TOPLINE: 

A neural network analysis of emergency department (ED) visits for ankle sprain identified age (98.5%) and insurance (71.4%) as top predictors of opioid prescribing. Overall opioid use declined from 29.5% to 15% (2016-2021), whereas ED administration and discharge rates remained stable.

METHODOLOGY:

• Researchers conducted a retrospective analysis of National Hospital Ambulatory Medical Care Survey (NHAMCS) data and examined 5,594,274 ED visits for ankle sprains between 2016 and 2021.
• Patients were predominantly women (55.2%) and White individuals (72.7%), with 31.5% aged 24-44 years and 40.8% covered by Medicaid.
• Researchers used a multilayer perceptron neural network (70/30 training/testing split with 10 iterations) to identify predictors of opioid prescribing, achieving 87% accuracy with an acceptable area under the curve of 0.77.
• The primary outcome was the change in opioid administration or prescribing rates among ED visits for ankle sprain.
• The secondary outcomes were trends in ED visits involving both opioid administration and prescribing and identification of predictors of opioid prescribing.

TAKEAWAY:

• From 2016 to 2021, 20% of all visits for ankle sprains involved opioid administration or prescribing, and the percentage of such visits significantly declined from 29.5% in 2016 to 15.0% in 2021 (P = .03).
• ED opioid administration remained stable at 6.22% of visits, whereas discharge prescriptions remained constant at 7.1% over time.
• Concurrent administration and prescribing occurred in 6.6% of visits, with no significant change over time.
• Age was identified as the strongest predictor of opioid prescribing (normalized importance, 98.5%), followed by insurance status (71.4%), consultation status (66.7%), hospital academic status (56.6%), and geographic region (47.3%).

IN PRACTICE:

"Our study found that while there was an overall decrease in presentations related to ankle sprains over time and an overall decrease in opioid utilization, there were no changes in ED administration, discharge prescriptions, and visits where both opioids were administered and prescribed," the authors wrote. "The results of this study can be used to help further inform the need for targeted interventions that can improve opioid prescribing patterns in the ED and outcomes for patients in this population," they added.

SOURCE:

The study was led by Christine Ramdin, PhD, Rutgers New Jersey Medical School, Newark, New Jersey. It was published online on September 16, 2025, in The American Journal of Emergency Medicine. 

LIMITATIONS:

The NHAMCS database lacked information on medication dosage, treatment duration, and injury severity, preventing analysis of morphine milligram equivalents. Additionally, insufficient data on included hospitals hindered the determination of potential differences in opioid prescribing patterns between rural and urban settings. 

DISCLOSURES:

Information on study funding was not provided. The authors reported having no conflicts of interest.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.