Taction Software®
Healthcare IT Glossary

What is MPI?
Master Patient Index

A single patient might have records in the EHR, the lab system, the radiology department, the billing platform, the patient portal, and three different outside hospitals. Each system assigns its own medical record number. Without a way to link all those records to the same person, clinicians make decisions with incomplete information, labs get filed under the wrong chart, and claims get denied for mismatched demographics. MPI is the system that ties it all together.

Certifications

Tell Us Your Requirements

Our experts are ready to understand your business goals.

What is 1 + 1 ?

100% confidential & no spam

Definition of MPIHow MPI Works in HealthcareKey MPI Standards and SpecificationsImplementation ConsiderationsHow Taction Helps with MPI

Definition of MPI

MPI, which stands for Master Patient Index, is a database that maintains a unique identifier for every patient registered across a healthcare organization’s systems, linking all of a patient’s records — regardless of which department, facility, or system created them — to a single identity.

The MPI stores core demographic attributes for each patient: name, date of birth, gender, Social Security number (where available), address, phone number, and medical record numbers from each connected system. When a new patient is registered or an existing patient presents at a different location, the MPI determines whether this person already exists in the system — and if so, links the new encounter to the correct existing record.

In a single-facility hospital, the MPI links records across departments — the EHR, the lab information system, the radiology information system, pharmacy, billing, and registration. In a multi-facility health system, the MPI extends across all sites, maintaining a single patient identity across hospitals, clinics, and ambulatory practices. In a health information exchange (HIE), the MPI operates at the community or regional level — matching patients across independently operated organizations that use different systems and different medical record numbers.

The consequences of MPI failure are serious: duplicate records (the same patient appears as two different people, splitting their clinical history) and overlay errors (two different patients are merged into one record, mixing their clinical data). Both create patient safety risks and operational problems.

In simple terms: MPI is the identity backbone of healthcare IT — the system that ensures every record, result, image, and claim gets connected to the right patient.

How MPI Works in Healthcare

The MPI operates through patient matching algorithms that evaluate demographic data to determine whether two records represent the same person.

Real-world MPI workflows include:

ADT message processing. Every admit, discharge, and transfer message contains patient demographics. The MPI validates the patient identity in each ADT message, linking the event to the correct patient record across all connected systems.

Deterministic matching
The simplest approach — two records match if specific fields match exactly. For example: same last name, same date of birth, same SSN = confirmed match. Deterministic matching is fast and highly specific, but brittle — a misspelled name, a transposed digit in the date of birth, or a missing SSN prevents the match even when the records clearly represent the same patient.
Probabilistic matching
The more sophisticated approach — the algorithm assigns weighted scores to each demographic field comparison, accounting for exact matches, partial matches, phonetic similarity, and missing data. The total score determines whether records are a likely match, a possible match (requiring human review), or a non-match. Probabilistic matching handles real-world data quality issues better than deterministic matching, but requires careful threshold tuning to balance match sensitivity against false positive rates.
Referential matching
An emerging approach that compares patient demographics against external reference databases — credit bureaus, insurance databases, state identity registries — to improve matching accuracy. Referential matching can resolve ambiguous cases that probabilistic matching alone cannot.
New patient registration
When a patient arrives and provides demographics, the MPI searches for an existing match. If found, the new encounter links to the existing patient record. If not found, a new MPI entry is created.
Cross-organizational matching
When a patient’s data arrives from an external source — a C-CDA document from another hospital, lab results from a reference lab, claims data from a payer — the MPI matches the incoming patient identity against its existing population. At the HIE level, this cross-organizational matching is what enables query-based exchange: an ED physician queries the HIE for a patient, and the MPI identifies which participating organizations have records for that patient.
Duplicate detection and resolution
MPIs continuously scan for potential duplicate records — two or more records that appear to represent the same patient based on demographic similarity. Identified potential duplicates are queued for human review and manual merge. The merge process consolidates the records and sends HL7 A40 (Merge Patient) messages to all connected systems to update their records.

