High Wire is Maia Szalavitz's reported opinion column on drugs and drug policy.
Last week, two prestigious think tanks released recommendations related to the opioid crisis. One came down in favor of piloting a highly controversial harm reduction treatment—prescribing heroin for opioid addiction—based on a careful review of the data. The other reached the opposite conclusion on harm reduction, claiming that syringe exchange programs and distributing naloxone to reverse overdose actually make matters worse—while ignoring decades of public health data to the contrary.
How did the Brookings Institution get such radically different results from the Rand Corporation? The answer lies in a clash between economics and medical and public health research, which has infuriated addiction experts to the point of demanding a retraction from Brookings. With overdose deaths at an all-time high and thousands of lives at stake, getting these recommendations right matters. Politicians, policy makers, and their staff often rely on summaries of these reports to make decisions, rarely digging into the details themselves.
(As of this writing, Brookings has merely amended the article to acknowledge that the previously unmentioned public health research exists, but this has not satisfied its critics).
Brookings’ article, titled “Research roundup: What does the evidence say about how to fight the opioid epidemic?” immediately angered experts, who took to Twitter to complain that its conclusions did not match those of the researchers and physicians who actually study and treat people with addiction. It has taken decades for harm reduction to move from being a fringe idea into an essential part of drug policy, most of which were spent using empirical evidence to fight against the idea that anything other than making the consequences of drug use worse would increase it.
The Brookings authors—Jennifer Doleac, associate professor of economics at Texas A&M University, Anita Mukherjee, assistant professor of risk and insurance at the University of Wisconsin, and Molly Schnell, a postdoc at the Stanford Institute for Economic Policy Research—are all economists. And their roundup was almost comically focused on the economic literature.
Watch More From VICE News:
Based on two studies—one a working paper and the other not yet published or available for review—they indict harm reduction via needle exchange and naloxone, which literally has a literature that includes thousands of studies. On needle exchange, they claim:
New work by Packham and Wells (2018) suggests that syringe exchange programs—a staple of harm-reduction efforts—reduce HIV rates as intended but unintentionally increase opioid-related mortality by making it easier, cheaper, and safer to use heroin.
In the original version of their post, the authors didn’t mention that virtually all public health authorities—from the US to China and Iran—support needle exchange based on reviews by the World Health Organization, the Centers for Disease Control and Prevention, the Institute of Medicine (now the National Academies of Science, Engineering and Medicine), a Cochrane review, and many others.
The support comes from data cited in those reviews showing that needle exchange reduces injection drug use and makes people more likely to seek addiction treatment—both of which lead to fewer overdoses, not more.
In the updated version, they added a paragraph noting the discrepancy while still failing to cite the data and the weight of the consensus that needle exchange doesn’t do harm.
The results of the Doleac and Mukherjee study suggest that broadening access to naloxone can unintentionally increase opioid abuse by reducing the cost of overdose. That is, while naloxone saves lives in the moment—which is clearly a priority for policymakers and public health officials—it may increase rates of drug abuse and possibly even mortality.
Once again, actual research on the ground contradicts these claims and remains unmentioned. For example, a study on naloxone programs by Alexander Walley and colleagues found that in Massachusetts, areas with low levels of naloxone availability and training reduced overdose death by 27 percent and those with high levels cut it by nearly half, compared to places that did not implement naloxone programs.
“What’s interesting is that they appear not to read anything that’s not published by an economist,” says Richard Frank, professor of health economics at Harvard Medical School. He notes that the Doleac naloxone study measured availability of the medication indirectly, by assuming that laws passed to increase access actually do that in a timely fashion and by looking at the number of searches for naloxone on Google and the number of drug arrests.
But when Frank and his colleagues used a more direct measure—comparing the results of expanding Medicaid and legal changes on actual naloxone sales—they found that, as the public health data suggests, more naloxone leads to fewer overdose deaths, not more. (His study is also not yet published in a peer-reviewed journal, but some of the data was published on the website of Health Affairs, in an article raising concern about Doleac’s naloxone research.)
Doleac would not respond to specific questions I sent about the Brookings article and her naloxone work, but released a statement to several journalists who inquired, saying:
“I find it utterly disheartening that a discipline as important as public health is filled with so many people who collectively have so little understanding of rigorous research methods. Advocates should acknowledge that many of their strongly-held priors are not evidence-based. Anecdotes and personal experience are valuable but are not a substitute for rigorous causal inference methods.”
Doleac’s contention appears to be that the only data that counts is economic research that uses mathematical techniques to tease out whether harm reduction tactics cause or reduce harm—or randomized controlled trials. This rejects nearly all of public health—tens of thousands of studies that compare communities or individuals before and after an intervention, using controls to manage pre-existing differences and other observational methods.
