Predicting Response to Mental Health Treatment using Environmental Data and the PREDICTIX Digital Algorithm ̶ A Technology Overview
Mental health conditions affect 1 in 5 people globally, with demand on health systems increasing driven by COVID-19. With longer waiting times to see psychiatrists and General Practitioners (GPs), patients have turned to more accessible online solutions to receive care.
The PREDICTIX Digital technology harnesses artificial intelligence (AI) to analyze the interrelationships between complex environmental interactions – combining clinical and demographic data sources considered important factors in determining response to medications – to predict efficacy and adverse effects of treatments with high accuracy, starting in the field of depression.
Based on an analysis of the worlds largest prospective datasets, PREDICTIX Digital predicted response to antidepressant medication for a participant with an average accuracy of 65.29% (SD 4.24) across medications, while the best performance in antidepressant prediction was seen with Bupropion (70.68%) across the validation-test set.
For comparison, STAR*D’s average response rates across relevant medications for the same group of participants (i.e., validation-test sets) was 48.3%. PREDICTIX Digital’s average accuracy of 65.29% is comparable or superior to today’s pharmacogenomic tests (39%-64%)i.