Based on risk, cost or opportunity targets – provide “person-based scoring technology." These are person-by-person real-time updates from highest likelihood to lowest for key outcome targets. Stratification scores for every person raise and lower in real-time as new data feeds the platform. These “stratification scores” provide clients a focused “priority list” of where to best apply resources to optimize outcomes. Analysis on “why” (reasons) each person received the stratification score that they did is also performed.

The intervention engine consumes data from assessments, model risk scores, raw data such as demographics and diagnostics, as well as derived data such as generated features and comorbidities. The engine uses that data to determine which interventions are appropriate and orders them based on risk from the models. Suggested interventions (actions) are managed from a diverse and growing repository. The engine provides a navigation pathway, sending insights to the correct team resource. Ultimately, all the “performed actions” for every person across all points in time become part of the platform’s automated feedback loop and deep learning cycle.

The feedback loop compiles outcome results of each accepted or declined action & stores this data back for pre/post analysis. Over time, the platform learns which actions work and which do not. The only way to significantly improve performance metrics is to measure the effectiveness of all actions taken – every day for every person. That involves tremendous complexity. The platform handles this complexity & delivers that needed information to optimize future outcomes.