Advice to DSM-V . . . Change Deadlines and Text, Keep Criteria Stable

Advice to DSM-V . . . Change Deadlines and Text, Keep Criteria Stable

There is no magic moment when it becomes clear the world needs a new DSM. The publication dates of previous DSMs were determined by revision dates of the International Classification of Diseases (ICD). Thus, DSM-I appeared with ICD-6 in 1952; DSM-II with ICD-8 in 1968; DSM-III with ICD-9 in 1980; and DSM-IV with ICD-10 in 1994.

The projected publication date for DSM-V (May 2012) was chosen to be consistent with an earlier expectation that ICD-11 would be published that year—but recently announced delays in ICD-11 will postpone its arrival until at least 2014. Under normal circumstances, DSM-V and ICD-11 should be published simultaneously. But there is another problem caused by a coding change going into effect in October 2013. ICD-9-CM—which is now the official method of diagnostic coding used to specify all medical encounters in the United States—will then be replaced by a completely revamped ICD-10-CM. Publishing DSM-V much before or after October 2013 would result in great confusion. By far the most convenient date for DSM-V would be just before October 2013.

Even were this not the case, the extra time is needed to ensure that DSM-V will not cause serious unintended consequences. The fatal flaw in the current work on DSM-V is its intention to conduct field trials without having first posted for review the specific wordings of all the options being considered. The most important step in the development of any DSM is the first draft—presenting all the criteria sets in a systematic form. Only the Task Force working as a whole can discipline and reconcile the often inconsistent outputs produced by the different work groups. It is a very reliable rule of thumb that work groups are always more willing to make changes than is desirable. Experts in any given area tend to have their pet ideas and to worry more about missed cases than about creating potential false-positives. By ruthlessly applying the necessary rule of empirical documentation, the Task Force must provide a useful check on work group enthusiasm.

Once a complete first draft of DSM-V is posted, it will take months for the field at large to critique the explicit wordings of all DSM-V suggestions. It will then take more months for the work groups to incorporate the comments into revised criteria sets. Finally, it will take the Task Force months to reconcile the inconsistencies in the work group revisions.

No field testing should begin until this process is completed. Field testing is itself a time-consuming process. First, the methodology should be posted, reviewed, and revised based on suggestions received. Next, sites must be recruited. The obvious time-saving temptation is to use samples of convenience, but these will almost certainly not generalize well to the actual environments in which DSM-V will be used. Human participants’ approval will almost certainly cause unpredictable delays. Raters and administrative staff must be trained; data management systems installed; patients recruited; data cleaned and analyzed; and results posted, reviewed, interpreted, and incorporated into the evolving drafts of DSM-V. Then the revised drafts must be reviewed again by the field and by the Task Force.

Finally comes the very time-consuming task of writing the text and having it reviewed. Then DSM-V must be approved through the APA’s governance structure. All this takes time—and the schedule is far from predictable.

The secrecy surrounding DSM-V prevents us from knowing the precise state of its current development and from understanding the rationale for beginning field trials now. However, it is a fair guess that the reason no first draft of DSM-V has yet been posted is that the criteria sets are still far from being presentable to the field, even though field trials were meant to begin now. We can only assume that the crucial step of vetting criteria sets before doing field trials is being skipped to allow the Task Force to save time to meet the arbitrary and now clearly very inconvenient publication deadline of May 2012.

If anything like proper care is taken to accomplish the steps listed above in their proper order, DSM-V cannot possibly be ready for publication by May 2012. Even if all the many previous problems in the DSM-V methodology were immediately corrected now and every future step of the revision were done with perfect efficiency, the publication date would probably have to be delayed at least until October 2013—far more convenient because DSM-V can then be coordinated with the introduction of the new ICD-10-CM codes. Of course, it is possible that the careful preparation of DSM-V may take even longer.

Why not do the obviously right thing and substitute a postponed or flexible date of publication to ensure there is adequate time to avoid the many sticky problems that a rushed DSM-V will cause? Publishing profits are the only possible driver of a fixed and implacable 2012 publication deadline, and this is obviously not acceptable.

The DSM-V Task Force and work group members are dedicated people doing their best under difficult circumstances. They need sufficient time to ensure that DSM-V will be a worthwhile contribution—and not the cause of enduring problems for the field and for our patients.

Keep criteria sets stable

Criteria sets should not be changed except when there is a great reason. The most important criteria sets have a long track record with substantially constant wording since the publication of DSM-III in 1980. They have survived the test of time and have become familiar to generations of clinicians and educators. Any arbitrary changes will be needlessly disruptive.


Loading comments...
Please Wait 20 seconds or click here to close