The Addiction Puzzle, Part 5: Bringing Scientific Rigor to Addiction Research

“Alcohol is the third leading cause of preventable death in the United States. We need high-quality science to address this problem. We don’t do belief-based medicine in cardiology, and we shouldn’t do it here.”

One of Richard Saitz’s most satisfying moments as a scientist came last June at the annual meeting of the Research Society on Alcoholism. Presenting results at the podium, a scientist finished his remarks by saying, “I think even Rich Saitz would agree that our conclusions are sound.” The audience chuckled and Saitz, sitting anonymously among them, beamed. “I’m usually the first person at the microphone grilling the researchers,” he says. “Even beyond my own work, I’m pushing people to improve their science.”

Saitz (CAS’87, MED’87) a School of Public Health professor and chair of community health sciences and a School of Medicine professor of medicine, relishes his role as a globe-trotting gadfly. In his 25 years of researching addiction, he has strived to push his field—and sometimes drag it, kicking and screaming—into the world of evidence-based medicine. His recent work has challenged two popular addiction interventions, to both the admiration and the dismay of colleagues.

“Rich has no problem calling it straight. And sometimes that bothers people,” says colleague Jeffrey Samet (SPH’92), a MED professor of internal medicine. “But he has engendered a lot of respect in the field because he is true to the data. So I think if there’s flak, he wears his flak jacket well.”

Saitz was drawn to addiction research in the late 1980s, during his medical residency at Boston City Hospital (now Boston Medical Center), where many patients were drug or alcohol dependent. While he and his colleagues treated patients’ acute conditions—liver failure, seizures, infections—they had frustratingly few tools to tackle the underlying addiction. “Somebody would come in with alcoholic pancreatitis, we would get them improved, and then we would discharge them. And then a few days later they would come back with alcoholic pancreatitis,” he recalls. “We did a lot for the medical condition, but it wasn’t clear to me what we were supposed to be doing for the cause of the problem, the alcohol use.”

Saitz found few scientific studies to guide him, partly because addiction had long been considered a moral or social problem rather than a treatable illness. So he tackled the questions himself, publishing his first influential paper, which refined the standard treatment for alcohol withdrawal, in the Journal of the American Medical Association (JAMA) in 1994. He chose a rigorous scientific design for this study—a randomized, double-blind, controlled trial—even though such studies are notoriously difficult in behavioral science. This scientific rigor has become a hallmark of his work. “Alcohol is the third leading cause of preventable death in the United States,” he says. “We need high-quality science to address this problem. We don’t do belief-based medicine in cardiology, and we shouldn’t do it here.”

Challenging popular beliefs

Recently, two of Saitz’s studies have challenged some popular beliefs in addiction treatment. The first involves a practice called Screening, Brief Intervention, Referral, and Treatment (SBIRT). The premise is simple: in a primary care setting, during an annual physical, say, the doctor asks if a patient has had five or more drinks at one sitting in the past year (for women, it’s four). If they answer yes, the doctor asks a few more questions, counsels the patient for 10 to 15 minutes, and sometimes refers him or her for more treatment. In this setting, for identifying healthy people with risky alcohol use, SBIRT has been proven to work.

Since SBIRT worked so well, in 2003 the federal government instituted a massive, ongoing program to implement it in many different health care settings (like hospital wards and emergency rooms) to screen people for risky drug use and other problems. To Saitz, the federal program seems a huge leap of faith, and it drives him batty. “It strikes me as very bizarre,” he says. “Behavioral interventions are affected by their context, and they’re just really different depending on where and when and how you do them.”

But Saitz isn’t just a critic; he’s a scientist. So, armed with funding from the National Institute on Drug Abuse, he set out to test whether SBIRT could actually lower risky drug use in a primary care setting, where it’s supposed to work best. In the study, researchers randomly assigned 528 drug users to either a standard brief intervention, a stronger intervention with a trained psychologist, or nothing at all. Six months later, the researchers followed up. Saitz has not yet published the results, but first presented his findings at the Society of General Internal Medicine’s 2013 meeting in Denver. “It was negative, negative, negative,” says Saitz. “Overall, people didn’t decrease their drug use six months after entering the study. In all three groups, none of them did.”

