Meanwhile, our data science team was working on a project to demonstrate that IAPT data can be used to predict patient outcomes, and by providing insights into the likelihood of disengagement with therapy, enabled clinicians to take action to intervene and improve attendance / dropout rates. Around 30% of patients disengage before treatment begins and a further 35% disengage during treatment. These predictive tools are already in live use, offering insights that will help more people with common, debilitating mental health conditions access treatments that can have a profound impact on their lives.