Key MPI Standards and Specifications

Legacy
IHE Patient Identity Cross-Referencing (PIX)
The IHE PIX profile defines how patient identity cross-references are managed across multiple domains. When a patient is registered in one system, the PIX manager notifies other systems of the cross-reference — enabling each system to maintain its own local patient identifier while linking to the enterprise MPI. PIX supports both HL7v2 and FHIR-based implementations.
Legacy
IHE Patient Demographics Query (PDQ)
The IHE PDQ profile defines how systems query the MPI for patient demographic data. A clinical application sends a query with known demographic parameters (name, DOB, gender), and the MPI returns matching patient records. PDQ is used for patient lookup during registration, order entry, and clinical documentation.
Legacy
HL7 ADT Messages for Identity Management
MPI operations are communicated through specific HL7v2 ADT event types: A28 (Add Person Information), A31 (Update Person Information), A40 (Merge Patient — Merge Patient ID), and A41 (Merge Account). These messages propagate identity changes across all connected systems.
Modern
FHIR Patient Resource
The FHIR Patient resource provides a modern, API-based approach to patient identity management. FHIR patient matching operations ($match) enable probabilistic matching via API calls. The Patient $link operation manages cross-references between patient records across systems.
Legacy
ONC Patient Matching Initiative
ONC has prioritized improving patient matching accuracy through standardization of patient demographic data collection, adoption of patient matching algorithms, and exploration of a national patient identifier (a topic with significant political and privacy implications). Healthcare organizations should monitor ONC guidance on patient matching best practices.
Building an MPI integration? Let’s talk.
Book a free call

Implementation Considerations

MPI implementation involves algorithm selection, data quality management, governance processes, and integration architecture.

HIPAA and identity data. MPI databases contain sensitive personally identifiable information — names, dates of birth, SSNs, addresses. Access to the MPI must be strictly controlled, all queries logged, and data encrypted at rest and in transit.

Data quality at registration is the foundation
The MPI is only as good as the data it receives. If front-desk staff misspell names, enter incorrect dates of birth, or skip optional fields, matching accuracy degrades. Standardize data entry practices — require specific name formatting, validate date of birth against government ID, and capture as many demographic fields as possible. Real-time address standardization (USPS validation) and phone number normalization at registration significantly improve match rates.
Algorithm tuning is an ongoing process
Match thresholds must be tuned to your specific patient population. A threshold set too low produces excessive false positives (merging different patients). A threshold set too high produces excessive false negatives (creating duplicates). Start with conservative thresholds, monitor match/review/reject rates, and adjust iteratively based on outcomes.
Duplicate rate benchmarking
Industry benchmarks suggest that well-managed MPIs maintain duplicate rates below 5%. Many organizations operate with duplicate rates of 10–20% or higher. Measure your duplicate rate regularly and invest in both prevention (better registration data quality) and remediation (duplicate detection and merge workflows).
Enterprise vs. community MPI
An enterprise MPI operates within a single health system. A community or regional MPI operates across organizations within an HIE. Community MPIs face greater matching challenges — inconsistent demographic data standards, different system formats, and no shared internal identifiers. Plan for higher review queues and more complex matching rules.
Merge propagation must be complete
When duplicate records are merged in the MPI, every connected system must process the merge. If the lab system, PACS, billing platform, or patient portal doesn’t process the A40 merge message, those systems retain split records — creating ongoing data integrity problems. Build monitoring to verify merge propagation across all connected systems.

How Taction Helps with MPI

At Taction, our integration team builds and optimizes MPI systems for health systems, hospitals, and HIEs that need accurate, reliable patient identity management across complex multi-system environments.

What we do:

Whether you’re deploying a new MPI, cleaning up an existing one, or building cross-organizational patient matching for an HIE, our healthcare engineering team delivers the identity management precision these critical systems demand.

MPI implementation
We deploy and configure enterprise and community MPI solutions with probabilistic matching algorithms tuned to your patient population and data quality profile.
Patient matching optimization
We analyze your current duplicate rates, tune matching algorithms, and implement data quality improvements at registration to reduce duplicates and improve match accuracy.
Merge workflow automation
We build automated duplicate detection, review queue management, and merge propagation workflows that distribute HL7 A40 messages across all connected systems with verification.
HIE identity management
We build cross-organizational patient matching for health information exchanges — implementing IHE PIX/PDQ profiles and FHIR patient matching operations.
Data quality programs
We implement registration data quality tools — real-time address validation, demographic standardization, duplicate checking at registration — that prevent MPI problems before they occur.

Explore Related Terms

What is Interoperability?What is Consent Management?

Ready to discuss your MPI project?

Schedule a free call

Ready to Discuss Your Project With Us?

Your email address will not be published. Required fields are marked *

What is 1 + 1 ?

What's Next?

Our expert reaches out shortly after receiving your request and analyzing your requirements.

If needed, we sign an NDA to protect your privacy.

We request additional information to better understand and analyze your project.

We schedule a call to discuss your project, goals. and priorities, and provide preliminary feedback.

If you're satisfied, we finalize the agreement and start your project.