A large new study from Australia found that DSM-5 would cause a skyrocketing 60-percent increase in the rate of alcohol-use disorders. The authors are neutral on whether the consequences of this huge jump would be positive or negative. Their study was not designed to determine whether the new "problem drinkers" caught in the wide net cast by DSM-5 would benefit -- or whether they lack clinically significant impairment and might be harmed by misidentification and unnecessary stigma.
Some thoughts:
1) This kind of comparison between rates of diagnosis using DSM-5 versus DSM IV urgently needs to be done for every change suggested in DSM-5. Otherwise, its makers will have no way of judging the possibly enormous impact of DSM-5 in pushing the boundary of mental disorder deep into what has heretofore been considered normality. Note that this study on rates of alcohol disorder was performed independent of the work on DSM-5. Unaccountably (and irresponsibly), the DSM-5 field trials have altogether avoided studying impact on rates and will therefore not have this necessary information for any of its changes in other diagnoses.
DSM-5 will be flying blind to its impact.
2) If, as seems probable, the new DSM-5 proposals for other diagnoses encourage similar large jumps in diagnostic rates, the concept of psychiatric disorder will be trivialized beyond recognition. We already have a diagnostic system whose low thresholds pin a diagnostic label on 45 million Americans every year. The further watering down of definitional standards will make psychiatric diagnosis so ubiquitous as to be almost meaningless -- and divert scarce resources away from the patients with severe psychiatric disorders who really need them.
3) The obvious next step is to determine more about the risks and benefits of such a huge swing in diagnostic practice -- both to the individuals newly labeled as mentally disordered by DSM-5 and to the nation's health-care policy.
4) Decisions of such huge import to people and policy should not be left to a small group of narrowly focused experts. Experts are always biased to prefer lower thresholds so as to avoid missed diagnoses. They are consistently insensitive to the risks of overdiagnosis. I discussed this in more detail in a recent blog.
5) DSM-5 changes should not be made in a fast-draw, shoot-from-the-hip manner, without any study of their effect on rates and possibly profound negative consequences. The design of the DSM-5 field trials should be revised now, before it is too late, to determine how each proposed change will influence rates of disorder and to assess the risks of overdiagnosis.
Leave a comment