The prodromal stage may last from several weeks to several years, and some patients may never develop schizophrenia. Thus, one of the first challenges of prodromal treatment is to identify the most vulnerable patients. In recent years, Alison Yung and Patrick McGorry, both professors at the University of Melbourne in Australia, described three features that can put patients at high risk for schizophrenia: unusual perceptions or beliefs that patients don’t yet hold in rigid ways, such as paranoid thoughts that they’re being followed or watched by others; a family history of schizophrenia; and a tendency to experience fleeting hallucinations and delusions that they can still distinguish from reality.

The Australian scientists produced a detailed diagnostic questionnaire that Thomas McGlashan, a professor of psychiatry at the Yale University School of Medicine, modified for research settings. McGlashan’s “structured interview for prodromal syndromes,” or SIPS, has become a standard diagnostic tool in the United States and in some European countries. Patients who undergo a SIPS evaluation might be asked whether a voice they hear is real or “just in their head.” They might also be asked whether anyone else can hear the voice. A patient’s answers enable clinicians to characterize prodromal symptoms and their severity and decide what to do, if anything.

By making it possible to identify vulnerable patients consistently, the SIPS evaluation helped fuel a dramatic expansion in prodromal research. There are prodromal studies at more than 30 research sites worldwide. Robert Heinssen, associate director for prevention research at the National Institute of Mental Health (NIMH), heads the North American Prodrome Longitudinal Study, a collaboration of eight independent research projects with a total of more than 850 subjects. McFarlane’s program, launched in 2000, is now among the largest schizophrenia prevention efforts in the United States. pier counsels teachers, clinicians, social workers and others who interact with young people on how to recognize prodromal warning signs: sudden declines in performance at school, social withdrawal, poor hygiene and trouble concentrating, among others. Patients receive two to four years of family counseling designed to build specialized coping skills and reduce stress. They also receive focused guidance at school or work and medications that often include antipsychotic drugs. “Most of our kids keep getting better,” McFarlane asserts. “Our experience has been that after a couple of years, they don’t need a lot more help.” McFarlane predicts that most patients will be able to come off medication, but he can’t yet say just when.

Other programs have similarly positive outcomes. Still, Barbara Cornblatt, a professor of psychiatry at the Albert Einstein College of Medicine in New York City, cautions that this approach is not ready for routine use. “We don’t know enough yet about appropriate treatment,” she says. “It may turn out that antipsychotics are useful only in combination with psychosocial approaches, or that we don’t have to use those drugs at all. If we do use them, we have to figure out when and for how long.”

Whether to prescribe antipsychotics is the burning question in prodromal treatment. The drugs’ side effects, including weight gain and heightened risk of diabetes, can themselves pose dangers to a patient’s health. Antipsychotics may also interfere with normal development of the adolescent brain, and McFarlane’s willingness to use them disturbs many scientists, though he insists that when side effects are observed, the PIER physicians immediately change the drug regimen. What’s more, the SIPS questionnaire, though sophisticated, is hardly foolproof, and there’s no blood test that predicts psychosis. Physicians must base a diagnosis on sometimes vague behavioral symptoms and subjective responses to questions, and part of the time they’re going to be wrong.

How many prodromal patients will in fact become psychotic? McFarlane thinks about a third will “convert” in the year after they’ve been identified. A recent NIMH study, published in the Archives of General Psychiatry on Jan. 7, 2008, found prodromal behaviors could predict psychosis correctly from 35% to as much as 80% of the time in high-risk youth, depending on the number of symptoms and their intensity. Those odds are high enough to justify treatment, McFarlane says. But conversion estimates gathered at individual program sites vary widely, with some as low as 20%. If the real conversion number is just one or two in five, suggests Diana Perkins, a professor of psychiatry at the University of North Carolina, prescribing antipsychotics may be too perilous, subjecting 60% to 80% of prodromal patients to medications without known benefits.

Only one study has investigated whether psychosocial interventions without drugs might effectively treat prodromal patients. Anthony Morrison, a professor of clinical psychology at the University of Manchester in England, found that giving high-risk patients cognitive behavior therapy (CBT) for as long as six months reduced the likelihood of psychosis by more than 90%. Morrison published his findings in the British Journal of Psychiatry in 2004; however, his study was limited to 58 patients, and their average age was 22. Younger patients diagnosed as prodromal tend to have much more severe symptoms.

There is also just a single study, published in 2006 by McGlashan and his colleagues at Yale and other centers, that has investigated antipsychotic medication as the sole therapy. The findings suggested that a drug called olanzapine could cut prodromal patients’ conversion rates in half. But those results are considered inconclusive because such side effects as weight gain and fatigue led to a high dropout rate in the treatment group, resulting in a less-than-optimal sample size.