2.2 Master Patient Index (MPI) & Demographic Management

The Master Patient Index (MPI) is the cornerstone of health information management (HIM) and data integrity within any healthcare organization. The MPI is an enterprise-wide database that maintains a permanent, accurate historical record of every patient registered across a health system. By assigning a single, unique Medical Record Number (MRN) to each individual, the MPI links all clinical documentation, laboratory findings, radiological images, billing records, and administrative encounters to one distinct patient identity. When multiple facilities—such as hospitals, outpatient clinics, urgent care centers, and specialty groups—share a unified database, the system is known as an Enterprise Master Patient Index (EMPI). For the CEHRS, maintaining MPI integrity is paramount; errors in patient identity matching can result in catastrophic clinical mistakes, compromised privacy, and severe financial losses.

Core Demographic Data Elements in the MPI

To establish a unique record and prevent identity confusion, the MPI relies on a standardized core data set. The American Health Information Management Association (AHIMA) and federal health IT standards recommend including the following primary data elements within every MPI entry:

• Internal Unique Identifier: The primary Medical Record Number (MRN) automatically generated by the EHR.
• Full Legal Name: Legal first name, middle name, legal last name, and any applicable suffixes (Jr., III).
• Aliases and Maiden Names: Historical names, maiden names, or social aliases used for cross-referencing.
• Date of Birth: Mandatory 8-digit numerical date (MM/DD/YYYY).
• Social Security Number (SSN): A key unique identifier, though fuzzy matching algorithms accommodate missing SSNs.
• Gender and Biological Sex: Current biological sex at birth alongside administrative gender.
• Mother's Maiden Name: Historically used as a secondary identity verification attribute.
• Address and Contact Data: Primary residential address, phone numbers, and emergency contact details.
• Account / Encounter Numbers: Unique episode-of-care numbers generated for specific visits, which map back to the single master MRN.

Identity Matching Algorithms: Deterministic vs. Probabilistic

EHR systems employ sophisticated algorithms during registration to compare newly entered patient data against existing MPI records to prevent duplicate creation.

• Deterministic Matching: The simplest matching methodology, which requires an exact, character-for-character match across specific key fields (e.g., exact match on SSN, Date of Birth, and Last Name). While deterministic matching produces near-zero false positives, it frequently fails when minor typographical errors, hyphenated names, or missing SSNs occur, resulting in undetected duplicate records.
• Probabilistic Matching: Advanced statistical algorithms that evaluate multiple demographic fields simultaneously, assigning weighted mathematical scores based on the likelihood that two records belong to the same person. Probabilistic matching handles phonetic variations (e.g., "Kathleen" vs. "Catherine"), transposed DOB digits (e.g., 04/12/1980 vs. 04/21/1980), name inversions, and missing data elements. When a match score falls within a designated threshold (e.g., 85–95% probability), the EHR alerts the CEHRS to manually review the prospective match.

Critical MPI Integrity Errors: Duplicates, Overlays, and Overlaps

Errors in demographic management create three distinct, high-risk data integrity anomalies:

1. Duplicate Records: Occurs when a single patient is assigned two or more distinct MRNs within the same facility (e.g., MRN 10045 and MRN 88301 both belong to John Doe). Duplicate records split the patient's medical history; a physician reviewing MRN 88301 may fail to see severe drug allergies, chronic conditions, or recent diagnostic results documented under MRN 10045.
2. Overlay Errors: The most dangerous MPI error, occurring when one patient's record is mistakenly overwritten or merged into another completely different patient's MRN. For example, if "Jane A. Smith" is registered under "Jane B. Smith's" MRN, Jane A.'s clinical notes, lab results, and prescriptions become mixed into Jane B.'s chart. This creates immediate clinical harm (e.g., administering wrong medication) and constitutes a major HIPAA privacy violation.
3. Overlap Errors: Occurs in enterprise systems when a patient has separate records created across different facilities within the same healthcare organization (e.g., Hospital A and Hospital B) that have not been linked by the EMPI.

Best Practices for Search, Resolution, and HIM Reconciliation

Preventing MPI errors requires strict adherence to front-office search protocols and specialized HIM reconciliation workflows:

• Mandatory Pre-Registration Search: Before clicking "Create New Patient," the CEHRS must perform a comprehensive multi-parameter search. Standard protocol mandates searching by Date of Birth plus the first three letters of the legal last name.
• Checking Alias and Maiden Name Fields: Searching former names, hyphenated names, or previous addresses if an initial search yields no results.
• HIM Merge/Unmerge Workflows: Front-office staff should never independently merge records. When a potential duplicate or overlay is identified, it must be flagged for the Health Information Management (HIM) department. HIM specialists analyze audit trails, compare historic documents, execute formal record merges for duplicates, or conduct complex record "unmerges" to separate overlaid charts.