Follow Us:
Opioid 101:
Contact us:
Thank you! We’ve sent you an email to confirm your address.
Drug overdose deaths decline for sixth straight month
By ROGER PARLOFF|October 22, 2018
CDC's Provisional Drug Overdose Death Counts
Quick Takeaway
  • Drug OD deaths, including opioid OD deaths, have declined for six straight months.
  • "If opioid overdose deaths in 2018 drop or stay about the same as 2017, I believe it will be the first time since 1995."—Andrew Kolodny, MD

Drug overdose deaths in the United States have now fallen for six straight months, according to the CDC’s most recent data, dropping 2.8 percent from their peak.

Similarly, the subset of those deaths attributable to opioid drugs has steadily declined over the same period, falling 2.3 percent.

These modest, but steady, declines are reflected in “provisional” data published by the Centers for Disease Control and Prevention, which are subject to adjustment as additional data come in. The promising new figures corroborate and extend a trend we noted two months ago in an article entitled, “Have drug overdose deaths peaked?”

Fatalities from heroin overdose (a component of the total opioid death tally) have now been falling for an even longer stretch—eight straight months. They are down 6.9 percent from their peak. Fatalities from the most common forms of prescription opioids*—like oxycodone and hydrocodone—have, likewise, declined for eight consecutive months. They’re now 7.2 percent below their highest level.

“Still too soon to say, but it sure looks like we may finally have crested.”—Andrew Kolodny

On the other hand, deaths associated with the deadly synthetic opioid fentanyl**—50 times more powerful than heroin—continue to rise. Also still rising are deaths from cocaine and methamphetamine***—substances that are increasingly being contaminated with fentanyl, according to multiple news accounts.

But even as fentanyl deaths continue to rise, the steepness of their climb has flattened over the last six months. That has allowed the combined drops in heroin and prescription drug fatalities to outpace the rising fentanyl deaths, bringing down the total drug overdose numbers. That, in turn, has resulted in the first monthly declines in the three years for which the CDC has been tracking these monthly figures.

“Still too soon to say,” says Andrew Kolodny, MD, co-director of the Opioid Policy Research Collaborative at Brandeis University’s Heller School, “but it sure looks like we may finally have crested. If overdose deaths in 2018 drop or stay about the same as 2017, I believe it will be the first time since 1995 that opioid overdose deaths didn’t increase.”

Opioid use disorder vastly undercounted: Massachusetts study

Rather than publishing monthly death tolls—data that might be misleading due to seasonal variations—the CDC publishes death tolls for rolling 12-month periods. Each new period’s data become available after a six-month lag. For instance, the data released last week concern the period from April 1, 2017 to March 30, 2018.

CDC Provisional Drug Overdose Death Data (interactive graph at

During that stretch, the CDC found that 71,073 people died of overdoses from all categories of drugs. That’s 1.6 percent less than the 72,240 that died in calendar 2017. It’s also 2.8 percent fewer than the 73,133 that died during what we can all fervently hope was the epidemic’s peak—the 12-month stretch between October 1, 2016 and September 30, 2017.

Looking at the opioid death toll in particular, 48,400 died during the most recent period, down 1.3 percent from calendar year 2017, when 49,038 died, and down 2.3 percent from the peak stretch ending in September 2017, when 49,541 died.

CDC’s Provisional Opioid Drug Death Data
(interactive graph at

(All these figures are “predictions,” or statistical projections, based on the hard data the CDC has received and analyzed to date. The agency warns that “true declines or plateaus . . .  cannot be ascertained until final data become available.” Figures for 2017 will become final in December, but figures for 2018 won’t become final until late 2019.)

Drilling down to the level of individual states, the data are also heartening. When we wrote just four months ago about the latest CDC data (for the period ending November 30, 2017), only eight states showed year-over-year reductions in drug overdose deaths. And only one of those eight—Rhode Island—was among those that had been hard hit by illicit fentanyl.

In the data released last week, a remarkable 20 states showed year-over-year reductions in overdose death, including such fentanyl-ravaged locations as the District of Columbia, Kentucky, Ohio, Massachusetts, New York, Pennsylvania, Vermont, and, again, Rhode Island. DC’s deaths were down 14 percent; Vermont’s, 11.9 percent; Rhode Island’s, 9.7 percent.

