AI-Assisted Coding and Billing
Rural healthcare organizations manage medical coding and billing processes that are part of a broader workflow that includes claims submission, denials and appeals, payer interactions, and payment processing. Medical coding translates patient care into standardized codes that healthcare organizations use for billing and reimbursement. These codes are essential for submitting claims to insurance companies and generating revenue. For more information on billing and reimbursement, see our Rural Healthcare Payment and Reimbursement topic guide.
Medical coding is a complex and time-consuming process. The International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) includes over 70,000 codes, with hundreds added, deleted, or revised each year. Errors in medical coding can lead to revenue cycle delays, revenue loss, insurance claims denials, and increased risk of audits.
AI-assisted coding and billing tools can reduce the burden of common rural challenges, such as workforce turnover and shortages, limited broadband, and resource limitations, by streamlining workflows and applying medical codes with greater accuracy and speed than humans. AI may also help identify and correct billing errors, analyze trends in billing data, and support compliance or regulatory updates.
Examples of Rural Programs Using AI-Assisted Coding and Billing
- Nebraska Methodist Health System uses AI-enabled revenue cycle tools to support medical coding accuracy and streamline claims processing. These tools are integrated at multiple points in the revenue cycle, including pre-bill coding review and post-bill claims management, where they can identify potential documentation and coding issues. The health system has also used AI to automate routine claims workflows and continuously monitor large volumes of information, allowing staff to focus on complex claims, denials, appeals, and other activities that require human judgment. Over two years, these changes freed up the equivalent of 24 full-time employees, who were redeployed to higher-value work, and contributed to increased revenue and improved efficiency.
- Community Health Network (CHN), a nonprofit health system serving central Indiana, uses an AI-enabled platform integrated with its electronic health record (EHR) to support real-time and accurate Hierarchical Condition Category (HCC) coding. The cloud-based AI tool reviews data from clinical notes and EHRs, helping identify relevant diagnoses and coding suggestions. The system evaluates current patient information and suggests appropriate HCC codes that reflect the patient's current health status. This tool has led to increased provider satisfaction, streamlined workflows, and supported more accurate clinical documentation.
Considerations for Implementation
Successful implementation of AI-assisted coding and billing depends on several factors, including technology access, organizational readiness, staff training, human review, privacy and oversight, and monitoring performance.
Technology access. Unreliable broadband Internet access can make it more difficult to access and use cloud-based AI coding and billing tools. Intermittent internet access can cause submission errors, workflow delays, or increase likelihood of challenges with syncing data across sources. AI platforms require EHR integration for the most seamless application. Systems or practices using different EHR systems or those using paper or fax-based workflows may have challenges implementing AI-assisted billing support.
Organizational readiness. It is important to ensure there is transparency around decisions made with AI and data sources. Rural healthcare organizations must also incorporate the context of their healthcare organization into the process. For example, it is important to ensure that any specific billing practices (for example, those for Federally Qualified Health Centers, Rural Health Clinics, and Critical Access Hospitals) are considered in AI-assisted processes. Identifying champions can help foster buy-in. Emphasizing AI as support to free up time, rather than replacing staff, can help address workforce concerns.
Staff training. Staff will require training on AI-assisted features. In-house staff or vendors can provide training and answer questions from providers and administrative staff. Staff must be aware of all AI policies and protections.
Quality assurance and human review. Human review is essential to ensuring the reliability of AI-assisted billing. Healthcare organizations should implement quality assurance processes and conduct pilots using AI to test processes before they are implemented at scale. Human reviews are essential for identifying AI hallucinations, AI-driven over-coding, or other errors. For example, organizations using AI-assisted billing may want to start using AI on straightforward functions, validate those decisions, have clearly written validation and audit procedures, and ensure that humans review complex cases. Additionally, conducting a documentation assessment to identify and address coding and billing inconsistencies will help ensure that AI-assisted analyses make consistent and "correct" decisions.
Privacy and oversight. As billing and coding require that AI platforms have access to patients' protected health information, it is critical to ensure that any platform or cloud services comply with federal privacy policies or data governance policies (for example, encryption). Also, vendors accessing the data should have the appropriate agreements in place (for example, Business Associate Agreements).
Monitoring performance. Rural healthcare organizations often operate on thin financial margins. It is important to monitor return on investment and the performance of AI-assisted billing and coding, such as assessing reductions in billing delays and whether more claims are processed in a specific timeframe. Providers should consider what success looks like within their specific context, whether that is faster reimbursement, reduced burden, improved accuracy, or other outcomes. Establishing metrics and reviewing performance of AI tools can help ensure they deliver value and align with the healthcare organization's priorities (see Module 5: Evaluation Considerations).