Evaluating patients who present with pain complaints to a community hospital emergency department: Opioid prescription tracking software versus provider gestalt.

Objective: This study aimed to compare provider gestalt to assigned Narx score, a common prescription drug monitor-ing program (PDMP) component that gauges the patient's risk of misuse or abuse.

Design: This is a prospective, anonymous survey from advanced practice providers (APPs), emergency medicine resi-dents, and emergency medicine attendings.

Setting: Data from two emergency departments (EDs) within the OhioHealth network were included. One hospital is a 213-bed academic, community hospital. The other hospital is a 434-bed academic level I trauma center.

Patients, Participants: The survey was open to all providers. Exclusion criteria for patients included prior knowledge of the patient and/or their Narx score, or cancer-related pain.

Interventions: Surveys were collected over a 3-month period. Variables included provider type and level of experience, participant demographics, provider gestalt, and the patient's actual Narx score.

Main Outcome Measure(s): Primary outcome was the ability of providers to accurately estimate a patient's Narx score. Groups were defined as Match = No (gestalt and actual score did not match) and Match = Yes (gestalt and actual score matched). Various characteristics were compared between these two groups.

Results: Providers were able to accurately estimate actual Narx score (72.7 percent). The Match = Yes group was younger (p = 0.01). Dental pain was more common in the Match = No group, 11.5 percent versus 0 percent (p = 0.02). Match = No group also had a higher incidence of triggers. Specifically, any trigger (p = 0.006), explicitly asking for pain medication (p = 0.03), and asking for opioids by name (p = 0.03). Every 10-year decrease in age showed a 1.5 times increased likelihood of accurately estimating Narx score (p = 0.02). Having no triggers showed a three times increased likelihood of accurately estimating Narx score (p = 0.02). Prescribing was largely unchanged after viewing the actual Narx score.

Conclusions: Providers are able to accurately estimate Narx score, though there are limiting factors. Older patients, those with dental pain, and those who give specific triggers are more difficult to estimate. Providers did not change their prescribing patterns after viewing the actual Narx score. Overestimation versus underestimation of Narx score was not directly studied.