EHR data helps predict opioid use
A prediction model leveraging electronic health record data could be used to help providers identify hospitalized patients who are at highest risk of progressing to chronic opioid use after they’re discharged from the hospital.
Researchers at the University of Colorado Anschutz Medical Campus, who developed the model using patient data from the Denver Health Medical Center, say the model could be integrated into the EHR and activated in the form of an alert when a physician orders opioid medication, informing the doctor of their patient’s risk for developing chronic opioid use.
“The goal was to identify who these patients were to let providers know at the time of care that these patients are at higher risk, so they think twice before they prescribe an opioid or think about other ways to manage their pain in the hospital setting,” says Susan Calcaterra, MD, a fellow in addiction medicine at the CU School of Medicine and lead author of a study published last month in the Journal of General Internal Medicine.
Calcaterra contends that prior to the study by her and her colleagues, no prediction model had been published to identify hospitalized patients at high risk of future chronic opioid therapy (COT)—defined as either receiving a 90-day or greater supply of oral opioids with less than a 30-day gap in supply over a 180-day period, or filling 10 or more opioid prescriptions over the course of a year.
“Our model accessed EHR data to predict 79 percent of the future COT among hospitalized patients,” found the study. “Application of such a predictive model within the EHR could identify patients at high risk for future chronic opioid use to allow clinicians to provide early patient education about pain management strategies and, when able, to wean opioids prior to discharge while incorporating alternative therapies for pain into discharge planning.”
“Hospital providers are really busy, and oftentimes, they don’t know their patients very well because it may be the first time they’ve ever met them,” adds Calcaterra. “The nice thing is that all of the data required to assess risk are available already documented in the electronic health record, and providers do not need to ask for additional information from patients.”
Researchers conducted a retrospective analysis of EHR data from 2008 to 2014 using logistic regression and identified patient-specific factors that were most associated with the progression to COT, including having a history of substance use disorder; past year receipt of a benzodiazepine; an opioid medication or a non-opioid analgesic; as well as receipt of an opioid at hospital discharge and high opioid requirements during hospitalization.
However, undergoing a surgical procedure during the hospitalization was not associated with progression to COT, according to Calcaterra.
“The purpose is to provide physicians with information to make their clinical decision,” she concludes. “We developed the model in one specific patient population, a safety net hospital. The next step would be to validate the model in a different healthcare system—which we’ll probably do at the University of Colorado, with that patient population, and see how well it does in terms of predicting people that progress to chronic opioid use.”