Machine Learning

Machine Learning

Empiric believes that the application of Machine Learning and Artificial Intelligence advances our ability to quickly identify and reduce unwarranted clinical variation in physician practice.

Empiric’s use of Machine Learning begins in the building and validation of new cohorts and new cohort rules. Empiric Health has over 40,000 algorithms using both structured and unstructured data and a library of over 370 cohorts. However, the creation of new cohorts is time consuming and the reliance on human experts can lead to inconsistencies in cases where clinical opinion varies. Empiric has used clinically-informed Artificial Intelligence algorithms to automate the creation of cohorts and Machine Learning models to improve the consistency of the cohort definition process.

Empiric has also utilized the power of Machine Learning to build analytical models on payor claims data to automatically identify patterns of variance in physician practice. Supervised Machine Learning algorithms have been applied to determine clustering of surgeons, supplies, medications, tests, etc., that lead to an optimum alignment of cost, quality, and outcomes, all done within the context of a given cohort.

We believe our NLP and Machine Learning capabilities create a powerful platform to drive physician change through clinically meaningful data analytics.

← Back

Check out Megan Bultema and Justin Schaper present an overview of “Removing Unrelated Claims from a Total Cost of Care Analysis” at HAS20