By Carm Huntress

Now that CMS has announced their plan for implementing Medicare Advantage rule changes over the next three years, it’s more important than ever to streamline the ways we retrieve and interpret medical records.

The CMS plan includes the removal of more than 2,000 diagnosis codes from the current risk adjustment model, requiring providers to get much more precise with their diagnoses or risk missing out on premiums.

Unfortunately, traditional systems for medical record retrieval, storage, and analysis are not sophisticated enough to adequately meet providers needs under these new rules.

Shortcomings of the current system
Even with the implementation of EHRs, 78% of hospitals are still “often or sometimes” using mail and fax to receive medical records, according to a 2021 report from the Office of the National Coordinator for Health Information Technology.

And when (or if) those records arrive, they’re often incomplete, illegible or contain far too many pages for a provider to read and analyze before the clinical encounter.

Because of this, at-risk providers often go into a clinical encounter with limited-to-no understanding of the patient’s history. This makes it nearly impossible to do accurate and comprehensive HCC coding.

When a provider does identify diagnoses and related treatment plans, these are typically documented in a progress or SOAP note. In fact, up to 70% of clinical value is contained in these notes.

After the clinical encounter, an independent coder reviews the provider’s documentation. Unfortunately, because progress notes are unstructured, there’s no good way for these coders to extract the information they need from the note.

Accurate diagnosis matters more than ever
When a provider incorrectly diagnoses a patient — or presents a diagnosis without a treatment plan — it puts them at risk of payment retraction or fines for any discrepancies found during a Risk Adjustment Data Validation audit.

On the other hand, when a provider misses a diagnosis, it lowers the patient’s RAF score and negatively impacts the payment the provider receives from CMS for that patient’s care.

These consequences were already a significant source of concern for at-risk provider groups. But now that the list of HCC codes is shrinking, it’s more important than ever to code accurately and to catch all diagnoses.

Accelerating the risk adjustment timeline
As health systems are looking for ways to improve their HCC coding processes, a natural starting point is to enable providers to start the risk adjustment process before they ever meet with the patient.

If, for example, a provider could request a comprehensive medical history for a patient and receive it in one easy-to-read document — rather than the tens to hundreds of pages of records a patient will accrue throughout their lifetime — they could walk into a clinical encounter with a list of diagnoses and make sure to address each one while talking to the patient.

This “shift-left” on the timeline - starting risk adjustment pre-encounter - adds enormous value for at-risk providers and also improves outcomes for patients, as their caregivers will be armed with a better understanding of their medical histories and current needs.

Can artificial intelligence offer a solution?
With an enormous amount of data available and the need for a system that can process that data in a variety of formats, artificial intelligence and machine learning are uniquely positioned to change the way we compile a patient’s medical history.

Rather than expecting providers to sift through every document from a patient’s lifetime to look for potential diagnoses, these large language models could complement the work of manual risk adjusters, summarizing key information and pulling out diagnostic codes.

There’s also enormous potential for AI to help coders extract value from the unstructured progress or SOAP notes, including diagnoses, prescriptions, and even specific treatment plan details.

Whatever solution health systems choose, it’s important to remember that the clock is ticking. If risk adjusters want to remain compliant and ensure they’re receiving maximum value for each patient, we must find a medical records solution that supports accurate, compliant coding as soon as possible.

About the Author: Carm Huntress is the founder and CEO of Credo, the leader in automated patient medical record retrieval. Credo is working to radically simplify and update the current medical record and retrieval process to support patients receiving better overall care.