Revenue Cycle Management: The Case for Starting with Coding

Healthcare spending reached an estimated $5.3 trillion in 2024, and yet health care organizations still operate on thin margins and leak revenue. They can’t afford inefficiencies. But revenue cycle management processes struggle to keep pace with the volume and complexity of modern healthcare. 

Revenue cycle management, the process of turning clinical services into payment, determines whether organizations thrive or fail financially. Within this complex process, medical coding stands out as the single highest-impact area for improvement. 

Coding determines what gets billed, how accurately, how fast, and whether it’s compliant. It shapes denial rates, audits, A/R days, and provider workload. No other RCM function touches as many levers. Understanding why starts with understanding how revenue cycle management works.

What is Revenue Cycle Management?

The revenue cycle in healthcare [LINK] spans every step of a patient visit, from the initial patient contact to the final payment collection. Here’s a breakdown:

Front-end operations include patient registration, insurance verification, and prior authorization.

Mid-cycle functions include clinical documentation and medical coding, where providers document patient encounters. Depending on the organization, professional coders and/or providers usually code charts. This process involves translating the details of provider-patient encounters into standardized alphanumeric codes used by payers to reimburse health systems for care delivery.

Back-end processes involve managing claim submission, payment posting, denial management, and collections.

Medical coding sits at the center of this continuum, serving as the connection between clinical care and revenue recognition. 

Why Coding Represents the Main Revenue Opportunity in RCM

When considering how to improve revenue cycle management [LINK], medical coding delivers an outsized financial impact compared to improvements in other areas because it’s so closely connected to the entire revenue cycle. 

Nearly 39% of hospitals operated with negative margins in 2023, with 22% posting margins below -5%. Those margins leave zero room for revenue leakage. Yet coding, the function that determines what gets billed and how accurately, remains chronically under-resourced. In many high-volume specialties, providers code their own charts and coders review only about 30% of them, according to Arintra’s internal data. The remaining 70% are not reviewed, creating a massive blind spot in the revenue cycle.

When professional coders don’t code or review charts, several things are more likely:

• Missed diagnosis codes that support medical necessity
• Overlooked procedures that warrant separate billing
• Failure to capture the complexity that justifies upcoding
• Variability in coding across dozens or hundreds of providers

These risks grow as payers demand greater specificity. The shift from using general codes to more granular codes, like the HCC transition from V24 to V28, raises the bar for accurate documentation. When providers use general codes instead of specific ones, they leave revenue behind.

A primary care visit may only represent $150 in potential revenue, but even small coding gaps compound quickly across thousands of monthly encounters. Claims denials alone, for example, cost hospitals about $262 billion a year, and each denied claim requires $25-$181 to rework and resubmit. 

These revenue cycle management challenges [LINK] compound because coding also impacts the effectiveness of other revenue cycle stages. For example, accurate coding reduces denials, meaning that back-end teams spend less time on appeals. Timely coding also accelerates claim submission, shortening accounts receivable (A/R) days.

The Human Stakes: Impact on the Healthcare Workforce

Most revenue cycle departments employ just a handful of professional coders who are only able to review a small percentage of charts in high-volume healthcare settings. It’s the most difficult revenue cycle role to fill. These professionals handle enormous workloads while staying current with frequently changing guidelines and payer-specific requirements. For example, the move from ICD-9 to ICD-10 expanded the code set from 14,000 to 70,000 options. 

Each payer also adds its own documentation requirements, while every healthcare system has its own business requirements involving coding adjustments. Medicare policies also differ from Medicaid, which varies by state. It’s a complex job, and many coding teams are overwhelmed with high chart volumes.

Coding and revenue cycle management [LINK] challenges don’t stay contained in one department. They spill over to providers, who spend around 16 hours a week outside of working hours doing administrative work, including coding. Instead of spending their off time resting and renewing their minds for the difficult work they do, many spend weekends answering questions from the billing department about their coding choices. Provider-led coding also creates other risks, such as undercoding that reduces compensation or overcoding that creates compliance exposure.

These workforce challenges point to a fundamental truth in revenue cycle management: organizations can’t simply hire their way out of coding shortages. They must rethink how coding work gets done.

