img

How AI Can Help with HCC Coding

83% of healthcare providers expect AI to “eventually” lighten their workload. But most believe this is still a far-off possibility – and don’t realize they could already be using AI to save time and money.

The perfect example is HCC coding: by automating this process, AI can make the Centers for Medicare & Medicaid Service (CMS) risk adjustment process faster, easier, and more effective.

This article explains exactly how your organization could do that – and the powerful benefits it will bring.

AI and HCC Coding: An Overview

Why Do Providers Want to Automate Coding?

Every value-based care (VBC) provider understands the importance of accurate risk adjustment, but HCC coding presents several common problems:

  • Time: Recording, documenting, verifying, and submitting HCC coding takes a lot of time – and physicians are already overworked.
  • Accuracy: With more than 72,000 ICD-10 codes that map to 86 HCC codes within the CMS-HCC risk adjustment model, errors are common and can be very costly for providers. Equally, many providers lack specialist knowledge of a particular medical area, and therefore choose the wrong code for a particular condition.
  • Documentation: The CMS requires providers to maintain accurate documentation of all conditions included in their HCC coding, which creates an extra hurdle for organizations that struggle with fragmented IT systems and data silos.

The net result? Manual HCC coding is time-consuming and difficult. It takes providers’ attention away from patients – and doesn’t even produce optimal coding. It’s therefore unsurprising that a growing number of providers want to automate the process; the question is how they can achieve that goal.

How Does AI Support HCC Coding?

The basic process of automating HCC coding using AI requires the following three steps:

  • Input Data: The system ingests a large quantity of patient data. This will typically be from a range of sources, existing electronic health records (EHR), handwritten notes, and health information exchanges (HIEs).
  • Analyze Data: The system uses natural-language processing (NLP) algorithms that are trained to recognize patterns within the data and can identify relevant diagnoses.
  • Recommend Codes: The system can then match these diagnoses to the relevant HCC codes and produce coding recommendations for the provider.

Importantly, this is not the same as completing coding for the provider; it still allows the provider to assess and verify the selected codes – ensuring any inaccurate or irrelevant codes are ignored.

What are the Benefits?

Removing the laborious manual coding process helps VBC providers:

  • Save Time: Physicians save hours each week that used to be spent on repetitive HCC coding tasks. This time can be reallocated to patient care and improving the patient experience.
  • Reduce Costs: Providers no longer need to hire human HCC coders to shoulder the excess coding workload, reducing overall staffing costs.
  • Improve Accuracy: Human error is eliminated and many commonly missed or overlooked diagnoses are included in the coding.

All of which suggests every provider should be pursuing AI solutions – so why aren’t they?

3 Common Barriers to AI Adoption for Risk Adjustment

Despite the clear benefits risk adjustment AI offers, many providers struggle with:

1. Data Gaps

AI-based solutions require high volumes of accurate, reliable data – but many providers struggle to access that information. This could be the result of actual data gaps, such as incomplete medical histories or poorly recorded and documented diagnoses. Equally, it could be a technical problem where data is stored across multiple IT systems or pieces of software that do not interoperate – creating persistent data silos.

As a result, many providers fear they will be unable to feed AI-based solutions the data they need to produce accurate coding recommendations.

2. Usability

Automating HCC coding can save a lot of time – but only if providers can easily access the software’s recommendations. Many AI solutions are difficult to integrate into existing IT systems, such as EHRs, and do not easily fit into the provider’s workflow. Providers therefore spend almost as much time trying to access coding recommendations as they would doing the coding themselves.

3. Trust

HCC coding determines a VBC provider’s Medicare reimbursements, which makes it a highly sensitive task. Many providers fear missing out on funding they are entitled to – and don’t trust a machine to be responsible for this crucial function.

How HCC Assistant Solves These Problems

HCC Assistant is an NLP-based tool that helps VBC providers:

  • Close Data Gaps: By ingesting both structured and unstructured data, the tool enables providers to create a single source of truth for patients’ medical histories. This eliminates data silos, increases visibility, and ultimately ensures you have the maximum amount of reliable data to produce accurate HCC coding.
  • Seamless Integration: With easy integration through all popular EHRs, HCC Assistant fits directly into your providers’ existing workflows. From coding recommendations to automatic care plans, providers can quickly click the HCC Assistant app within their EHR and get the information they need in seconds.
  • Provider Confirmation: HCC Assistant presents recommendations, leaving the provider to sign-off on all codes and ensure you are never simply “blind” to the output. This means you can trust the system without giving up control of your risk adjustment process.

Ultimately, this has helped providers across the country save countless hours and increase their risk adjustment factor (RAF) scores by 35%.

Want to see it in action?

Book a Demo

Click to rate this post!
[Total: 0 Average: 0]