Patterns of Opioid Use and Risk of Opioid Overdose

By | July 6, 2017

Opiate overdose is now the leading cause of accidental death in the United States, killing more than 50,000 people in 2015. About 20,000 of those deaths were attributed to the use of prescription opiate medications.

As a physician, I want to alleviate my patients’ pain, but I have also taken an oath to do my patients no harm. Clearly, based on these statistics, far too many people have been harmed by prescription opiate medications, either by developing an addiction or through a fatal overdose. We must better understand why this is happening and what we can do to stop it.

Renu K. Garg and colleagues at the University of Washington set out to illuminate risk factors for opioid overdose death among a Medicaid population in their paper, published in this month’s issue of Medical Care: Patterns of Opioid Use and Risk of Opioid Overdose Death Among Medicaid Patients.

The researchers initially looked at all adult Medicaid enrollees in the state of Washington from 2006 to 2010 who had one or more opioid prescriptions, a sample of 328,445 individuals. After removing patients with cancer, methadone maintenance, and other exclusion factors, they were left with a sample of 150,821 patients. Garg and colleagues then identified overdose by reviewing those death certificates and identified 316 with a cause of death attributed to prescription opiate overdose.

They examined each patient’s average daily dose of opiate medication, use of long-acting or short-acting medication, duration of use, as well as recent and concurrent sedative-hypnotic and other sedating medications such as skeletal muscle relaxants. You can find a complete list of the medications analyzed here [Word file].

They found, intuitively, that individuals on higher doses of opiates had higher rates of overdose death, demonstrating a clear dose-response relationship. Compared to the lowest dose category of less than 20 mg/day, patients taking 50 mg/day or more had a significantly higher risk of death (see figure below).

Source: Created for the blog from data in Table 3 of Garg et al., Med Care. 2017:55(7);661–668.

They also found that patients concurrently taking sedative-hypnotics and muscle relaxants had a substantially higher risk of overdose death, even among patients taking less than 20 mg/d of opiates. This finding is extremely important and is thought to be the first report of increased overdose death with these drug combinations.

Clinicians often try to take a multi-pronged approach to pain, and muscle relaxants and benzodiazepines are sometimes used to alleviate muscle spasm in combination with opiates. Benzodiazepines are also used in anxiety disorders, which can be co-morbid with chronic pain. Garg and colleagues’ findings illuminate the very real risks of using these medications in combination, and may allow us to alter our prescribing habits as well as counsel patients on these risks.

Furthermore, we are learning in this epidemic that many medications aside from opiates have a street value – meaning, patients may give away or sell pills remaining after they have used what they need for their pain or anxiety, and others may use them recreationally. This is particularly true of the benzodiazepines that Garg and colleagues studied. Discretion in prescribing appropriate quantities is part of our responsibility as clinicians.

Almost every state maintains an online database that tracks prescriptions of controlled substances to assist us in that task. This allows prescribers to ensure that patients aren’t getting controlled substances from multiple doctors and increases the safety of prescribing these medications.

Starting in August, Massachusetts will also track prescriptions for gabapentin, an anti-convulsant (anti-seizure medication) that is also widely used for neuropathic pain. Individuals sometimes use it with opiates to augment or prolong a high. I am unaware of any studies demonstrating whether gabapentin does indeed increase the risk of overdose, but based on these findings, I suspect it may. Further studies on this question are needed.

Lastly, we must continue to work toward ending the stigma against individuals suffering from substance use disorders. Such stigma is preventing the provision of a database of controlled substances for clinicians in Missouri, the only state without one, where a legislator once stated that overdose “removes them from the gene pool.”

A friend recently sent me an obituary for a young man, Garrett R. Moody, who died of an overdose. His courageous and loving family wrote beautifully about him and his struggle, words that we all must keep in our hearts: “It’s important to not allow addiction to define a person. An addict is a person with a disease. They are someone’s child, sibling, significant other, best friend – they mean something to someone.”

Audrey Provenzano

Audrey Provenzano

Audrey M. Provenzano, MD, MPH is a General Internist in the Boston area. She cares for patients in a community health center and works in quality improvement. She is interested in primary care practice reform and quality improvement and healthcare policy. She is the host and producer of the primary care and health policy podcast Review of Systems, which can be found at
Audrey Provenzano

Latest posts by Audrey Provenzano (see all)