By Pat Anson, PNN Editor
A recent study published in JAMA Network Open made headlines when it claimed that use of prescription opioids more than doubled the risk of causing a fatal car accident.
Researchers at Columbia University Mailman School of Public Health looked at death certificates and toxicology tests on drivers involved in over 18,000 fatal two-vehicle accidents from 1993 to 2016. They found that 1,467 of the “crash initiators” tested positive for hydrocodone, morphine, oxycodone or other prescription opioids.
The researchers reported their findings as “compelling evidence that use of prescription opioids by drivers is a significant contributing factor” in fatal crashes, which was a dog whistle for media outlets like CNN to warn, “Opioid epidemic spilling over onto roads.”
It turns out the evidence was not so compelling after all.
Like many states, Maryland faces growing rates of opioid abuse and addiction. But when researchers looked at opioid-related crashes in Maryland from 2006 to 2017, they found no significant increase in fatal accidents.
“The fact that opioid crashes in Maryland over the last 10 years have been more or less steady was a surprise,” Johnathon Ehsani, PhD, a professor at the Johns Hopkins Bloomberg School of Public Health told The Washington Post.
“That is striking, because Maryland is one of those states that has been quite severely affected by the overall opioid epidemic.”
Ehsani and his colleagues took another look at the toxicology data and realized it was misleading. The reason was simple: Because hospitals and first responders often give opioid medication to patients injured in accidents, autopsies will detect those opioids in patients who later die from their injuries. In other words, the drivers may not have been under the influence of opioids before the crash.
When researchers only looked at a sub-sample of drivers who died at the scene, they found that the use of prescription opioids was “considerably lower than those who died hours or days following the crash.”
“We applied a stricter definition than our other colleagues did,” said Ehsani, who published his findings in the less prominent journal Accident Analysis & Prevention. “We were more conservative in our case definition.”
The JAMA study apparently over-counted the number of opioid-related crashes by not differentiating between drivers who died at the scene and those who died after getting medical treatment with opioids.
"When determining the prevalence of opioid use in a population, it can be tricky to untangle the two circumstances," says Ehsani. "Thinking differently about the way in which researchers count cases has implications that can offer policymakers and public health professionals more meaningful results."
No one is suggesting that it’s okay to drive a vehicle while impaired by opioids. But it’s also not okay for researchers and prominent medical journals to use flawed data to suggest that patients using prescription opioids are more likely to cause fatal accidents.
In Maryland, they found just the opposite may be true.
“It could be that a number of people who are impaired by opioids choose not to drive,” Ehsani said. “It could be that there is some self-selection going on in that population or it could be that some of those who are most seriously impaired are unable to or don’t have access to a vehicle to drive.”