DSM-5 Field Trials (Part 1)-Missed Deadlines Have Troubling Consequences

[Editor's note: Please click here to read Part 2 of this blog.]

This is a sad tale of completely unrealistic timetables, poorly executed work effort, consistently missed deadlines, and what will undoubtedly be a rushed and botched DSM-5. It all started at the annual meeting of the American Psychiatric Association in May 2009, when the DSM-5 leadership blithely announced it was ready to begin field testing in the early summer of 2009.

This was shocking news. The Work Group efforts were still at a primitive stage, not nearly ready to post for public comment-much less to test formally. The DSM-5 confidentiality agreements had cloaked all the preliminary discussions in such secrecy that the field had no idea what was being considered and no chance to provide corrective input. There had also been no public discussion of the appropriate goals and proper design of field testing, and no sites had been recruited.

In short, it was pure fantasy that field trials could begin in just 2 short months. Why the complete disconnect between the planned and the possible? The schedule for field testing obviously had nothing to do with the work that had (not) been accomplished and instead was targeted to what had become the equally impossible publication deadline of May 2012. DSM-5 had slipped so badly behind schedule that the leadership was forced to try to play catch-up by skipping the necessary steps of:

1.Posting draft suggestions to encourage public input
2.Identifying mistakes and inconsistencies
3.Honing and refining the draft options so that they would be in a clean, final form needing only confirmation through field testing. Instead, DSM-5 would be testing the very roughest of drafts and hoping to somehow modify them on the fly- -this is fully equivalent to attempting to fix a broken aircraft engine while the plane is in the air.

Robert Spitzer and I were alarmed by this recipe for disaster and wrote a letter to the APA Trustees on July 9 2009, warning them of the need to post and refine the proposed drafts before field testing them. We recommended that the field trials and the publication of DSM-5 both be delayed for 1 year to allow the process to catch up on these crucial missing steps. The Trustees fiddled as usual, but by the winter of 2009, it became clear even to them that DSM-5 was far too behind its schedule to meet the planned 2012 publication date. They were forced to announce the 1-year delays.

The extra year was badly used, mostly wasted. The DSM-5 drafts were painfully and mysteriously slow in coming-announced first for October 2009, then for December, then for January; they finally appeared about a year behind schedule in February 2010. And, after all this dithering, they were still a mess. With only a few exceptions, the posted criteria sets were written imprecisely, had many inconsistencies across disorders, and were far from ready for field testing. The public response to the postings was vigorous and should have led to major changes in the criteria sets, before any attempt to test them. But, for absolutely unaccountable reasons, only a few of the revisions of the criteria sets have been posted and most of the wordings are trapped in the usual DSM-5 black hole of secrecy.

At the May, 2010 APA meeting, history eerily repeated itself. The DSM-5 leadership again blithely announced that field trials were about to begin-this time in the early summer of 2010. The suggested field trial design is extravagantly and fatally flawed in ways to be addressed in later blogs. Here we will focus only on the subsequent missed deadlines that clearly predict the inevitability of future continuing delays and a final rush to what will be an extraordinarily sloppy DSM-5.

It was patently obvious from the moment of its announcements that the new DSM-5 field test timetable was also a product of fantasy that would not be met in the real world. First off, it should have been clear that the field trials could not possibly start on time 2 months after their announcement. Recruiting the sites, training the personnel, gaining human rights approvals, and pilot testing always take at least 6 months. Predictably, we are already in mid November 2010 and it is still not at all clear when the DSM-5 field tests will actually begin to enroll patients at all its sites.Then there is the design. Forget for the moment that it asks the wrong questions and will produce largely irrelevant answers. Forget that it is testing poorly written criteria sets that are in much too rough a form to be ready for testing. Again, our focus here is only on timetables and missed deadlines. The DSM-5 field trials are a masterpiece of cumbersome complexity-an administrative nightmare. They were originally scheduled to last 9 months starting July 2010 and ending in March 2011. Instead, the project will probably not start in full force until December 2010 or January 2011 (or later). By my reckoning (based on the experience with the DSM-IV field trials), it will take at least a year to complete from the date of first patient entry. And this assumes a maximum possible efficiency that is not at all likely given all the past laggard DSM-5 performances.

This phase of the DSM-5 field trials will probably be delivered at least 9 months late (really, 21 months late if you begin the clock with the original 2009 announcement). And again, these may be extremely optimistic projections because they are based on the risky assumption that there will not be continuing foul-ups and delays in executing what is an unusual and unnecessarily complex field trial design.

Because we are reaching a point of no return, these accumulating delays spell future disaster for DSM-5. The future schedule provides little room for error or forgivingness. The DSM-5 publication date of May 2013 is fixed in stone-both because the new ICD-10-CM codes will become official in October 2013 and because the APA budget depends on DSM-5 publishing profits. But the work to be done is enormous, not really do-able in the remaining time allotted. The result will be a rushed and jumbled DSM-5 that will create huge problems for our field and for our patients.

The only possible save at this point is a radical midterm correction that would: 1) abandon the many radical new diagnoses suggested for DSM-5; 2) revise and simplify the field trial design; and, 3) focus the remaining time on clean writing and eliminating mistakes. The first and obvious step would be to end the secrecy and post all the most current versions of the DSM-5 criteria sets.

Will any of this happen? Common sense says it must. Experience says it won't. DSM-5 will try to muddle through, as it has. The APA Trustees will look on haplessly from the sidelines. Eventually, DSM-5 will be a rushed patch-work. The only hope for a usable DSM-5 is for the Trustees to exert their authority to correct an errant process. But they will act only if there is mounting outside pressure and widespread public concern. There is still time to save DSM-5, but it is being wasted and is quickly running out.