Long-acting opioids were associated with a greater than 2-fold risk for unintentional overdose compared with short-acting formulations, according to a cohort study published online February 16 in JAMA Internal Medicine. Moreover, the risk was more than 5-fold greater in the first 2 weeks of using a long-acting opioid.

“If replicated in other cohorts, our findings suggest that clinicians weighing the benefits and risks of initiating different opioid regimens should consider not only the daily dose prescribed but also the duration of opioid action, favoring short-acting agents whenever possible, especially during the first 2 weeks of therapy,” write Matthew Miller, MD, ScD, from the Department of Health Sciences and Epidemiology at Northeastern University in Boston, Massachusetts, and colleagues. The study population was made up of veterans who were largely men older than 50 years.

Previous studies had found that higher doses of opioids were associated with greater risk for overdose, but few studies have considered the duration of action in overdose risk, the authors note.

The current study involved 840,606 veterans with chronic pain diagnoses who began opioid therapy between January 1, 2000, and December 31, 2009, who had not taken opioids in the previous 6 months. The researchers reviewed diagnostic and procedural codes as well as pharmacy data from the Veterans Health Administration. They note that they were not able to track the actual amount of medication taken, but considered patients to be using the opioids on the basis of filled and refilled prescriptions.

Overall, 18,887 patients used long-acting opioids, which included orally administered sustained-release morphine sulfate, methadone hydrochloride (but not the liquid methadone used for treating addiction), controlled-release oxycodone hydrochloride, levorphanol tartrate, and fentanyl patches. Short-acting opioids, taken by the majority of patients in the study (801,729), included codeine phosphate, hydrocodone, and oxycodone that were orally administered either by themselves or in a formulation with acetaminophen or aspirin. The authors note that patients who were given prescriptions for long-acting opioids were more likely to be given higher doses and were more likely to also be taking antidepressants and benzodiazepines.

Three hundred nineteen patients experienced an unintentional, nonfatal overdose. After adjustment for age, sex, opioid dose, and other covariates, the hazard ratio for overdose was substantially higher for those receiving long-acting opioids (2.33; 95% confidence interval, 1.26 – 4.32). The risk for overdose during the first 2 weeks of treatment was more than 5 times higher for those taking long-acting opioids (hazard ratio, 5.25; 95% confidence interval, 1.88 – 14.72).

The authors note that the rate of overdose is likely to be an underestimate, as events may go unreported if patients die, do not seek medical attention, or receive medical attention outside the Veterans Affairs system. They also note that they were unable to adjust for the severity of illness or substance use disorders and that other confounders may be present at baseline, although they write that baseline confounders are unlikely to explain the elevated risk at the beginning of treatment.

“The best solution is to avoid prescribing opioids for chronic pain because there is no high-quality evidence that they are effective for this indication, and the risk of adverse effects, including death from unintentional overdose, is great,” writes Mitchell Katz, MD, the journal’s deputy editor, in an editorial. Despite the possibility of confounding factors, he writes, “it would be wise to avoid long-acting agents when initiating opioid therapy for chronic pain…. Meanwhile, there is an urgent need for better nonopioid treatments for chronic pain.”

The authors and Dr Katz have disclosed no relevant financial relationships.

JAMA Intern Med. Published online February 16, 2015. Abstract