Fall 2025 Issue

Data Integrity: Quality AI Data Relies on HIM Expertise
By Todd Goughnour, MBA, RHIA
For The Record
Vol. 37 No. 4 P. 8

Smarter AI in health care begins with more accurate, complete, and relevant patient information. Health information professionals are uniquely positioned to ensure that data accuracy remains a top priority as health systems implement new technologies including AI. This article outlines three strategies for health information leaders to guide their organizations towards stronger clinical data practices as the foundation for effective and productive use of new AI tools.

Trust is the foundation for realizing the promised outcomes of AI in health care: operational efficiency, improved patient care quality, and stronger reimbursement. That trust, among patients, providers, and staff, depends on accurate patient data.

AI applications demand clean, reliable data to deliver meaningful results. When the information feeding large language models and other AI systems is incomplete or inaccurate, the outcome is poor performance, mistrust among end users, and potential patient safety risks. HIM professionals are uniquely positioned to ensure that data integrity remains a priority as provider organizations adopt new technologies. Their expertise in data governance and health information standards allows them to serve as critical partners in protecting data quality. This article outlines three strategies HIM professionals can use to guide their organizations toward stronger data practices while keeping executive focus on information integrity.

One: Do Your Homework and Tap Resources
A priority for HIM professionals is to fully understand the role of data quality in AI applications. This includes recognizing how data bias, privacy lapses, or gaps in integrity can affect not only patient outcomes but also organizational performance. National studies have demonstrated the risks associated with poor-quality data, from perpetuating inequities to exposing provider organizations to liability.

Fortunately, numerous resources are available. Associations such as The Joint Commission (in partnership with the Coalition for Health AI), the American Medical Association, and the National Academy of Medicine have published frameworks to promote the responsible and ethical use of AI in health care. Industry groups, including the Trustworthy and Responsible AI Network, also provide tools to support best practices.

These resources typically include the following:

• guidelines and principles for responsible AI;
• toolkits for assessing organizational readiness;
• use cases demonstrating both benefits and pitfalls; and
• codes of conduct and certification programs.

By consulting these sources, HIM professionals can ground their recommendations in authoritative guidance and avoid duplicating work that has already been vetted by national bodies. Establishing familiarity with these materials also builds credibility when HIM leaders present to executives or collaborate with IT teams.

Two: Identify Quality Gaps and Communicate Risks
Data quality issues most often originate from three areas:

• System migrations and mergers. Importing a new master person index (MPI) into an EHR or AI application can create mismatched patient records. This risk is heightened during system conversions or organizational mergers.

• Patient self-registration. Allowing patients to create records through portals introduces the risk of duplicates and inconsistent information.

• Manual entry. Inconsistent registration processes and keystroke errors result in duplicates that negatively affect operations, reimbursement, and patient care.

HIM leaders must anticipate these risks and act as data integrity champions. Specific actions include the following:

• Monitor for red flags. A duplicate rate above 1% in the MPI, high levels of technical denials, or frequent clinician complaints about EHR accuracy all signal underlying problems.

• Establish governance. Most organizations have committees focused on data governance or information integrity. HIM leaders should ensure representation on these groups or work with IT leadership to form them.

• Standardize processes. Registration policies and procedures must be consistent across all access points. The use of referential matching tools can significantly reduce MPI duplicates.

Once risks are understood, HIM professionals should make the case for keeping data quality a strategic priority across their leadership teams. The most effective approach is to connect data integrity directly to executive concerns:

• Patient safety. The most serious risk of poor-quality data is compromised patient care. Clinicians cannot make accurate decisions when records are incomplete or fragmented.

• Financial impact. Data errors translate into tangible costs. Financial impacts include malpractice lawsuits, claims denials, and repeated or duplicative medical care.

• Operational performance. Ongoing inefficiencies in registration, scheduling, and billing all trace back to poor data quality, eroding trust across the organization.

By quantifying both the risks and costs, HIM professionals can secure executive commitment and resources for long-term data quality initiatives.

Three: Sustain Ongoing Improvement
AI systems are continuously learning. Without reliable inputs, their outputs degrade quickly, eroding trust and limiting effectiveness. Continuous monitoring, validation, and correction of data are therefore essential. For organizations with multiple EHRs, the challenge is even greater.

Disparate systems increase the risk of fragmented records, inconsistent patient identifiers, and duplicate entries. HIM professionals must collaborate with IT to create processes that reconcile records before data is exposed to clinicians or payers.

Key strategies include the following:

• Proactive cleanup. Errors should be identified and corrected before clinicians use records in patient encounters or claims are submitted to payers.

• Cross-departmental partnerships. Clinicians and revenue cycle staff can serve as valuable allies in spotting mismatched or faulty data. HIM leaders should establish clear reporting workflows and provide training on how to escalate potential issues.

• Education and culture building. Data quality must be treated as a shared organizational responsibility, not solely an HIM function. By positioning themselves as educators, HIM professionals can help foster a culture where staff across departments understand the risks of inaccurate data and take ownership of flagging issues.

Maintaining data quality requires not only processes but also sustained executive attention. HIM leaders should provide regular reports that demonstrate how data integrity efforts improve patient safety, reduce costs, and support AI adoption.

Stay Current
The rapid expansion of AI in health care emphasizes one reality: high-quality data is its foundation. HIM professionals, with their expertise in HIM, are uniquely positioned to lead initiatives that safeguard integrity and reliability.

AI promises significant benefits, but without trusted data, its value cannot be realized. By staying current with industry guidelines, identifying and addressing risks, and building partnerships across departments, HIM professionals can ensure that AI adoption advances patient care, improves operational efficiency, and strengthens financial performance.

Sustained focus on data quality will determine whether AI builds trust and delivers outcomes—or fails due to errors and mistrust.

— Todd Goughnour, MBA, RHIA, is the senior vice president for HIM at e4health. He has worked on successful projects in a variety of scales and settings, with project management experience across the health care continuum from small hospitals to large multientity institutions.