Regency EF
Step-Wedge Design to Evaluate the Effectiveness of Opioid Prescribing Aids (304083)
*Caroline Ledbetter, University of Colorado
Keywords: step-wedge, simulations, opioids
In 2017, 47,600 people died of an opioid overdose (68% of all overdose deaths). Our study explores the effect of integrating the prescription drug monitoring program into the electronic heath record on physician prescribing behaviors in emergency department. Billions of dollars are being spent to reduce opioid addiction in the US. Stakeholders need data on the most effective efforts but given all the attention, guidelines and interventions, the actual effects of interventions can be difficult to disentangle from secular time trends. Instituting an intervention effort across multiple sites simultaneously can be expensive. Through simulations, we demonstrate that three sites can be used to evaluate the effects of an intervention using a step wedge design even in the presence of complex secular time trends and site effects. Our study found that introduction of the PDMP into the workflow did not reduce overall opioid prescriptions but did reduce the prescribing rates in high risk individuals. Using a shiny app, we allow users to model different ‘TRUE’ effects of time, site, and the intervention and observe the resulting simulation and model performance.