What’s new with AI (Artificial Intelligence) in Workers Compensation Claims Processing?
By Kevin Ross
It used to be that certain parts of the processing of a Workers Compensation Claim were 100% manual. The gathering, organization, and analysis of files – the first step in processing any new claim – was an unavoidably time-intensive task. A human got boxes of paper files or electronic pdfs (or other formats) related to a disability claim and had to decide which ones were pertinent or non-pertinent, which ones were redundant, and which ones went into which stacks or folders. This sorting operation could take hours – or days – and had to be completed before the doctor even started to review the claim.
AI has changed that. Increasingly, scanned documents and pdf files are now being automatically sorted by category, using Artificial Intelligence (AI). AI software works with Optical Character Recognition (OCR), Natural Language Processing (NLP), and other smart technologies to decide WHO the document came from, WHAT it says, WHERE it is going next, WHY it is going, and HOW the document will be used now and in the future.
For the same job, manual sorting speed is 2k~3k pages per hour, versus seconds for AI. And, unlike manual sorting, AI uses Machine Learning (ML)/Deep Learning (DL) to simultaneously compare what it understands to everything in the database. Based on continually evolving algorithms, the original data is ingested, curated for relevance, and organized. The end-user, such as an IME, carrier, or attorney receives an indexed, searchable hyperlinked master document and summarization for case review that can be sorted chronologically, by document type, by medical provider, and more. The technology can also be used predictively to analyze trends and inform business decisions going forward.
AI does not make the hard decisions around recommended outcomes. Those require a human (for now). What AI does do is answer the questions that inform those decisions.
Want to know if the person being treated has been treated previously? AI
Want to know if this is a duplicate document? AI
Want to get an idea of all of the previous medical conditions experienced by this person?
Yep. AI!
Beyond intelligent sorting, AI is being deployed in the Workers Comp area to transform a wide variety of disparate data into actionable information. AI Compared to manual processing, AI delivers faster results less expensively and handles many tasks in parallel. Deep Learning technology also improves accuracy as the database ingests additional information and considers all data, even that which might be missed by humans.
Kevin Ross is the senior vice president of sales and marketing for Numina-RapidCare Group where he focuses on e-Discovery and AI-powered technologies developed for medical, legal, and insurance verticals.