DTR workflows that work: Medical policy digitization

Let’s say you’ve cleared your compliance strategy’s first big technical hurdle—your FHIR APIs (CRD, DTR, and PAS) are ready to go. You’ve laid the groundwork for an interoperable prior authorization workflow, connecting systems for prior auth intake in alignment with CMS-0057-F. That in itself is a huge accomplishment.

But here’s the catch: that’s just the beginning.

While much of the industry is understandably focused on the complexity of API integrations, a recent WEDI poll¹ revealed the top two challenges among health plans in complying with CMS-0057-F:

1. Determining a cohesive enterprise strategy for interoperability
2. Digitizing prior authorization policies

These two top the list of several issues lurking beneath the surface of the API requirements, and they are very much intertwined. Your health plan’s medical policies are the crux of every authorization decision. How you digitize them is not just a DTR workflow question; it is a key component of your end-to-end strategy.

CMS-0057-F compliance and policy digitization offer a unique opportunity to go beyond checking a compliance box, drive provider efficiency, and improve the quality of care across your network.

The top two challenges identified by health plans are deeply intertwined. Digitizing your medical policies is the key to making interoperable workflows truly work. Yet this area of interoperability strategy is not well understood across the industry, and few are equipped to handle the digitization work effectively.

Let’s take a closer look at digitizing medical policies

It is common for health plan policies to exist in static formats, such as PDFs and text on your website. However, your policies must be converted into a compatible digital format to support the DTR workflow.

To complicate matters, the DTR requirement is vague, leading to multiple interpretations of determining clinical appropriateness within the workflow. Many are going down the path of converting a given policy into a set of clinical questionnaires to surface to the provider. But as we’ve seen, that method leads to inaccurate responses more than 30% of the time. To support quality decision-making, protect patient safety, and ease the burden on provider auth submitters, the best practice is to convert that policy to code vs. parsing it into a series of questions prone to gaming.

Tying it all together—how to approach policy digitization and automation

Due to our unique platform and business model, Cohere has spent years working through the ins and outs of medical policy digitization as a core component of our approach to prior authorization. We’ve found that getting this process right is essential if you’re looking to enable automation, improve compliance, or reduce provider abrasion.

Here’s how we approach it:

Step 1: Review how the health plan organizes policies and analyze what is needed to support the authorization workflow.

During policy review, you want a team that combines technical knowledge and clinical expertise. Whereas business rules are more straightforward, medical policy is developed to ensure clinical appropriateness is more nuanced. Purely technical teams without experience in utilization review won’t have the necessary understanding of the provider workflow they are trying to support. ML models can be developed and trained to help parse the policy to pull out components for the digitized version. Still, clinicians should always be involved to ensure that the results satisfy the intent of the clinical policy. Working together, clinical and technology experts can develop a strategy for identifying the indications and non-indications within these policies and a tailored template for digitizing them.

Step 2: Break the policy into components for your digital template.

Each medical policy you’ve identified for digitization needs to be stripped apart and re-composed into the digital format you have developed for interoperable workflows. There are three main components to look out for as you perform this step:

• • Code mapping
  • Effective dates
  • The text of the actual criteria

This is where that digital template gets leveraged. Creating discrete fields to enable mapping codes to the appropriate health plan(s), lines of business, effective date ranges, and states (where applicable) helps to improve the efficiency of determining what policy is in play for a given authorization. The clinical criteria (indications and non-indications) drive the determination of medical necessity.

Step 3: Convert the indications into logic to drive automated decisioning and clinical review.

The clinical criteria you identified in the previous step must be converted to code. This process is time-consuming. Policies often contain a variety of logical constructs, such as “this or that,” “2 or more of the following,” and so forth, that now must be standardized into code. You must convert each indication into a “checkbox” based on Boolean logic to support medical necessity and satisfy policy criteria. The only way to do this at scale is with a code-building tool developed and refined for this specific purpose. Some AI tools can help the conversion process, but a clinician should always review the work to ensure the results support evidence-based care.

As mentioned, many plans choose to convert a policy into a set of questions rather than take this step. The work looks similar and is time-consuming, yet it does not drive automation or reduce provider burden the way a truly codified policy can. While Cohere solutions can support a questionnaire-based workflow, we don’t recommend it as a best practice.

Step 4: Test your digitized policy.

Simulating the review experience (and auto-decisioning, if applicable) is crucial to ensure your digital policy operates as intended. This includes examining the workflow and auditing the test results to tackle any issues that may arise. It is common to encounter challenges at this step, so set aside ample time for quality assurance.

Step 5: Perform ongoing policy updates.

Updating these policies is an ongoing, time-intensive process as well. Looking at CMS Local Coverage Determinations alone—just one component of medical policy—anywhere from 20 to 100 determinations are updated weekly! Supporting these updates on top of shifting contracts, health plan organizational changes, and the myriad other factors that lead to updates is a challenge that many health plans and their vendors are not built to support.

You need to either build and staff a scalable, repeatable process for updating policies or find a vendor with the clinical and technical chops to support it. Part of this process would be creating a policy content library to store the policies and manage version updates. If developed effectively, elements of this library can be leveraged to shave time off future policy digitization efforts, using the code developed for a given indication in another policy that contains the same indication.

Policy digitization and management are in-depth, time-consuming enterprises that will become a permanent fixture of your utilization review efforts. To do it right and set your plan up for success in 2027, we strongly suggest working with teams skilled in managing prior authorization operations and building interoperable workflows.

What Cohere has learned over the years

Cohere’s business model is unique: we offer both delegated clinical specialty management and advanced automation solutions for health plans managing UM in-house. Our team has always been a hybrid of technical and clinical experts working to support a digital-first UM workflow. Since our inception, digitizing health plan policies has been central to our day-to-day workflow. And we always have clinicians review our digitized policies to validate them before use.

Here’s what we’ve built over the years:

• An in-house code-building tool that accelerates policy digitization (3-6 hours per policy)
• QA procedures to have clinicians trained in prior auth validate logic
• A digital content library with over 2,700 fully digitized policies
• A systematic approach to supporting frequent policy updates

To fulfill our vision of supporting an automated, end-to-end workflow, we’ve spent years setting ourselves up well to create clear, discrete, codified instructions for our ML models to parse clinical records for the indications in each codified policy. This allows us to extract member data directly from clinical records, obviating the need to surface clinical questionnaires to provider auth submitters (though we also support those workflows).

Cohere also performs ongoing policy updates for our health plan partners. Because of our novel business model, we are uniquely situated to “lay down the pipes” of a complete interoperability solution and manage our customers’ ongoing policy updates to support long-term success. Working with Cohere means you won’t have to determine the ongoing staffing needs to support digital policy workflows.

Most importantly, Cohere has a track record for delivering an exceptional provider experience. Our solutions have a 93% provider satisfaction rating and 67 NPS among providers.² The goal of CMS-0057-F has always been to improve the provider experience; your compliance solution should share that goal.

Medical policy digitization is a core component of a comprehensive CMS-0057-F compliance strategy.

Ready to set your plan up for long-term success?

Partner with the experts who’ve been doing this for years. Cohere’s hybrid clinical and technical team helps health plans deliver fully compliant, high-performing prior auth solutions that support digital transformation. To learn more, contact us for a demo.

¹WEDI Interoperability and Prior Authorization Survey, April 2025

²Cohere Provider Experience Survey, December 2024