Skip to content Skip to navigation

Predictors of death in opioid overdose cases come into focus

June 7, 2016
by Tom Valentino, Senior Editor
| Reprints

A study of opioid overdose cases from the past 10 years points to a history of previous addiction, mental illness and having other chronic diseases among the strongest predictors of which overdose patients are most at risk for death or other serious complications.

Researchers at Geisinger Health System (GHS), which covers a wide area of Pennsylvania from State College to Scranton, reviewed electronic health records of 2,039 patients admitted to its system from 2005 to 2015. Geisinger researchers presented their findings at the International Conference on Opioids on Sunday in Boston.

Among their findings:

  • 9.4% of patients studied died within a year of overdose
  • Patients had an average age of 52, were more often female (54% of the 2,039 patients studied), not married (64%) and unemployed (78%)
  • Concurrent chronic diseases included: cardiovascular disease (22%), diabetes (14%), cancer (13%) and the presence of one or more mental health disorders (35%)
  • Higher prescription opioid use, having concurrent chronic diseases and/or concurrent mental disorders, and concurrent use of other psychotropic medications were among the predictors of the worst patient outcomes (death, repeated overdoses, frequent health care service and higher related costs)

Another statistic—that patients in just 9% of cases in the report were given orders for naloxone after their overdoses—was particularly surprising to the researchers, study leader Joseph Boscarino, senior scientist and director of clinical research training at Geisinger’s Center for Health Research, told Addiction Professional in a phone interview.

Boscarino says he hopes the takeaways gleaned from the study can be incorporated into pinpointing at-risk patients and improving overdose outcomes.

 “We have [patients’] mental status, mental codes and mental health disorders, substance use disorder history—we have all that [information] we can see,” he says. “Based on the patterns in the electronic health records studied, we can develop diagnostic screens in the future.

“We can identify who [high-risk patients] are quite well statistically. The idea would be to, on a real-time basis, give that information to the patient while they are on a visit. They can get the intervention, which may be counseling and a naloxone prescription for them and bystander training.”

Topics