Why Traditional Solutions Fall Short

Manual coding depends entirely on human coders reviewing each chart. This works relatively well when there are enough coders to handle chart volume, but it scales poorly. Health systems can’t hire quickly enough to keep up with growing chart volumes. The cost of hiring enough coders to review every single chart, especially in high-volume, low-dollar specialties, becomes prohibitive.

Outsourcing coding to third-party coders moves the burden to external vendors but entails new problems. Organizations lose direct control over coding quality, generally don’t receive clinical documentation feedback, and lose insight into revenue integrity. Data security concerns are another issue with this approach, as protected health information is sent outside internal systems. Perhaps most critically, outsourcing still doesn’t solve the underlying capacity issue.

Early attempts at revenue cycle automation [LINK] via computer-assisted coding (CAC) emerged in the 1990s as a first attempt to meet scaling challenges in coding. But CAC tools:

• Are not proficient at reading unstructured text (such as physician notes), which limits their ability to produce accurate and complete coding.
• Essentially operate as black boxes, providing no explanation for their code suggestions. This can complicate audits and compliance efforts. 
• Still requires human review of every chart

CAC has improved coder productivity, but it doesn’t fundamentally alter the scalability issue. For decades, healthcare stayed stuck in this cycle. Not enough coders, no way to scale. Then generative AI broke the stalemate.

The Next Frontier: Autonomous Medical Coding

Recent breakthroughs in generative AI have redefined the future of revenue cycle management [LINK], creating something CAC never could: tools that code charts autonomously. This technology can interpret clinical documentation with human-like fluency, synthesizing both structured and unstructured physician notes. 

Unlike CAC-based code suggestion tools, autonomous coding is capable of coding most charts independently. Human coders only need to review particularly complex charts, which autonomous coding can recognize and route with specific instructions about what to check. Delivering on this promise requires more than throwing AI at coding. It demands purpose-built architecture, specialized clinical reasoning, and seamless EHR integration. 

Arintra’s autonomous coding solution was built to meet that bar. Built from the ground up with deep clinical context, it integrates directly within EHR systems like Epic or Athena, preserving existing workflows while coding charts autonomously. Key capabilities include:

• Coding ICD, CPT-10, and HCC codes in minutes
• Native EHR integration that maintains data integrity
• Modular architecture with specialized reasoning agents that handle specific concepts independently
• Rapid coverage expansion and guideline updates without retraining the entire system

This architecture shows what’s now possible with autonomous coding, especially in high-volume settings like ambulatory care. But what does that potential look like across the full revenue cycle?

How Autonomous Coding Transforms RCM 

Autonomous medical coding affects every stage of the revenue cycle management workflow [LINK]. Here’s the breakdown at each stage. We’ll start at mid-cycle, as it’s the most important for coding:

Mid-Cycle Transformation

Autonomous coding handles the vast majority of charts without human intervention, completing ICD-10, CPT, and HCC coding in minutes while maintaining full alignment with payer-specific requirements. Specialized reasoning agents process different aspects of each encounter: interpreting free-text notes, resolving clinical ambiguity, mapping symptoms to conditions, and identifying patterns aligned with payer rules and denial trends. These agents work as a coordinated system, each contributing domain-specific analysis while maintaining complete clinical context.

Autonomous coding delivers something else: consistency. When dozens or hundreds of providers code their own charts, organizations end up with the same number of different approaches to coding. Autonomous coding consolidates those decisions into a single, auditable methodology. That uniformity strengthens compliance, since the logic is programmed once and applied the same way every time. Consistency also makes problems easier to fix. When a denial trend emerges, the solution happens in one place, as opposed to retraining 50 different providers.

Clinical Documentation Improvement

Autonomous coding strengthens documentation before claims go out the door. When the system identifies gaps in clinical notes, it flags them and delivers provider-specific feedback. This creates a continuous CDI loop, catching missing specificity or unsupported diagnoses at the source. Providers engage with this feedback because it connects directly to reimbursement. Instead of generic queries, providers see exactly how documentation gaps affect their coding and revenue.

Back-End Impact

Back-end operations become more efficient when autonomous coding feeds them cleaner claims. Denial rates drop substantially. Every coding decision includes a clear, explainable trail showing the documentation supporting each code, enabling revenue cycle teams to challenge denials more effectively.

