1786. Separating the Rash from the Chaff: Novel Clinical Decision Support Deployed During the Mpox Outbreak

BACKGROUND: The mpox outbreak of 2022 posed significant challenges to clinicians: how to identify individuals with potential infection and then take appropriate next steps, all in the context of an emerging infection with which they were unfamiliar. By facilitating adaptable, guideline-concordant management, clinical decision support systems (CDSS) have been shown to support clinicians in such efforts. METHODS: We integrated the “Evaluate for mpox” (.evalmpox) CDSS into the EHR (Epic Systems) in a large, integrated healthcare system. .evalmpox was designed to facilitate identification of patients with epidemiological risk factors warranting Person Under Investigation (PUI) status (Figure 1). We previously described our CDSS prototype and the evaluation of 55 patients through 7/20/2022. Here, we describe substantive tool enhancements including improved symptom collection, patient triage, and automated isolation and de-isolation workflows (Figure 2). We also report use characteristics and tool performance through 4/13/2023. [Figure: see text] [Figure: see text] RESULTS: .evalmpox was used in 721 encounters (Figure 3) originating from 114 clinical locations, with 262, 249, 185, and 25 from the ED, outpatient, urgent care, and inpatient settings. 650 total non-duplicate patients were evaluated, with 265 classified as PUIs based on rash and the presence of at least one epidemiological risk factor, and 385 as non-PUI based on the absence of reported risk factors (Figure 4). Among 265 PUIs, 185 had mpox testing performed: 103 (57%), 69 (36%), 10 (5%), and 3 (2%) were negative, positive, indeterminate, or unable to be run. Among the 385 non-PUIs, 49 had mpox testing: 43 (88%), 4 (8%), and 1 (2%) were negative, positive, and indeterminate. The PPV of .evalmpox for a positive test in our EHR was 38% (95% CI 31%-45%) and the NPV was 99% (95% CI 97%-100%). Chart review revealed two of the four patients classified as non-PUI by .evalmpox who tested positive for mpox reported risk factors to clinicians that were not input into the tool. [Figure: see text] [Figure: see text] CONCLUSION: In the setting of an evolving outbreak, .evalmpox facilitated the evaluation of patients presenting with possible mpox across a large healthcare system in diverse clinical locations and performed with favorable test characteristics. DISCLOSURES: Jacob E. Lazarus, MD, PhD, SEED Biologics: Advisor/Consultant|UpToDate: Advisor/Consultant Kevin L. Ard, MD, Binx Health: In-kind research support Elizabeth L. Hohmann, M.D., gilead: Advisor/Consultant|Kowa Pharmaceuticals: Advisor/Consultant