ABSTRACT
Many mental health problems are recurrent, heterogeneous and fluctuating. Intensive longitudinal data from the experience sampling method (ESM) can help to better predict important changes over time. One of the most promising clinical uses of ESM is to identify precursors to future crises. This could ensure earlier detection and intervention, and avert escalations in symptomatology. In this chapter, the authors provide an applied framework for the use of prediction models in identifying the temporal association between mental health variables. We review recent studies that have employed ESM to reveal aspects of the short-term course and theoretical mechanisms of psychiatric disorders. We then go on to explore considerations for data collection, statistical analyses and stakeholder engagement, as well as the promise and potential pitfalls of preventative interventions that leverage predictive modeling in real-time data.
Word count: 5711.