System-Wide Results

Faster coding enables quicker claim submission, while fewer denials shorten A/R days. This is revenue cycle management optimization [LINK] in practice: coders have the capacity to shift to higher-value work, physicians can reclaim administrative time, and front-end staff get clearer guidance about the information needed for successful claim submission. So how do you know if your organization is ready to capture these gains?

Implementing RCM Transformation: Readiness Checklist

Organizations developing revenue cycle management strategies [LINK] for autonomous medical coding should evaluate their readiness across several dimensions. Here are some questions to consider.

Scale and Infrastructure

• Do you have 25 or more providers generating consistent chart volume?
• Does your organization use an EHR that allows native integration, such as Epic or Athena?
• Can your IT team support a low-lift integration that preserves data integrity?

Current Coding Operations

• Do you code in-house, allowing you to redeploy staff to complex cases, provider education, and revenue integrity work?
• If you currently outsource, are you prepared to bring coding back in-house for greater control and visibility?
• Do you manage high-volume ambulatory specialties like primary care or urgent care where most unreviewed charts accumulate?

Specialty Selection

• Which specialty carries your longest A/R days or highest denial rates?
• Which specialty has the highest physician dissatisfaction with coding?
• Does that specialty's team have the capacity to support vendor implementation?

Vendor Validation

• Does your chosen starting specialty include a breadth of code types (ICD, CPT, E/M) and reasonable complexity?
• Will this specialty test whether your vendor can scale into additional areas later?

Financial Metrics

• What is your denial rate by specialty?
• How many A/R days do you average?
• What percentage of charts receive professional coder review?

These baseline metrics enable accurate ROI modeling and provide concrete benchmarks for measuring improvement. One health system’s experience shows what’s possible when autonomous coding implementation is done right.

Case in Point: Mercyhealth's RCM Reinvention

Health systems using Arintra's autonomous coding platform report consistent results: 5.1% revenue uplift, 43% reduction in denials, and 12% improvement in A/R days. Mercyhealth's experience shows how these gains unfold in the real world.

Mercyhealth faced challenges familiar in healthcare: coding backlogs that delayed billing, strained their small coding team, and created revenue recognition problems. As a growing health system, they couldn't hire coders fast enough to keep pace with increasing patient volumes. The situation created both financial pressure and friction.

The organization implemented Arintra's autonomous medical coding solution and saw gains across its revenue cycle. Native Epic integration meant zero workflow changes for their team. Charts moved through the same processes, but autonomous coding handled the bulk of routine encounters while maintaining complete audit trails.

Within months, Mercyhealth achieved 5% revenue uplift while reducing work queue aging by 50%. They established comprehensive coding coverage without adding staff, enabling their existing team to focus on complex cases and revenue integrity projects. The compliance benefits proved equally valuable. Every autonomous coding decision involved clear documentation and reasoning, creating audit trails that satisfied both internal compliance requirements and external payer scrutiny.

As Mercyhealth’s VP of Revenue Cycle explained: “With Arintra, we’re ensuring compliance, simplifying our auditing with a solid reporting structure in place, and increasing revenue that depends on the documentation.” Mercyhealth’s story is a preview of where revenue cycle management is headed.

The Future of RCM Is Autonomous

Autonomous medical coding, powered by generative AI and large language models, represents a fundamental breakthrough in healthcare revenue cycle automation [LINK]. It solves the scalability, accuracy, and efficiency challenges that have constrained revenue cycle management for decades.

The financial stakes have never been higher. With 40% of hospitals operating with negative margins and more than 700 rural hospitals facing closure risk, organizations cannot afford to leave revenue on the table. Medical coding represents the highest-leverage point for improvement. Get coding right, and the benefits impact the entire revenue cycle.

Health systems that invest in autonomous coding today are building the financial resilience healthcare needs for tomorrow. They’re capturing every dollar earned, accelerating cash flow, ensuring compliance, and freeing their people to focus on delivering excellent patient care.

Ready to Accelerate Revenue Recognition?

Learn how Arintra's autonomous medical coding platform helps leading health systems reduce A/R days, cut denials, and achieve compliant reimbursement, all within Epic and Athena.

Book a demo. 

‍

‍