Far more Massachusetts residents suffer from opioid use disorder than has previously been understood, according to a new prevalence study by Boston Medical Center researchers.
It’s a startlingly basic discovery so late in an epidemic that has been so minutely analyzed, inspected, and dissected.
As of 2015, 4.6 percent of all state residents over the age of 11—about 275,000 people—had OUD, according to the new estimate. The research was published Thursday in the American Journal of Public Health, and first reported by the Boston Globe.
That prevalence figure was “nearly four times higher than current national prevalence estimates,” wrote lead author Joshua A. Barocas, MD, in the journal article. He is an assistant professor of medicine with the division of infectious diseases at Boston University School of Medicine and Boston Medical Center.
The previous estimates came from SAMHSA’s National Survey on Drug Use and Health.
“The finding is horrifying both for the suffering it represents and for how it reveals the inadequacy of our existing methods of tracking the epidemic.”
“The finding is horrifying both for the suffering it represents and for how it reveals the inadequacy of our existing methods of tracking the epidemic,” says Keith Humphreys, PhD, a professor of psychiatry at Stanford University School of Medicine, in an email to Opioid Watch.
“We have been relying on a household survey that excludes high-prevalence populations (e.g., prisoners) and assumes people all honestly report their addiction to a government surveyor. As a result we have grossly underestimated the severity of the epidemic.” Humphreys served as a drug policy advisor to both the George W. Bush and Barack Obama White Houses.
The most novel aspect of the research was that it employed statistical sampling techniques to count people who had the disease even if they had neither interacted with the health system nor identified themselves on surveys—the typical ways of measuring a disease’s prevalence.
“We hypothesized that national prevalence estimates were underestimating opioid use disorder prevalence,” Barocas explains to Opioid Watch in an email. People suffering from it might not seek health care or identify themselves on surveys because of stigma, fear of legal repercussions, and lack of medical insurance. (“Individuals with OUD are less likely than members of the general population to have medical insurance,” the authors write, “even in Massachusetts, where a large proportion of residents are insured.”)
The estimation technique, known as “capture-recapture analysis,” was developed originally to estimate wildlife populations.
“It is not commonly performed in human populations since we often need more than one database to use this method,” Barocas explains. “It is a powerful tool, however, for other states looking to get better estimates of their OUD prevalence.” (Barocas’ coauthors were Laura F. White, PhD; Jianing Wang; Alexander Y. Walley, MD; Marc R. LaRochelle, MD; Dana Bernson; Thomas Land, PhD; Jake R. Morgan, PhD; Jeffrey H. Samet, MD; and Benjamin P. Linas, MD.)
When applied to wildlife, the technique works like this. Researchers might capture 200 fish from a body of water, tag them, and release them. Then, after an interval, they’d capture 100 more fish—the “recapture.” Suppose, then, that 10 of the 100 fish are found to be tagged. The technique then assumes, the authors explain, that “the fraction of the tagged fish in the recapture (10/100) is proportional to the fraction of fish originally captured in the total population (200/n). Thus the total population can be estimated (n = 2000).”
That’s the idea anyway. The BMC study was obviously far more complicated than that. It took advantage of a unique Massachusetts database that afforded access to seven different datasets that could each identify cases of likely opioid use disorder. These included emergency room diagnostic codes, ambulance services data, death certificates, birth records (reflecting neonatal abstinence syndrome), addiction treatment facility admissions records, health insurance claims, and prescription drug monitoring data showing receipt of addiction-treatment drugs.
A startling 6.78 percent of all residents between ages 26 and 44 had opioid use disorder in 2015.
The study tracked data from 2011 to 2015. Prevalence rates rose sharply during that period, swelling from 2.72 percent in 2011 to 4.60 percent in 2015.
Prevalence was greatest for the age bracket 26 to 44, for which a startling 6.78 percent of all state residents had opioid use disorder in 2015.
The sharpest growth in prevalence, however, occurred among the young—ages 11 to 25. For that group, prevalence grew 76 percent from 2011 to 2012, and 42 percent from 2013 to 2015.
The study also found that prevalence was higher in rural counties than in urban ones. Rural Berkshire County had 6.06 percent overall prevalence, for instance, compared to 3.36 percent in urban Suffolk County, which includes Boston.
Overall, the researchers found that “fewer than 50 percent of people with OUD were ‘known’ to the health care system, a critical first step in the OUD treatment continuum.”
For the years of the study, it had already been known that opioid-related overdose fatalities in Massachusetts jumped 273 percent, from 676 to 1793 per year. But that surge had largely been attributed to the influx of highly lethal, illicit fentanyl into the market.
In light of the study results, however, the authors stress that “the high prevalence of OUD is probably also contributing to the stark increase in overdose deaths.”
Based on CDC overdose death statistics, Massachusetts was among the states hardest hit by the opioid epidemic. In 2015, its death rate was 25.7 per 100,000 population, compared to a national average of 16.3. In 2016 its rate was still worse: 33.0 per 100,000 compared to 19.8 nationally.
The CDC’s latest provisional death statistics provide some basis for optimism that drug and opioid overdose fatalities in both the US and Massachusetts may have crested. For the most recent 12-month period measured by the CDC—stretching from April 1, 2017 to March 31, 2018—Massachusetts had 2 percent fewer total drug overdose fatalities (including from nonopioids) than in the preceding 12-month period.
Still, the raw number is chilling: 2,419.