DSM-5: The End of One-Size-Fits-All Addiction Treatment?
When the dust of debate settles, the new revision's benefits should be clear: good science, better diagnoses, more individualized care.
Sometime this month, the DSM-5 will replace the DSM-IV as the coin of the realm for diagnosis of mental illnesses, including substance use disorders. Despite the unprecedented criticism that has accompanied the process, the final product’s changes are based on very solid epidemiological research, and they are likely to reduce ambiguity and confusion. But there may be some surprise, too, as received wisdom about the diagnosis and treatment of addiction is turned on its head. Let’s hope that this development will result in a more rational and nuanced approach to addiction.
When the DSM-IV was developed, it appeared that abuse and dependence were two distinct disorders. Substance abuse was defined according to four criteria; dependence, according to seven criteria. In practice, “abuse” was often used to denote a milder form of Substance Use Disorder (SUD); “dependence,” a more severe SUD.
In the case of opioids, “dependence” was confusing because almost anyone on opioid-based painkillers for any length of time develops physiological dependence (they will have withdrawal if they stop suddenly), whereas in the DSM-IV, “dependence” meant “addiction” (pathological, compulsive, harmful use). So pain patients prescribed opioids were mislabeled as opioid “dependent” even though they took their medication as prescribed.
Since then, a considerable body of research has shown that there are not two distinct types of substance misuse, but only one. More important, most DSM-IV “abuse” symptoms develop only in people with severe addiction, while “dependence” symptoms are among the earliest to develop. In the DSM-5, "abuse" and "dependence" are gone. In their place is the single "Substance Use Disorder."
With alcohol, for example, the earliest and most common problems are “internal” problems, such as going over limits, persistent desire to quit or cut down, and use despite hangover or nausea. The only “abuse” criterion that develops early is drinking and driving, but without a DUI. In the largest study of its kind, the NIAAA Epidemiological Study of Alcohol and Related Conditions (NESARC), 90 percent of people who met criteria for DSM-IV alcohol abuse—but not dependence—did so because of admitting drinking and driving. All other abuse criteria only occurred in people with the most severe and chronic addiction, and then late in the game.
In fact, legal problems occur so infrequently that this criterion was dropped from the DSM-5. This may come as a surprise to people working in the treatment industry because legal problems are the most common reason people seek treatment in rehab. But only about 12 percent of people with DSM-IV alcohol dependence ever seek specialty treatment, which suggests that the rest—who are not in treatment—have less severe disorders. People in rehab or AA are to alcohol use disorder what asthmatics on a ventilator in the ICU are to people with asthma: the most severe, treatment-refractory disorders as well as the most co-morbid psychiatric and medical problems. We’ve made a large error by assuming that everyone in the community who meets the criteria for a substance disorder has exactly the same disease as people in rehab or AA.
Addiction is not a well-defined term; it means different things to different people. Many people tend to apply the term only to the severe end of the spectrum.
NESARC and other studies have also demonstrated that SUD exists along a continuum of severity ranging from mild to moderate to severe. For example, almost three-quarters of people who develop DSM-IV alcohol dependence have a single episode, averaging about three or four years—and it never recurs. In other words, most addiction is not chronic or progressive. It is for many people in rehab and AA, but not for the rest.
The new SUD disorder is dimensional, meaning that the larger the number of criteria met, the more severe is the disorder and the associated dysfunction. For all DSM-5 disorders there is a range denoting severity: mild (two criteria), moderate (four criteria), and severe (six or more criteria). One of the challenges of this approach is that when you chart number of criteria by adverse consequences, you get a straight line. There are no clear “break points”—that is, no big jumps in dysfunction—as a result of an additional criterion or two.
In the absence of break points, decisions about categories such as mild or severe are somewhat arbitrary. The important clinical point, though, is that more criteria mean greater severity. Guided by this critical point, we can provide treatment according to severity, or stage, of illness, rather the give the same treatment to everyone with a diagnosis.
For example, the COMBINE trial, which was the largest clinical trial of treatments for alcohol dependence, had two groups: one drank an average of six to eight drinks per occasion, and met three or four out of seven criteria for dependence; the other group average 17 to 24 drinks per occasion, and met all seven dependence criteria. Yet everyone was treated as if they had exactly the same disorder. Several studies have demonstrated that there are multiple different illness trajectories after treatment. This suggests that people with different degrees of severity and treatment response should receive treatment specifically tailored to their needs.
Most people with SUD have mild or moderate forms, and most of it goes away without professional treatment. The proportions of mild, moderate, and severe vary by substance. For example, most heavy drinkers do not have alcohol use disorder but are at increased risk for it, and most disorders that develop are mild or moderate. On the other hand, there are fewer people with mild cocaine or heroin use disorders. Some of this is socially driven, in that the latter two substances are illegal in the U.S., but they are also more addictive and toxic than alcohol. Relatively few cannabis users develop a disorder, and if they do it tends to be mild or moderate.
What about “addiction”? This gets pretty muddy. Does everyone who drinks and drives more than once, and who goes over their self-imposed drinking limits a few times a year, have alcohol “addiction”? Do only people with severe disorders and physical withdrawal have “addiction”? Addiction is not a well-defined term; it means different things to different people. Many people tend to apply the term only to the severe end of the spectrum.
I view addiction along a continuum, including very mild forms of addiction. My advice is to abandon the clinical use of "addiction" in favor of “disorder.” But I suspect this ambiguity will cause a lot of confusion. “Dependence” and “abuse” are going to hang around too, simply because people are so used to using "substance abuse" as an all-inclusive term.
In sum, the DSM-5 is an advance that can help the field move beyond the “one-size-fits-all” approach so prevalent today. It is based on very solid science, and it works well clinically. Over the last 30 years an international consensus has developed concerning the core symptoms of substance use disorders. I doubt that a DSM-6 will look a lot different from the DSM-5, at least in terms of the clinical phenomenology of SUDs. Fortunately, however, we likely have a decade before that revision, and its attendant controversies, begins.