CDC Data on Percent Change in 12-Month Drug Overdose Death Totals By State
(interactive map available at

“The most important factor is treatment—particularly with buprenorphine.”

Public health and government authorities have, of course, started many programs to fight the opioid epidemic in recent years. These have included widening the availability of the overdose-reversal drug naloxone (Narcan); improving access to the addiction-treatment medication buprenorphine (like Suboxone), including in some prison and jail settings; reducing opioid prescribing, which has dropped 29 precent since 2011; improving disposal of unused prescription opioids; deploying needle-exchange programs, and many others. What’s working?

Kolodny says it’s hard to know what’s causing the overdose decline, but that it’s certainly a combination of factors. “More cautious prescribing is helping to reduce opioid prescription deaths, but I think the most important factor is treatment—particularly with buprenorphine. I think more Americans are accessing effective treatment.”

Keith Humphreys, PhD, a professor of psychiatry at Stanford University School of Medicine, was reluctant to jump to any positive conclusions when we reported on the CDC data two months ago. But this time he is more sanguine.

“You have longer term data from more places,” he writes in an email, “so, yes, that is more encouraging.”

He does observe that it’s hard to read much positive into the decline in heroin deaths, given the continuing increase in fentanyl deaths. When people die from heroin that’s been laced with fentanyl, he writes, coroners and medical examiners are probably now reporting those as fentanyl deaths, rather than heroin deaths.

In fact, the rise in fentanyl deaths (1,448 more in the most recent 12-month period than in the one that ended six months earlier) is still outpacing the decline in heroin deaths over the same period (1,095 fewer).

“The drop in prescription opioid deaths is particularly encouraging.”—Keith Humphreys

On the other hand, Humphreys says, “the drop in prescription opioid deaths is particularly encouraging.” It suggests, he believes, that we may be beginning to feel the benefits from the reductions in opioid prescribing.

“I have often talked about this in terms of stock and flow,” he writes. By “flow” he means the new influx of opioid use disorder cases that develop when people are exposed to opioids they obtained either directly from a doctor or from someone else who got them from a doctor. By “stock” he means those who are already addicted to prescription opioids.

To end an epidemic, you must both treat those who are sick (the stock), and try to stop new cases from breaking out (the flow).

While reductions in opioid prescribing have the beneficial effect of cutting down on new cases of OUD, Humphreys says, they run the risk of driving some number of the “stock” into the illicit opioid market, which could actually increase the number of overdose deaths in the short term.

“Because the stock is finite,” he writes, “and the flow over time is potentially infinite, with each year the benefits to the latter group will become larger relative to the costs of the former.”

One closing note of caution. While overdose deaths are the most grisly statistic of the epidemic, they are only one barometer of its toll. Other measures, like cases of neonatal abstinence syndrome in hospitals, or strains on state foster care programs, or nonfatal emergency room visits prompted by addiction, might give a different picture. The CDC’s latest data on emergency room visits due to nonfatal opioid overdoses, analyzing the period from July 2016 to September 2017, still showed those rates soaring by nearly 30 percent nationally, and by more than 54 percent in large metropolitan areas.


*The CDC calls this category “natural and semi-synthetic opioids.” It does not include synthetic prescription drugs, like tapentadol, tramadol, prescription fentanyl (Subsys, Duragesic, Actiq), methadone, or buprenorphine.

**This CDC does not track “fentanyl” per se, but rather a category that comprises all “synthetic opioids” (except methadone, which is tracked separately). The category includes a number of prescription drugs as well as unlawful drugs. But the CDC believes that, since 2013, the vast majority of overdose deaths associated with this category are being caused by illicitly manufactured and trafficked fentanyl or close chemical variations.

***The CDC data does not track “methamphetamine” per se, but a category including all “psychostimulants with abuse potential.” The drug from this group most frequently causing overdose death, however, is believed to be methamphetamine.

Filed under: Addiction/ Fentanyl/ Research