Changes to the diagnostic definition of autism will be published in the fifth edition of the "Diagnostic and Statistical Manual of Mental Disorders" - DSM V - but exactly what those changes will be is a key point of discussion. There are still a lot of qualifying issues in a lot of areas for a publication that has already been a long time in the making.

At stake? Apparently a lot of money.  Autism was once rare enough that a definition was not rigorous but it also was not crucial - some leeway was allowed.  As a result, recent increased instances, either to more occurrences or more accurate diagnoses or even mis-diagnosis, have made the new definition for DSM-V a hot topic.

Unlike DSM-IV, when the race was on to add diagnoses for everything, concern about DSM-V is in making things more scientific and less social interest or advocacy-based.  If 80 percent of people are something, for example, it isn't a psychological condition, and autism has been extrapolated so far that virtually anyone can be considered 'on the spectrum' or one of its derivatives if they want to be.  That's bad for people who actually have it because limited treatment resources are stretched beyond control.

But a push is also on to make sure cases are not lessened. Yale School of Medicine researchers at a meeting of the Icelandic Medical Association state that getting an autism diagnosis could be more difficult in 2013 when a revised diagnostic definition goes into effect - labels mean services and services mean money.

Whether or not the changes actually are bad is the debate.  Accuracy in diagnoses should be paramount (and are, to most) but autism is a booming industry.  Cutting those back is going to be a negative development to people working in that industry and to parents who are concerned.  Yale Child Study Center director Fred Volkmar, M.D., and colleagues Brian Reichow and James McPartland analyzed people (without intellectual disabilities) who were evaluated and determined to be autistic during the 1994 DSM-IV field trial and they estimated that about half those might not qualify for a diagnosis of autism under the proposed new definition. If so, was DSM-IV too loose or is DSM-V too strict?

Obviously, looking at old diagnoses, rather than actual people, and matching those to new proposed criteria and claiming they might not qualify is not entirely rigorous but autism was less faddish in 1994 so they may have a point.  If autism is being reconfigured by negotiation and lobbying rather than science, DMS-V will be basically useless for deciding what should be covered by insurance companies, which means their coverage will be arbitrary. Labels matter when it comes to services and people want services for free. It's impossible that those economic factors aren't being considered.

The cases that might not be covered now are not cognitively disabled people.  They may instead be like a lot of people who are casually diagnosed.  In science and math, they may be like a lot of people who suddenly find themselves being labeled autistic for not liking to drink or socialize. Basically, the problem DSM-IV ran into was that they wanted to call everything a disorder, which meant to the public and insurance companies it really meant nothing so coverage was inconsistent. 

What about analyses like this, that seek to compare diagnoses across eras and DSMs?  It becomes even trickier.  An autism diagnosis in DSM-III was not the same as IV and a IV will not be the same as V.  The status quo can't remain, of course, DSM-IV had too many flaws, but the social sciences have a hard time maintaining science rigor and most (Four Horsemen of the Questionable like Stapel, Hauser, Kanazawa, etc., aside) in the field want that to improve.  DSM-V is the most thoroughly negotiated document in psychology history, as it should be.   For its 2001 version, the United Nation Intergovernmental Panel on Climate Change (IPCC) threw out everyone who disagreed with working group leaders on previous versions and that was not good for climate science, as the resulting scandals and credibility dips showed - psychology is sticking to its guns and instead saying DSM-V is "ready when it's ready" and that is to be applauded.

Any change is going to be criticized and the American Psychiatric Association (APA) seems to be ready for that, but they haven't been at the forefront of a cultural storm before so we'll need to give the new criteria some time.