Vanderbilt study finds use of methadone to treat pain carries increased risk of deathJan. 20, 2015, 3:39 PM
Outside the hospital, use of methadone to treat pain carries a 46 percent increased risk of death when compared to the equally effective but more costly alternative, morphine SR (sustained release).
That’s according to a study from Vanderbilt University Medical Center, appearing online this week in JAMA Internal Medicine.
The authors note that there were some 4.4 million methadone prescriptions written in the U.S. in 2009 for treatment of pain. They calculate that methadone’s higher risk ratio translates to 72 excess deaths per every 10,000 person-years of treatment — again, compared to morphine SR.
“That’s quite high by medical standards. And as far as we know, there’s no clinical benefit to using methadone as opposed to morphine SR, so for that reason these deaths are particularly concerning,” said the study’s first author, Wayne Ray, Ph.D., M.S., professor of Health Policy. “It’s a lot of increased risk for a drug that happens to be cheap, but confers no other benefits.”
As a consistent side effect, opioids repress respiration, but methadone is apparently unusual in that the analgesic effect is more short-lived than the respiration effect.
“This means that patients may sense the need for more medication even though there is still enough methadone in their bodies to cause respiratory problems, so they essentially inadvertently overdose themselves,” Ray said.
Another basis for safety concerns is that methadone is a pro-arrhythmic drug that can trigger lethal ventricular arrhythmias.
Citing these concerns, in 2006 the Food and Drug Administration cautioned clinicians concerning methadone use. Subsequently, Centers for Disease Control and Prevention scientists recommended against using methadone as a first-line treatment for pain. The new study’s authors write that their findings “support recommendations that methadone should not be considered a drug of first choice for non-cancer pain.”
Using Tennessee Medicaid records for 1997 through 2009, the team followed 6,014 beneficiaries who took methadone and 32,742 who took morphine SR. People entered and exited follow-up as they started and stopped filling these prescriptions. More than three-fourths of the total prescriptions were for back pain. In all, there were 477 deaths during 28,699 person-years of follow-up.
The sole previous study comparing these two drugs produced a nearly opposite result: 44 percent decreased mortality with methadone.
“That study included a lot of very sick patients, such as patients with cancer, and this may have skewed their results. Our findings are much more consistent with the existing body of data,” Ray said.
To root out potential confounding variables, people in the Vanderbilt study were excluded from follow-up while in the hospital and for 30 days following hospital discharge, and people with life-threatening illnesses or cancer were excluded outright, as were those over age 74 and those in nursing homes.
In calculating the risk ratio, the team accounted for some 196 covariates: demographic factors, comorbidity, medications, recent medical care utilization and so on (this information was updated at the time of each medication fill).
The authors found that methadone’s excess risk was present in the lower half of the dosing range.
“That’s what’s interesting,” Ray said, “because this is consistent with the potential of methadone to accumulate and for patients to inadvertently overdose themselves. Apparently the higher you go with the dose, the less difference there is between methadone and another opioid. That’s what’s particularly concerning in some ways: that a clinician might consider the low dose to be relatively free of overdose risk, but because of methadone’s particular pharmacologic properties that may be incorrect.”
Joining Ray for the study were Cecilia Chung, M.D., MPH; Katherine Murray, M.D.; William Cooper, M.D., MPH; C. Michael Stein, M.B.Ch.B; and Kathleen Hall.
The study was supported by the National Institutes of Health (grants HL081707, AR064768).