TRANSFORMATIVE AI TO REVAMP PRIOR AUTHORIZATIONS

Tech enablement drives lower costs, lessened provider abrasion

Health plans are unnecessarily burdened by administrative tasks such as prior authorization—a costly bottleneck to optimized care and operational outcomes. Recent American Hospital Association (AHA) research states that that 95% of providers spend increasing time seeking prior authorization approval. An astounding 78% of hospitals and health systems said their relationship with commercial insurers is getting worse. What’s needed is relief at this high-impact payer-provider touchpoint. AI-enabled prior authorizations can help. According to McKinsey analysis, AI-enabled prior authorizations can automate up to 75% of manual tasks. This tech enablement solves for many of the inefficiencies while working to improve payer and provider synergy, reduce costs, improve case turnaround time, and, most importantly, drive better health outcomes for patients.

Drivers to Digital Utilization Management

The case has never been clearer for a straightforward, faster, and impactful PA process. High-impact PA ups the ante with an analytical approach to flag clinical decisions that need priority nurse review. The algorithmic decision support can identify high dollar clinical usage patterns, FWA patterns and CPT conversion/alternative treatment. Traditional, siloed UM programs are not adequate to meet current market conditions: a high-cost clinical resource pool and financially pressured payers and providers. Add to that this simple fact: our internal research shows that 80+% of authorizations reviewed are eventually approved without any modifications, and only a few of the cases need a detailed review for necessity. This leaves a lot of room for digital intervention in the provider interaction channel to instantly auto-approve and provide an outcome via digital assists for better member care.

Experts such as healthcare business process management partners have the combined digital solution and skillset: experienced clinicians supported by AI workflows to effectively manage the process and cost. Backed by process re-engineering, automation and digitization of the prior authorization process will ease provider burnout and change the perception of this process. Traditionally, the prior authorization step has been viewed as cumbersome, with high administrative costs running into billions of dollars across the US healthcare sector. BPM partners with experience across both payers and providers can collaborate and bridge gaps to improve the overall ecosystem of population health and give a better dollar accountability and visibility. As a result, transformed, high-value utilization reviews will drive to lower-cost per review using state specific/CMS guidelines and a 99%+ calibration with medical directors on review decisions.

UM Augmented with Transformation Levers

The front- and back-office digital suite of solutions use automated speech recognition (ASR), NLP of medical entities, contextual insights, computer vision, and machine learning feedback to evaluate prior authorization requests against automated guideline rules. The outcomes are provided via digital nurses on some channels and nurse assist flows on others, each with comprehensive criteria summary and recommendation on approval or referral to a medical director.

Outcomes and Future of Clinical Reviews

BPM partners with experienced clinical resources combined with AI workflows can manage the PA process cost effectively and find transformation with process effectiveness. For the first launch of the back-office solution for a top payer client, Sagility’s prior authorization medical record review automation leveraged in-house asset Intelligent Content Processing (ICP) for a 20% efficiency gain across the entire document review process, with a line item gain of 75% error reduction, in some cases. With the buildup of this platform to transform clinical data into intelligence to reduce process complexity, enhance accuracy, and speed TAT which is 7-10 minutes saved per case transaction. Through machine learning, these savings will compound over time, for up to 40% cost savings for identified scenarios. This powerful tool will also drive significant TAT improvements as well as enriched provider and member experience (potential Star and NPS ratings impact) from improvements on quality with reduction in nurse error and nurse glide path, enhanced productivity, decrease in regional medical director (RMD) routes, increase in identification of potential approvals for nurse to review and also decrease in number of cases referred to physician.

Transforming the front office with the voice assist and context bots results in auto-approvals for high volume procedures. This directly reduces call volume reduction by 10-15%. The next-best action and access to unified knowledge bases have driven a shortening of complex prior authorization calls, for 15-20% savings via handle time reduction. The continuous digital bot and agent monitoring metrics have also resulted in an internal training glide path reduction by 50%, ensuring the quality of outcome at above 99%+. Overall savings in the front office can be at least 20% as skilled resources need no longer work on administrative tasks anymore.

As next steps, any outcomes of clinical decision support that sources from a unified data lake of claims, medical records, and member profiles can be successfully linked with population health outcomes and patient profile stratification. This will result in insights on high risk population analysis, population trends and wellness management, and, finally, risk scoring of individuals and proactive population management. Weaving in provider profiles can also help provide in extension insights on the segmentation, scorecards, and comparative analytics to evaluate provider utilization and abuse, if any.