And the reason why randomized controlled trials haven’t been conducted on needle exchange and naloxone is the same one as why parachutes have never been tested against simply jumping out of planes at altitude. It would be unethical for the control group to be randomized to a potentially deadly study condition like not getting clean needles or access to a drug proven to reduce overdose.
“The dispute is primarily disciplinary—although it is taking on metaphysical proportions given the inability or unwillingness of Doleac and colleagues to acknowledge the validity of other methodologies or to reflect constructively on the weakness of their own methods,” says Ricky Bluthenthal, professor of preventive medicine at the University of Southern California. “Ignoring findings from other methodological approaches is uncommon in public health. We are multidisciplinary by necessity and training.”
Or, to paraphrase Donald Rumsfeld, you have to use the data you have, not the data you might want or wish you had.
When advising on policies that affect the lives of vulnerable people, understanding the science across disciplines and the political environment is essential. And that’s where The Rand Corporation’s research on heroin prescribing, also released last week, stands out.
Heroin by prescription is, to put it mildly, a tough sell for the public and policy makers. For people raised on DARE and the war on drugs, it seems like “enabling” and just giving up on helping.
But in fact, most of damage and destruction associated with heroin addiction is linked to its illegality—and the research on heroin shows clearly that when you remove the danger and drama of chasing down the drug and dodging the cops, people’s lives tend to become much more calm, productive, and caring.
Without having to raise money for drugs (often through crime) and without having to spend hours waiting for dealers and then working to hide their drug use, people on heroin prescriptions have time to think. The research shows that this means they’re more likely to get a job, repair family relationships, and, yes, get into traditional treatment and recovery than if they’re simply left to risk their lives on the street.
To explain this to policymakers, Rand published four reports laying out the research on heroin and another harm-reduction approach, safe injection facilities (SIFs), which are rooms where people can consume drugs under medical supervision. Rand found that the data supporting heroin prescribing was far stronger and that, while safe injection sites are promising, it would be hard to create enough facilities to make a significant difference.
No fewer than ten randomized controlled trials have been conducted on prescription heroin. (Curiously, this approach wasn’t even mentioned in the Brookings piece). And as far back as 2011, a Cochrane Review, considered the gold standard for medical evidence, was published favoring heroin prescribing. That review, which covered trials including more than 2,000 patients in several countries, concluded that:
The available evidence suggests an added value of heroin prescribed alongside flexible doses of methadone for long-term, treatment refractory, opioid users, to reach a decrease in the use of illicit substances, involvement in criminal activity and incarceration, a possible reduction in mortality; and an increase in retention in treatment.
In other words, providing heroin as an addition to methadone cuts crime, reduces drug use, and helps people turn their lives around. Over the years, I’ve interviewed a number of participants in heroin treatment and the first thing most of them said to me was that it saved their lives.
The research does caution that heroin treatment must be more carefully supervised than current medications like buprenorphine and methadone because it is riskier. As Beau Kilmer, co-director of Rand’s drug policy research center, put it in a press release, “Given the increasing number of deaths associated with fentanyl and successful use of heroin-assisted treatment abroad, the US should pilot and study this approach in some cities,” adding, “This is not a silver bullet or first-line treatment. But there is evidence that it helps stabilize the lives of some people who use heroin.”
But Rand went even further than just reviewing the literature (and reviewing the reviews of the literature). It also conducted interviews with 150 community members who might be affected by heroin prescribing, including people who take opioids, treatment providers, and law enforcement. Not surprisingly, many expressed fear that heroin prescribing would “enable” continued drug use without helping—and the researchers concluded that educating communities about the research showing that it does not prolong addiction would be important.
If think tanks are to do their job and inform policy makers of the relevant data, they can’t simply ignore cross-disciplinary research, even if they disclose that they have done so. Drug policy can be a matter of life and death—so extreme care must be taken. Or as Frank put it, “I think a lot of humility is called for and I think that’s what’s missing here.”
Update 12/14/18: A spokesperson for the Brookings Institution provided the following response:
"Brookings supports the academic freedom of our resident and nonresident scholars, including their right to publish controversial research that adheres to our quality and independence standards. However, Brookings as an institution does not take positions on issues, nor do we endorse Doleac’s response to the criticism and feedback she received. Public health experts, researchers, and advocates have made important and critical contributions to our society, and their efforts have saved countless lives by informing America’s response to the opioid crisis."
Sign up for our newsletter to get the best of Tonic delivered to your inbox.