The question of how best to revise the 'bible' of American psychiatry once again has tempers flaring. The manual, the Diagnostic and Statistical Manual of Mental Disorders (DSM), is significant because it is used to determine clinical diagnoses, insurance reimbursements and research agendas throughout the United States, and is often used as a reference in other countries.

For the fifth edition, David Kupfer hopes to expand the remit of the psychiatry manual.

Planning for the upcoming fifth edition of the manual (DSM-V) began in 1999, but as work has picked up during the past year, critics have alleged that the process has been too secretive, and that working groups have been pushed to meet an unrealistic 2012 publication date. Some, including the architects behind the last two editions of the DSM, also complain that project leaders are pushing for the premature inclusion of changes meant to incorporate recent genetic and neurobiological advances, before they are ready for the clinic.

Supporters, meanwhile, say that the changes will not be too drastic, and are meant to make the manual more flexible for future revisions.

Even light tweaking of definitions in the DSM can bring about radical changes in psychiatric practice, as Allen Frances, professor emeritus of psychiatry at Duke University in Durham, North Carolina, wrote in a recent commentary in Psychiatric Times. Frances, chairman of the committee that produced the fourth edition of the manual in 1994, acknowledged that changes in the definitions of autistic disorder and attention-deficit/hyperactivity disorder made then may have contributed to the recent surge in diagnoses of these conditions.

In addition, the manual sometimes has an outsized influence on research directions, says Steven Hyman, provost of Harvard University and a member of the DSM-V task force that is overseeing the revision. Hyman became interested in reforming the DSM when he was director of the National Institute of Mental Health in Bethesda, Maryland, and witnessed the control that it exerted over grant review panels. "I was spending taxpayer money on grants that were being forced into categories that might or might not conform to nature," he says.

There just hasn't been time to do this in an organized way.

The latest revisions come as financial ties between prominent psychiatrists and pharmaceutical companies are being closely scrutinized. A 2006 analysis of potential conflicts of interest among those who participated in the last revision showed that 56% of panel members had financial links to the pharmaceutical industry (L. Cosgrove et al. Psychother. Psychosom. 75, 154–160; 2006). For the DSM-V, the American Psychiatric Association, which publishes the manual, vetted potential members of working groups under a new conflict-of-interest policy. But the process introduced delays, and working groups were not finalized until 2008.

Then word broke last July that working group members had signed non-disclosure agreements, agreeing to refrain from distributing pre-publication materials or divulging the content of group discussions pertaining to the rewrite. The agreements were intended to prevent members from publishing material to be used in the DSM-V, says Darrel Regier, vice-chair of the DSM-V task force and director of the American Psychiatric Association's research division. But Robert Spitzer, a professor of psychiatry at Columbia University in New York who oversaw the DSM-III, says that his request to look at minutes of a DSM-V meeting was denied because of these confidentiality agreements.

Since then, DSM-V working groups have begun posting regular summaries of their activities online. "All of us have been encouraged to be as public as we can be," says William Carpenter of the Baltimore-based University of Maryland School of Medicine and chair of the working group on psychotic disorders. "But just to have [the confidentiality agreements] — that never would have been considered in the previous revisions," says Spitzer.

From the start, Hyman and DSM-V chairman David Kupfer have also planned to change how the DSM-V evaluates mental disorders. Rather than relying strictly on categorical diagnoses — one either has depression or does not, for example — they have pushed to add 'dimensional' criteria to ascertain to what extent a person is depressed. Such criteria could also address similarities among different disorders, reflecting, for example, neuroimaging studies that suggest multiple anxiety disorders can affect the same region of the brain.

For this reason, Carpenter and others have said the DSM-V will represent a "paradigm shift" — an expression that alarmed critics, who say the science behind such dimensional assessments is not yet ready to be incorporated into clinical assessments. In March, Duke University psychologist and epidemiologist Jane Costello resigned from the working group on child and adolescent disorders after receiving a memo from Kupfer and Regier about including the dimensional approach. Adding these assessments would require a great deal of extra research, she says, at a time when working groups were already behind schedule for their 2012 publication deadline. "There just hasn't been time to do this in an organized way," she says. "This is a huge job."

Hyman says that the changes will not necessarily be so drastic, and could take the form of a few additional, and optional, diagnostic criteria without replacing the old methods.