Objective: Two forms of early intervention in psychosis are currently being proposed--intervention before the onset of illness with at-risk individuals and intervention after the onset of psychosis.
Method: The risks and potential benefits associated with these two approaches were examined using published data and Bayes probability theorem.
Results: Claims for benefits from early intervention in established psychosis go back more than two hundred years to the nineteenth-century advocates for asylum construction. Theoretical support for early intervention in psychosis rests on data suggesting that a longer duration of untreated psychosis (DUP) is associated with poor outcome. The association of DUP with outcome, however, appears to be inseparable from the confounding influence of the good outcome expected in recent-onset psychosis. Although some researchers advocate treating people with premorbid features of psychosis or other high-risk indicators, adequate screening measures and effective interventions are not yet available. The most promising current screening measure, if applied to the general population, would be accurate only two percent of the time. The accuracy of these measures can be improved by screening only those patients who are referred to a clinic, but this strategy will result in many cases in the population-at-large being overlooked.
Discussion: The risks and potential benefits associated with the two early intervention approaches are very different. The provision of optimal treatment early in psychosis could produce benefits, but people with brief, good-prognosis psychotic episodes are likely to receive unnecessary treatment. Screening the general population or patients referred to a clinic will have little public health impact on the incidence of schizophrenia. There is likely to be considerable negative impact on those who are inaccurately labeled as being at imminent risk of psychosis and are treated accordingly.