From Burden to Confidence: CHRISTUS Health’s Path to Better HCC Coding

Key Takeaways: 

• CHRISTUS Health wanted to break free from labor-heavy Hierarchical Condition Category (HCC) workflows that slowed teams and couldn’t scale to meet the organization’s accuracy, compliance and value-based care (VBC) needs. 
• Partnering with Premier’s Stanson Health, the organization embedded CodingCare, a tech-enabled solution providing real-time, artificial intelligence (AI)-driven insights to help ensure accurate HCC coding before and after patient encounters.
• CHRISTUS Health eliminated thousands of unnecessary reviews, strengthened compliance and transformed HCC coding into a clear, controlled and confident process.

In many health systems, the path to accurate Hierarchical Condition Category (HCC) coding documentation begins long before a patient ever steps into a clinic. Conditions evolve. Documentation varies across clinicians and specialties. And workflows differ by setting. 

For CHRISTUS Health, a mission-driven, nonprofit health system that spans more than 600 primary and specialty care locations and serves communities across multiple counties, accurate, compliant HCC coding is more than a financial necessity. It is foundational to success in a value-based care model and fulfillment of the organization's commitment to patients.

But like many large health systems, CHRISTUS Health once found post-visit chart reviews exhausting, retrospective processes inefficient, and audit risk ever-present.

Today, the obstacles that once loomed large are routine checkpoints.

By partnering with Premier’s Stanson Health and leveraging its artificial intelligence (AI)-enabled CodingCare solution, CHRISTUS Health transformed its risk adjustment program from reactive and resource-heavy to proactive, precise and audit-ready.

Building Confidence in AI Without Compromising Compliance

When it comes to accurate HCC coding, Jeannine Bumford, Director of Risk Adjustment Operations at CHRISTUS Health, is clear: AI alone isn’t enough. Governance, customization and compliance must come first.

“You absolutely need to have governance and structure and be comfortable with what you’re implementing,” she said. “If a vendor only talks about ROI and not about compliance or OIG (Office of Inspector General) high-risk conditions — run.”

That compliance-first mindset made Premier a strong fit.

“What I love about Premier’s Stanson Health CodingCare solution, specifically with the pre-visit prep, is the confidence it gives me in what we're surfacing to physicians,” Bumford said. “One of my biggest concerns was whether the AI would pick up conditions inaccurately. Premier has always been transparent with their logic and very conservative in their approach."

Pre-Visit: Smarter Preparation, Fewer Surprises

CHRISTUS Health leverages its electronic health record (EHR) software, Epic, and Premier’s AI-enabled solutions to create real-time, workflow-integrated nudges that feel natural, not disruptive.

Pre-visit, CodingCare scans EHRs to identify missing or inaccurate HCC codes before the patient arrives. Instant notifications surface only what’s relevant, allowing clinical documentation improvement (CDI) teams to proactively address opportunities. 

Rather than flooding providers with noise, CHRISTUS Health made deliberate decisions about alert suppression.

“We did a lot of suppression,” Bumford said. “We don’t even surface cancers anymore in our OPAs (OurPractice Advisories). We’re very mindful about acute conditions. What we show providers are the chronic conditions that truly need to be addressed.”

The result? Fewer — and better — queries.

“With that suppression, we actually have fewer queries now,” she said. “What we’re surfacing are real chronic conditions that need to be addressed. It’s a win-win.”

As the CHRISTUS team’s confidence in CodingCare accuracy grew, so did automation.

“At the beginning, we reviewed 100 percent of what CodingCare flagged,” Bumford said. “But as we built trust, we were able to automate more. It’s gone so well that we’re now moving toward automating much of the pre-visit workflow. At the start, I wanted to check every single item — but CodingCare has earned my trust.”

“Moving toward automation has been a huge win for efficiency.”

Post-Visit: Reducing Review Volume, Increasing Accuracy

Even with strong pre-visit and concurrent support, post-visit review remains critical, especially in today’s risk adjustment data validation (RADV) audit environment.

CodingCare analyzes every documented HCC after the encounter closes, validating whether sufficient MEAT (Monitoring, Evaluation, Assessment, Treatment) documentation supports each code. Visits with strong documentation can bypass manual review; those with gaps are prioritized.

The impact has been substantial: Approximately 20 percent of visits no longer require manual review, eliminating thousands of unnecessary touches and equating to roughly two full-time equivalents in workload savings.

“We were able to bypass review on around 13,000 encounters, because CodingCare had already confirmed the HCC conditions were supported,” Bumford noted.

“That’s 13,000 charts my team didn’t have to review.”

Today, only about 30 percent of charts now require human review, allowing CHRISTUS Health’s certified coders to focus on high-risk or complex cases.

At the same time, CodingCare identifies missed HCCs where documentation exists but the diagnosis was not added to the claim. Over seven months, CHRISTUS Health added more than 400 codes and achieved an 83 percent acceptance rate.

“I haven’t had to increase staff,” Bumford said. “Now, my team can concentrate on the work that truly needs their eyes and certification.”

Embedding Risk Adjustment Data Validation Readiness into Everyday Workflows

AI has turned readiness into a daily practice, not a last-minute scramble.

“We’ve turned hard lessons from past RADV audits into a proactive, structured approach,” Bumford said.

With 100 percent of notes scrubbed through CodingCare, CHRISTUS Health implemented a grading workflow to instantly triage documentation strength and extrapolation risk. A financial error calculator mirrors CMS methodology, allowing the team to forecast potential exposure in real time.

“We’re no longer waiting for audit notices,” said Lisa Reyes, CPC, CRC, CDEO, Risk Adjustment Quality Manager at CHRISTUS. “We’re anticipating and neutralizing risk before it escalates.”

As AI handles routine reviews, roles have evolved:

• Coders increasingly function as auditors.
• Auditors shift toward governance and systemic oversight.
• Educators focus on closing documentation gaps.

From Hesitation to High-Trust Clinical Adoption

CHRISTUS Health’s journey with Premier and CodingCare reflects a broader shift in healthcare risk adjustment:

• From retrospective cleanup to concurrent precision.
• From manual review to targeted, AI-supported validation.
• From audit anxiety to audit readiness.
• From skepticism to trust.

“You can have the best product in the world, but if you don’t have physician buy-in, it’s dead on arrival,” said Bumford.

The health system embedded CodingCare into CDI and coding workflows, suppressing noise, elevating human expertise and reinforcing governance at every level. As a result, CHRISTUS Health transformed HCC optimization from a reactive exercise into a proactive, defensible strategy.

The dread of post-visit chart review has been replaced with something else entirely: clarity, control and confidence — and, most importantly, getting coding right the first time, at the point of care.

“Throughout the implementation, Premier customized everything to our needs,” Bumford said. “That flexibility from day one was extremely important to us, and they absolutely delivered.”

For more:

Learn about what Premier's Stanson Health can do for your organization.

Get the insights you need to stay ahead in healthcare: Subscribe to Premier’s Power Rankings newsletter and receive our experts’ original content in your inbox once a month.