When a medical claim is submitted, the insurance provider must process the claim to determine the correct financial responsibility of the insurance provider and the patient. The process to determine this is broadly known as claims adjudication. It involves creating a claims processing workflow that checks each claim for authenticity, correctness, and validity based on coverage. Some of the steps in this workflow involve working with unstructured data which requires manual steps in the workflow to extract the information buried in the unstructured notes. To process volumes of claims in a cost effective and scalable manner, healthcare payers are increasingly looking at machine learning to reduce dependency on humans and rely on automation as much as possible. Additionally, analyzing and interpreting health claim data is powerful in driving improvements in population health to address issues related to cost, quality and outcomes. According to the CDC report analyzing claim documents will help identify certain behaviors would help in preventing or delaying the development of a medical condition. AWS provides a comprehensive list of machine learning and analytics services that allow developers, irrespective of their background, to start integrating machine learning and analytics technology into their applications. Through this session, we demonstrate how we can use two AWS AI services, Amazon Textract and Amazon Comprehend Medical to automate the claims adjudication workflow and run analytics on top to extract entities using Amazon Athena.
AWS services: Amazon Textract, Amazon Comprehend Medical and Amazon Athena
Speaker: Joinal Ahmed, Associate Solutions Architect, AISPL