How ProAct increased reviewer productivity by up to 40% with AI-assisted clinical data extraction
During an initial pilot of ClinIntel within PAHub, ProAct reviewers completed prior authorization reviews up to 40% faster while maintaining strong AI accuracy and expanding use across additional product lines.
Less manual review. More focus on clinical decisions.
ProAct needed a more efficient way to review lengthy clinical documentation, locate supporting evidence, and answer repetitive questions as authorization volumes increased.
The Customer
ProAct is a growing pharmacy benefit manager focused on delivering clinically driven pharmacy benefit management while maintaining exceptional service and operational efficiency.
The Challenge
Reviewers were spending valuable time manually reviewing clinical documentation, locating supporting evidence, and answering repetitive questions before making informed authorization decisions.
- Reviewing lengthy clinical documentation
- Searching for supporting clinical evidence
- Manually answering repetitive questions
- Spending time on administrative work
Our Solution
ProAct implemented ClinIntel within PAHub to analyze submitted clinical documentation, identify relevant clinical evidence, and pre-populate answers to clinical questions.
- AI-assisted review of submitted documentation
- Relevant clinical evidence surfaced automatically
- Answers pre-populated for clinical questions
- Reviewer validation before final submission
- Reduced administrative work
How ClinIntel Works Within PAHub
Strong performance across multiple drug therapies
ClinIntel delivered an overall 88% completion rate and 83% AI answer acceptance across the enabled questions measured during the study period.
Ready to transform prior authorization reviews?
ClinIntel, a feature within PAHub, uses AI-powered clinical data extraction to reduce manual effort, accelerate reviews, and help health plans and PBMs scale prior authorization operations with confidence.
*Human intervention includes user corrections and completion of unanswered questions. Results reflect the ProAct study period and enabled question configuration.