Transitioning to Updated Substance Use Disorder Criteria in DSM-5

Wilson M. Compton, MD, MPE
Director, Division of Epidemiology, Services and Prevention Research; National Institute on Drug Abuse, Bethesda, Md.

First published in Psychiatry Weekly, September 2013, 8(19).


The recently published DSM-5 introduces several key changes to the section on substance use disorders (SUD). Substance dependence and substance abuse, which were treated as discrete concepts in DSM-IV, are united in DSM-5 to create a single category of SUD, easing the way for the new section’s incorporation of dimensional assessments. DSM-5 also added “craving” as a formal diagnostic symptom, whereas DSM-IV had only recognized craving as an ancillary symptom often observed in clinical practice.


Dr. Wilson Compton, who recently published a study comparing the compatibility of the old and new sets of SUD diagnostic criteria, says it’s quite clear from the literature that addictive disorders exist on a spectrum.


“So instead of having the labels of abuse and dependence, we now use ‘mild, moderate, and severe’ to suggest the level of severity of the disorder,” says Dr. Compton. “That approximates the nature and reality of SUD better than what has proven to be an artificial and somewhat ill-advised distinction between abuse and dependence.”


By the addition of craving to DSM-5 SUD criteria comes a construct long validated by clinical neuroscience researchers. The question is: does this particular criterion provide helpful information? According to Dr. Compton, there is a modest difference in some substances between SUD symptom clusters with and without craving as a criterion. The criterion was accepted largely because of its clear neuroscience implications, making it a ready target for further treatment research.


Strictly speaking, the diagnostic criteria for SUD overlap greatly between DSM-IV and DSM-5. Change is still change, however, and Dr. Compton and colleagues undertook a study to see how well the SUD category of DSM-IV would apply to the new criteria set, and how to make the most effective crosswalk between DSM-IV and DSM-5.

Analysis, Methodology

The first key part of Dr. Compton’s analysis was to find a broad-based sample of SUD from the general population that covered the full range of severity in both the clinical and general populations. Most importantly, the sample needed to include the only recently formalized criterion of craving, in addition to all other DSM-IV diagnostic criteria for SUD. The chosen sample derived from the National Longitudinal Alcohol Epidemiologic Survey, conducted in 1991–1992 (n=42,862).


“The next key part of our methodology was to determine the likelihood of subjects diagnosed with a DSM-5 SUD being captured by the DSM-IV criteria for dependence,” says Dr. Compton. “Because DSM-IV required at least 3 criteria, we started with a base of at least 3 criteria from DSM-5, and then we looked at how closely the criteria agree between the two editions when ?3, ?4, ?5, or ?6 DSM-5 criteria are endorsed. In other words, what is the threshold that creates the greatest agreement between DSM-IV and DSM-5.”


Dr. Compton and colleagues looked specifically at subjects with an opioid, cocaine, cannabis, or alcohol SUD.


“The agreement across all of them was pretty similar,” says Dr. Compton. “For instance, we showed that there was a very great overlap between the DSM-IV and DSM-5 approaches to opioids at every criterion threshold, but that the best overlap occurred when we had ?4 criteria from the DSM-5 system. So if you diagnose an opioid SUD using ?4 criteria from the DSM-5 system, you have a high likelihood of meeting criteria for DSM-IV dependence.”


Alcohol and cocaine SUD, too, were in best agreement when ?4 DSM-5 criteria were endorsed. There was less agreement, however, for cannabis, which required ?6 DSM-5 criteria for the best agreement with DSM-IV. This, Dr. Compton says, is probably because craving was added as a new criterion for SUD in general, and especially because cannabis withdrawal symptoms are newly allowed by DSM-5 as a consideration when determining whether a severe cannabis use disorder is present.


Dr. Compton, who participated in The DSM-5 Substance-Related Disorders Work Group, says it is worth noting that the work group did not eliminate the building blocks of this particular diagnostic category, and, in some ways, it simplified diagnosis. “Those who trained in earlier editions of the DSM should find the adjustment to the fifth edition quite comfortable,” he says. “Our results show that, when diagnosing SUD with DSM-5 criteria, there is a high likelihood that an individual would meet DSM-IV criteria dependence, minimizing missed cases and the incidence of false-positives.”


Disclosure: Dr. Compton has stock holdings (less than $10,000) in 3M Corporation, General Electric, and Pfizer Inc.


Compton WM, Dawson DA, Goldstein RB, Grant BF. Crosswalk between DSM-IV dependence and DSM-5 substance use disorders for opioids, cannabis, cocaine and alcohol. Drug Alcohol Depend. 2013;132:387-390.