Should the DSM-V include a diagnostic category for psychosis risk
It is one of the most hotly debated issues in the development of the next
version of the diagnostic manual. On the one hand are those who point to a
robust prevention science that has identified risk factors highly predictive
of psychosis and who say that early intervention before the first psychotic
break can dramatically alter the long-term course for at-risk individuals.
On the other hand are those who point to the risk of stigmatizing
adolescents—the age group that would typically be receiving a diagnosis
of risk syndrome for psychosis—and who say that the putative risk
behaviors are too widely distributed among the general population to reliably
distinguish them from normal behavior. They also note the possibility of
exploitation by the drug industry eager to open a wide new market for
Allen Frances, M.D., has included the point in criticisms he has leveled in
Psychiatric Times against the DSM-V Task Force, saying that
the science of prevention is not advanced enough to warrant a new diagnostic
"No set of descriptive criteria can be nearly specific enough to
avoid a huge false-positive problem—particularly once the drug companies
sink their teeth into it," Frances told Psychiatric News."
Prevention is great if you have a specific test, but in our field this
will almost certainly have to be a biological test, because the descriptive
items are nonspecific and widely distributed in the general population.
Discussing this as an addition to the appendix would have been useful. Going
beyond to suggest that prodromal diagnosis is ready for prime time is a
misleading overselling of our current capacity to predict the
Leaders and members of the DSM-V Psychosis Work Group say the
inclusion of a subsyndromal category for pre-psychosis is by no means certain.
The topic was also the focus of a Web seminar conducted by Psychosis Work
Group Chair William Carpenter, M.D.
"In general medicine there is a tradition of treating at-risk
patients," Carpenter told Psychiatric News. "You don't
have to have cardiovascular disease to receive a statin drug, and it's
legitimate for a doctor to code for a risk category. Whether we are in a
position to anticipate that in psychiatry is debatable.
"We know we can identify people at risk for psychosis, and we know
that if we follow them for a year or two, 20 percent to 35 percent will
convert to one of the psychotic diagnoses," Carpenter said. "There
are legitimate concerns about stigmatizing and excessive prescription of
drugs. But my own sense is that once a psychosis is established, there is an
adverse effect on the long-term course, and so there are a lot of reasons to
be interested in the data on intervention that suggest you can make the
long-term course a lot better."
Carpenter said that if included, a risk syndrome category related to
psychosis risk would likely be written as a separate chapter to underscore
that it is not a diagnosis of psychosis and need not be tightly linked with
schizophrenia. Other sub-threshold risk categories could be added as
prevention science advances.
The most vigorous response to Frances came from Thomas McGlashan, M.D., a
pioneer in prevention of psychosis. He said Frances's criticisms ignored the
substantial literature on the value of early intervention, belittled the
demonstrated predictive capacity of proposed risk syndrome criteria, and
slighted the harm that can be done to true positives by failing to receive a
diagnosis of risk.
"The risk syndrome criteria we are proposing for DSM-V have
already proven capable of identifying a clinical entity within a help-seeking
population in which 1 out of 3 individuals develops a bona fide
DSM-IV psychotic disorder within two and a half years," he
said. "This amounts to a true positive rate of 33 percent."
McGlashan said that compares favorably with the predictive power of
hypertension for stroke or of hyperlipidemia for a coronary event.
Regarding Frances's concern about the "false-positive"
individual, McGlashan countered: what about the true-positive person who"
regularly gets forgotten in the surge of self-righteous
indignation" about false positives?
"For the true-positive person, the consequences of not monitoring
risk over time could well be an unexpected chaotic first psychotic break
destructive to reputation if not to physical well-being, leading to a negative,
adversarial, revolving-door relationship with the treatment system, not to
mention the ultimate deficit of chronic schizophrenia. These risks, and the
evidence that early detection and intervention can modify them, get ignored
regularly in the exclusive focus on the false-positive 'victim' of early
Carpenter said consideration of "stigma" requires more nuance
than Frances's anxiety about false positives in the general population;
rather, it needs instead a careful weighing of distress and disability among a
help-seeking population and of the harm or benefit associated with several
"The number of false positives in population sampling is
meaningless," he said. "It is the number of false positives in a
help-seeking group experiencing disability or distress that matters. Then, the
question remains: to what extent are these false positives harmed versus
helped by placing them in a risk syndrome class? Some will proceed to other
diagnoses [other than a psychotic disorder], some to schizotypal, and some
will not proceed to a diagnostic category. In each instance, what is the good
and the harm? But the debate is meaningless if one simply presumes all false
positives are harmed."
The online forum of the Psychosis Work Group on inclusion of a
category for the prodrome can be accessed at<www.schizophreniaforum.org/for/live/detail.asp?liveID=68>.▪