After two decades of research, prevention of psychosis becomes increasingly accepted in clinical psychiatry. However, there are still unmet scientific and clinical needs. Therefore, guidance for prediction as well as prevention is required, reflecting their current capabilities, but also their requirements and limitations.

Evaluating the current state of risk estimation and prevention.

Developing clinical recommendations for the prediction and prevention of psychosis.

42 samples, mainly defined by ultra-high risk criteria and/or the basic symptoms criterion ‘COGDIS‘, were included into meta-analyses of prevention, 15 studies into meta-analyses of prevention.

The pooled conversion rate at >4-year follow-up was 37.0% in UHR and 61.3% in COGDIS samples. The 12-month pooled risk ratio was 0.44, the NNT 10. Psychosocial functioning seemed not to improve, however results were inconclusive due to methodological issues of the trials. Both meta-analyses indicated age related differences.

Several recommendations were developed to guide prediction and prevention, emphasizing age-adapted strategies; details will be presented and discussed during the symposium.

Regarding future steps to further improve prediction and thus prevention, neurocognitive and neurobiological parameters of information processing, i.e. mismatch negativity, P300 and processing speed, as well as support vector machine based analysis of structural MRI seem to be most promising. Furthermore, with regard to current developmental models of psychotic disorders, risk should be conceptualized as dynamically modulated over time and thus presumably non-linearly related to future outcome. Therefore, studies need to consider the fluid interplay of risk and resilience factors to advance prediction significantly.