How to Ensure Data Integrity During Health System Change

Hospital merger and acquisition activity continues to rise as healthcare leaders seek to increase value, scale operations, and improve patient outcomes. Yet behind every merger lies a crucial challenge: protecting the accuracy of the master person index (MPI). When multiple systems merge, duplicate records and inconsistent data jeopardize care coordination and operational efficiency. 
 
Effective MPI management begins long before systems merge. Early and rigorous planning allows teams to identify potential data integrity issues, implement corrective measures, and prevent costly post-conversion cleanup. Planning not only ensures compliance with regulatory standards but also safeguards continuity of care by maintaining accurate clinical, financial and operational information across systems. 

For health information (HI) professionals, success depends on striking the right balance between technology and human judgment. With thoughtful planning and a collaborative approach, organizations can achieve data integrity that sustains both clinical and business performance. 

MPI Conversion Planning: Best Practices

Successful MPI conversion requires deliberate preparation, thorough validation, and clear alignment across clinical, financial, and technical teams. The goal is to ensure patient identity data is accurate, complete, and ready for use in the target system without introducing errors or workflow disruptions. 

These best practices form the foundation of a reliable MPI conversion: 

• Capture every available identifier associated with each record including legacy and alternate medical record numbers to preserve demographic and clinical linkages throughout the conversion process. 
• Confirm that leading zeros and character formats are maintained during transfers. Even minor formatting changes can create duplicates or unintended overlays once loaded into the new system. 
• Determine in advance how to handle deceased patients, test accounts, and outdated records. This prevents unnecessary volume in the conversion and aligns practices with regulatory and organizational requirements. 
• If moving from a person-centered environment to a patient-centered environment, create mapping rules that correctly link identifiers and visit history without interrupting daily operations. 
• Conduct frequency analyses on names, dates of birth, addresses, and phone numbers to identify patterns or inconsistencies that may interfere with accurate record matching. 
• Apply sampling and error analysis, along with system-to-system comparisons, to validate data and resolve discrepancies before they enter production. 
• Set clear expectations for duplicate cleanup. Establish how many duplicates are acceptable, how they will be prioritized, and who handles escalations so teams know what must be resolved before go-live. 

Organizations that follow these practices significantly reduce the likelihood of costly rework, delayed timelines, and identity errors that affect patient care and revenue integrity. Proper planning ensures the MPI enters the new environment as a dependable asset rather than a risk that must be corrected later.

Integrating Data Successfully with AI 

Successful data integration requires people, processes, and technology working together. Artificial intelligence (AI) can support this work by increasing confidence in matching decisions and reducing manual effort, but AI alone is not enough. Many organizations are learning this firsthand. A 2024 Black Book Research survey found that 92 percent of more than 900 healthcare IT leaders said their AI systems were “not accurate or actionable enough for clinical use.” Poor or incomplete data remains a major barrier, and without clean inputs and strong governance, AI cannot deliver reliable results.  

To address this reality, for example, Harris Data Integrity Solutions introduced the concept of Caring Algorithms, which are AI systems developed and monitored with ethical principles, transparency, and empathy. Human oversight, often referred to as the “human in the loop” (HITL), ensures that AI-driven insights are validated by professionals who understand clinical and operational context. When combined with robust governance, such algorithms enhance both efficiency and accountability. Technology identifies likely matches and patterns, while human experts manage exceptions, validate accuracy, and provide critical feedback that strengthens AI models over time.  

Harris Data Integrity Solutions analyzed 137,080 potential patient record duplicates processed through automated matching, and a meaningful portion of the recommendations required manual correction due to conflicting or incomplete data. While automation accelerates review, these discrepancies confirm that it cannot independently resolve every identity decision and may introduce errors such as overlays. By understanding where automation falls short, organizations can determine which records warrant expert review and apply algorithms more effectively across the MPI. This partnership between technology and human judgment strengthens MPI accuracy, reduces avoidable rework, and ensures that AI supports, rather than replaces, responsible decision making. 

The following chart illustrates how those 137,080 records were resolved, showing that while automation successfully handles many matches, a significant percentage required human correction due to conflicting or incomplete data.

The success of AI in healthcare depends on governance. Clear policies ensure ethical use, protect privacy, and maintain compliance. Effective governance includes transparency in how algorithms make decisions, accountability through cross-functional oversight committees, and continuous improvement through regular audits and retraining.

Sustaining MPI Health Post-Merger

Once a merger is complete, maintaining MPI integrity shifts from a project milestone to a core operational responsibility. Organizations must continuously monitor and resolve duplicate records, update matching logic as data and workflows evolve, and conduct routine audits to sustain accuracy. Aligning internal processes with national frameworks from AHIMA, Project US@, and Patient ID Now ensures consistency and helps organizations remain aligned with evolving best practices. 

Equally essential is empowering teams who collect patient data at the point of entry. Registration staff form the frontline of MPI accuracy, and ongoing education equips them to prevent errors before they propagate through the system. Clear communication and feedback loops reinforce how their work supports patient safety, revenue integrity, and overall data quality. Through a combination of strong governance, optimized technology, and engaged frontline staff, organizations can sustain MPI health long after a merger is complete. 

As healthcare mergers accelerate, data integrity will remain a defining factor of success. The most resilient organizations will be those that combine advanced AI with disciplined human expertise. HI professionals will continue to serve as stewards of patient identity, ensuring that data feeding these systems remains accurate and complete. The future of patient identification is not a choice between humans or machines. It is a partnership between both, working together to ensure every patient’s record tells one accurate, trusted story. 

Candice Neisen, MS, RHIA, is Director, Professional Services, and Mackenzie Higgins, RHIA, is a strategic growth manager, both at Harris Data Integrity Solutions.