complaint of abdominal pain.

This single-center retrospective study was conducted at a tertiary care ED, where PIT was implemented in July 2023 with 15 hours of daily attending coverage. Inclusion criteria were patients assigned an emergency severity index (ESI) score of 3 who presented with a chief complaint of abdominal pain. To meet the target sample size requirement, 14 randomly selected weeks from calendar year 2024 were abstracted using a standardized abstraction tool. This included age, gender, admission status, race, presence of CT abdomen pelvis (AP) imaging, and presence of a PIT assessment. Multivariate logistic regression analysis was performed to evaluate the association of PIT exposure and likelihood of CT AP imaging, with data presented as odds ratio (OR) with 95% confidence intervals (CI). Using a previously published metric, each CT AP was also assessed for clinical significance. A blinded second investigator reassessed a random 15% sample of CT AP studies, with interrater reliability assessed using Cohen’s kappa.

There were 860 PIT encounters and 631 non-PIT encounters in the cohort. In the PIT group, 481 (55.9%) had CT AP performed, as compared to 305 (48.3%) in the non-PIT cohort. In multivariable logistic regression adjusting for age, sex, race, and admission status, evaluation in the provider-in-triage (PIT) model was independently associated with CT utilization (adjusted OR 1.27, 95%CI 1.01–1.60, p = 0.039). In the PIT cohort, 38.7% of patients had positive CT AP findings, as compared to 46.9% of patients in the non-PIT cohort, with an absolute risk difference of 8.2% (95% CI 1.1%-15.3%), p=0.03. Cohen's kappa between raters for clinical significance of CT AP exams was excellent, k=0.86 (95% CI 0.76-0.96).

Patients presenting with a chief complaint of abdominal pain were more likely to receive CT AP imaging if they were assessed in PIT. A higher proportion of CT AP exams in the PIT cohort were negative for clinical pathology. Further study is needed to explore additional potential confounders to further investigate this relationship between PIT exposure and CT AP imaging.