Reviewing medical records: the modern Sisyphus.
The legendary king of Corinth, Sisyphus, rolled a heavy rock up the hill, again and again, becoming the symbol of laborious and repetitive work.
Today's Underwriters and Claim Analysts who review and summarize voluminous medical records for risk assessment can undoubtedly relate to his plight.
The US insurance industry is reviewing ~13M medical records per year for underwriting and claim management (8-10M in Life/A&H and 4-5M in P&C). More than 50M hours of human labor is applied to this task at the cost of more than $3B….year after year.
The process of manually reviewing medical records is labor-intensive and error-prone, but that's only natural. Reviewing hundreds or thousands of pages of medical information, day after day, is extremely tiring and challenging, even for trained professionals.
DigitalOwl had its NLP system compete with insurance physicians analyzing random medical records. In all cases, the NLP system identified twice as many significant medical findings as to the physicians (blood clots, hypertension, biopsy, smoker, etc.) and completed the analysis in 3-5 minutes, with the physicians taking 2-4 hours to complete each case.
Using a state-of-the-art NLP engine can save insurance companies, TPAs, and other insurance-related entities hundreds of thousands of hours of labor a year and reduce errors to improve risk management. It's also allowing them to focus their highly trained people on the important work of making and communicating decisions.
This reduction in costs and improvement in risk results can also translate to more competitive pricing, directly benefiting consumers. A real "win-win" enabled by advanced technology.
To read DigitalOwl's full Executive Briefing, "Harnessing AI for Faster and More Accurate Underwriting and Claim Management," click here.