MIT Health cut prior auth time in half with AI
MIT Health cut prior authorization handling time by 45% with Notable's Sidekick AI co-pilot. See how they processed 750+ cases faster with fewer denials.
reduction in prior authorization handling time
cases processed since go-live
The challenge
As a campus health center not anchored to a hospital system, high-acuity events such as ED visits, hospitalizations, and advanced imaging are often handled elsewhere. This makes timely documentation for prior authorizations and follow-ups more challenging, increasing the risk of denials, delays, or missed follow-ups.
Authorization decisions depend on recent provider notes, imaging history, and allergies, requiring all data elements for a usable clinical packet.
Gathering those critical details became cumbersome for several reasons.
- Manual processes: Staff manually gathered documentation, toggling between the EHR and the Carleon portal to support medical necessity, leading to duplicate searches for the same information.
- Time-consuming workflows: Prior authorizations involved high-touch workflows that required substantial staff effort, including printing and retrieving documents from the EHR, manual uploads, and answering detailed portal questions.
The solution: Notable Sidekick AI Co-pilot
Sidekick is Notable’s AI co-pilot tool designed to support healthcare staff through complex workflows that traditionally require time-intensive clinical reasoning and deep document analysis. The assistive AI agent uses natural language processing to help staff with complex tasks like interpreting multi-year EHR data, responding to payer clinical questionnaires, and generating comprehensive documentation packets that improve approval rates.
For MIT Health, this support is embedded directly into the authorization submission workflows across primary care and specialty clinics. This reduces the number of peer-to-peer reviews and denials.
Sidekick offered a path to:
- Put all relevant orders and chart context in one place
- Pre-assemble clinical packets
- Answer clinical questions with context from the patient’s chart
A high-touch, manual workflow required staff to print or retrieve documents from Oracle, manually upload them, and answer detailed clinical questions in the Carelon portal. MIT Health and Notable defined the workflow for Sidekick using an iterative, builder-centric approach to streamline the inefficient process.
The new workflow centralizes clinical data, automates documentation assembly, and guides staff through prior authorization submissions. In addition, missing allergies, packet-generation failures, or other automation errors create human tasks with clear instructions.
Results
Since Sidekick’s implementation, MIT Health has seen measurable improvements in the authorization process, including greater payer compliance, higher-quality work, and increased tool adoption.
Time savings
Staff complete authorizations faster, reducing manual abstraction and enabling them to focus on high-value work, such as interpreting patterns for more effective interventions.
A Sidekick-supported authorization can be completed in an average of 6 minutes, an estimated 45% reduction from the previous average ~11-minute handling time, when all information is available.
Staff experience and quality of work
Sidekick provides comprehensive information and proactively guides staff through the workflow. The assistive AI agent facilitates rapid responses to clinical questions in the portal and efficient copying and pasting, rather than repetitive re-entry, which is well-received by MIT Health team members.
Peer-to-peer and denials
MIT Health is quantifying the impact as volume and use cases build. Staff reports fewer requests for peer-to-peer review, attributing this to Sidekick’s AI-assisted clinical responses, which help reduce denials.
“I have fewer requests for peer-to-peer review for sure.” – Jessica S. Another team member expressed that Notable’s Sidekick “has all the necessary information” and is “helpful as we don’t have to go back and forth to the EHR.”
Adoption and scale
MIT Health actively monitors AI use and adapts based on feedback to reinforce a commitment to continuous improvement.
Since go-live, MIT's Companion Authorization Submission flow for Carelon has run more than 750 times, reflecting operational momentum and leadership support.
Sidekick is helping MIT Health deliver a more efficient, streamlined process with real-time authorization insights and smarter workflows.
Key outcomes
- An estimated 45% reduction in prior authorization handling time
- Fewer peer-to-peer reviews and denials
- Improved staff efficiency and satisfaction
- ~750 cases processed since go-live