Fewer fraud leaks, faster prior auth
A health insurer needed to spot suspicious patterns earlier and get clean prior-auth decisions into the hands of providers faster.
A health insurer needed to spot suspicious patterns earlier and get clean prior-auth decisions into the hands of providers faster.
Fraud and abuse were often found in retrospect. Prior authorization was a bottleneck: manual review, back-and-forth with providers, and delayed care. The plan had claims and clinical data but no unified way to score risk at the point of submission or to auto-approve straightforward prior-auth requests against policy and guidelines.
We built two modules that integrate with their core systems. First: a fraud and abuse detection layer that scores claims and referrals using patterns (provider, member, procedure, geography, history). High-risk items are routed to SIU and investigation workflows; the rest flow normally. Second: a prior-auth assistant that checks incoming requests against eligibility, medical policy, and simple clinical rules. Clear approvals are automated with an audit trail; edge cases are sent to nurses with pre-filled context so they can decide faster. Both modules are designed for their regulatory and privacy constraints and run in their environment.
SIU is catching more suspicious activity earlier and wasting less time on false positives. Prior-auth turnaround for standard requests improved significantly, and provider satisfaction with the process went up. The plan is now looking at extending the same approach to prepay vs postpay and to more complex procedures.