AWS Public Sector Blog
Acentra Health processes 35M Medicare documents 50% faster with IDP on AWS
Acentra Health, an Amazon Web Services (AWS) Advanced Partner, works with 45 Medicaid agencies and 25 federal agencies across 12 departments, including the Centers for Medicare & Medicaid Services and the Department of Health and Human Services. As one of only two Beneficiary and Family Centered Care-Quality Improvement Organizations (BFCC-QIO) in the US, the company helps more than 140 million Medicare beneficiaries access and protect their Medicare rights to improve quality of care.
As a BFCC-QIO, Acentra Health helps Medicare beneficiaries file quality of care complaints and appeals regarding early hospital discharge or early termination of skilled services. Processing these cases requires meticulous data entry from patient records and forms related to prior authorization notices, patient care, and case management, which often consist of complex medical history.
To better support its clients, Acentra Health implemented an intelligent document processing (IDP) solution using AWS services. In this post, learn how Acentra Health’s IDP solution reduced document processing times by more than 50 percent and lowered document processing costs by 40 percent.
Innovating document processing with AWS
Before building their AWS powered solution, most of Acentra Health’s document processing used legacy optical character recognition (OCR) systems that required manual search and data entry across multiple systems—a time-consuming process susceptible to human error. When the volume of documents ballooned, costs increased and care providers experienced delays in accessing vital information.
Acentra Health leadership knew that cloud technology could provide an opportunity to streamline operations and innovate patient care, so they reached out to AWS. Together with the AWS Product Acceleration Team, they developed a solution: an IDP pipeline that uses OCR, powered by machine learning (ML) technology, to recognize, understand, and extract text automatically from digital images and scanned documents—at scale.
Building a serverless OCR solution, powered by ML
Acentra Health’s IDP solution for its BFCC-QIO work helps clients extract important data from documents and forms in seconds, supporting efficient resolutions to appeal and quality of care cases for Medicare beneficiaries. The solution transforms scanned image files and faxed documents into a searchable format for quick access to evidence in support of these cases. The crux of the solution is built around Amazon Textract, an ML service that automatically extracts text, handwriting, layout elements, and data from scanned documents, images, forms, tables, and more.
To reduce processing time and costs, Acentra Health’s solution uses serverless, event-driven architecture. When a user uploads a scanned document into the solution, the file is sent to Amazon Simple Storage Service (Amazon S3), which triggers a process managed by AWS Step Functions, a visual workflow service. AWS Lambda, a compute service that runs code in response to events, sends the documents to Amazon Textract for analysis. Amazon Textract extracts key pieces of data from the documents, which is then sent back for storage in Amazon S3 along with the original documents. The documents’ metadata is stored in a serverless, NoSQL Amazon DynamoDB database for cost attribution, analytics, and audit purposes. The solution also supports keyword-based bookmarking, so users such as healthcare practitioners can find necessary information more efficiently.
“Document OCR and bookmarking medical records have benefited the BFCC-QIO team greatly,” said Rose Hartnett, director of operations for Acentra Health. “They have increased our efficiency, as we no longer have to manually search through records for the needed information. They allow us to easily navigate through lengthy medical records and locate specific sections or important information. This saves us a considerable amount of time and ensures that we can provide accurate and timely reviews.”
A solution that cuts time by 50% and costs by 40%
In its initial phase, Acentra Health’s new IDP pipeline is expected to process almost 35 million pages a year, as the company begins supporting 29 states in its BFCC-QIO network. Already, the solution has cut document processing times by more than 50 percent and lowered costs by 40 percent. By reducing human error in manual entries, the solution is also more accurate, helping to shrink clinician workloads and enhance the patient experience.
“The solution has fundamentally transformed our approach to handling inbound medical documents like images and faxes at scale,” said Sean Harrison, Chief Analytics Officer for Acentra Health. “By converting these documents into searchable text formats, we have greatly improved the efficiency and accuracy of nurses managing critical tasks and workflows. This innovation streamlines time-sensitive care management processes and enables our clinicians to operate at the top of their license.”
Getting started with IDP for state agencies
Acentra Health’s IDP solution improves the customer and patient experience in government healthcare by streamlining case management processes. Faster time to answers and improved data accuracy translates to accelerated case processing, increased clinician efficiency, and improved quality of care. And the benefits don’t stop there. Acentra Health is developing new software features using AI to summarize patient records across documents, generate notes from doctor visits, create customer support chatbots, and more.
With flexible and scalable ML services from AWS, organizations can rapidly build and iterate a customized IDP solution tailored to unique needs. A strategic document automation pipeline can optimize document workflows, generate efficiencies, reduce costs, and improve the beneficiary experience.
Learn more about Acentra Health or get started with your own intelligent document processing solution on AWS.