Study: Geisinger Researchers Utilize EHR Data to Predict Opioid Overdose Deaths

In a study, researchers at Danville, Pa.-based Geisinger Health System used data from electronic health records (EHRs) to profile patients who overdosed on opioids and developed a predictor for patients most at risk of dying or experiencing other serious complications from drug abuse.

According to a press release about the 10-year study, scientists at Geisinger analyzed the EHRs of more than 2,000 patients admitted to the hospital for overdoses between April 2005 and March 2015. Among them, 9.4 percent died within a year of hospitalization. 

Patients had an average age of 52, were more often female (54 percent), not married (64 percent) and unemployed (78 percent). Their concurrent chronic diseases also included cardiovascular disease (22 percent), diabetes (14 percent), cancer (13 percent) and the presence of one or more mental health disorders (35 percent).

Researchers found that marital status, mental health and employment status all weigh in as predictors.

Being married and having private health insurance were found to have a protective effect, while history of previous addiction, mental illness and having other chronic diseases were all found to be conditions associated with adverse overdose outcomes, including death, according to the study.

"Our study suggests opportunities for identifying patients at-risk for overdosing," study leader Joseph Boscarino, Ph.D, Geisinger addiction researcher and senior epidemiologist, said in a statement. "We've found that patients who are taking higher doses of prescription opioids combined with psychotropic medicines may need closer monitoring to avoid death and other serious complications."

Predictors of the worst patient outcomes–including death, repeated overdoses, frequent health care service use and higher related costs–were found to be higher prescription opioid use, having concurrent chronic diseases, having concurrent mental disorders, and concurrent use of other psychotropic medications, according to the press release about the study.

"These patients have a history of addiction and other serious mental illness both before and after their overdose, as well as current chronic diseases," Boscarino said.

Geisinger researchers will present results from "A 10-year Retrospective Study of Opioid Overdoses among Patients in a Large Integrated Healthcare System" at the International Conference on Opioids.

Opioids–including prescription opioid pain relievers and heroin – killed more than 28,000 people in 2014, more than any year on record, according to the Centers for Disease Control and Prevention.