Addiction expert Bertha Madras, a Harvard Medical School professor of psychobiology, calls Saitz’s data “impeccable,” but argues that the study focused on a population with additional risk factors, like poor family support, who were less likely to respond to SBIRT. Madras, who was the deputy director for demand reduction for the White House Office of National Drug Control Policy from 2006 to 2008, points to her analysis of the federal government’s SBIRT program for comparison. Her study, which had no control group, accounted for 459,599 people screened in many health care settings across six states, finding that among low-level drug users who received SBIRT, the rates of drug use dropped 67.7 percent after six months.

To Saitz, these results provide no evidence that SBIRT was responsible for the reduction in drug use. But he acknowledges that a more intensive or repeated form of SBIRT might work to identify and treat risky drug users, even those with compounding problems. But since such an approach is not yet proven, he argues against an expensive blanket approach that applies SBIRT to all comers.

A no-brainer, but startling results

Saitz’s second study is more disconcerting. Published in JAMA in September 2013, and funded by grants from the National Institute on Alcohol Abuse and Alcoholism and the National Institute on Drug Abuse, it asked whether chronic care management (CCM), a comprehensive, multidisciplinary approach to care, might be useful for treating alcohol dependence. In CCM, a patient with a chronic disease like diabetes might be assigned a nurse who visits the patient at home, teaches him to monitor his glucose, arranges visits to specialists, and generally helps manage the disease. “It just seems like good medicine,” says Saitz, and indeed, one cornerstone of health care reform is the idea that CCM, while expensive in the short term, will lead to lower costs overall and better health for patients.

Saitz thought that CCM, which has been proven to work for conditions like diabetes and asthma, could probably help patients addicted to drugs or alcohol, too. His team started with 563 people who were heavy users of alcohol, cocaine, or opioids, and assigned them randomly to one of two groups. One group got the “usual” care: phone numbers to treatment centers and an appointment with a primary care clinician. The other group got the blue-ribbon treatment: a team including an internist, a psychiatrist, a nurse, and a social worker, all focusing specifically on their addictions.

The study seemed like a no-brainer, so the results were startling. On the one hand, there was good news: the study had a surprisingly low dropout rate, and both groups showed marked improvement in many risky behaviors. Heroin use, for instance, dropped from 100 percent overall to just over 40 percent. “What’s shocking to me is how well it worked,” says Madras. “It was phenomenal.”

Despite this success, the overall results remain puzzling. Saitz had expected the CCM group to fare much better than patients receiving usual care. Instead, it doesn’t seem like the CCM blue-ribbon care worked any better than a phone number and a follow-up appointment. “We thought we nailed it,” says Saitz. “But we found no difference in abstinence. We found no difference in health care utilization. We found no differences in anything.”

He concluded that CCM worked no better than usual care, at least for heavy users, but critics caution against dismissing CCM out of hand. “Rich says the study was a total, abject, abysmal failure, but I disagree with his bleak interpretation,” says Thomas McLellan, CEO of the Treatment Research Institute, a nonprofit research and policy organization. Many of Saitz’s test subjects were heavy users recruited from detox centers, with low levels of education and few social supports—not the average primary care population. To McLellan and other critics, the study results are perplexing, but pave the way for more research. “These are rapidly changing times, and the guys at BU are right on the cutting edge,” he says.

Saitz still believes that CCM could help manage drug and alcohol addiction, but may need to target people who are less dependent and more motivated to change. He plans to continue research into CCM in the future, perhaps focusing on groups most likely to benefit. But that raises a bigger question: how do we treat the most serious addicts? If CCM can’t help them, what can? That’s a thorny problem, the kind that pushes Saitz ahead.

“There is complexity in this,” he says. “But that’s no excuse for not getting it right.”

Barbara Moran (COM’96) is a science writer in Brookline, Mass. She can be reached through her website WrittenByBarbaraMoran.com.

This Series

The Addiction Puzzle, Part 5: Bringing Scientific Rigor to Addiction Research

Also in

The Addiction Puzzle

• November 18, 2013

  The Addiction Puzzle, Part 1: An Overdose Lifeline

• November 19, 2013

  The Addiction Puzzle, Part 2: Could ADHD Meds Promote Future Cocaine Use?

• November 20, 2013

  The Addiction Puzzle, Part 3: The Search for the Alcohol Trigger

Series home