The Role of MPI Tools in Health Data Interoperability, Patient Matching

What is an MPI?

According to the Agency for Healthcare Research and Quality (AHRQ), an MPI is a database that "facilitates the identification and linkage of patients' clinical information within a particular institution."

Most MPIs use a patient matching algorithm that leverages patient demographic information, Social Security numbers, and other identifiers extracted from existing medical records to link patient data.

The system assigns each patient a unique patient identifier (UPI), a non-changing alphanumeric key associated with each instance of health data for a patient, according to AHRQ. Providers can use UPIs to link patient data across an institution.

MPIs can exist at an electronic system, facility, enterprise, or health information exchange (HIE) level. Stakeholders may refer to MPIs as Enterprise Master Patient Indexes (EMPIs) at the enterprise or HIE level, according to AHIMA.

In the greater healthcare arena, EMPIs for large integrated delivery networks, health information organizations, accountable care organizations (ACOs), and HIEs have additional layers of identifiers.

Within an organization, different facilities may have different medical record numbers (MRNs) for the same patient. Enterprise identification numbers (EIDs) link various MRNs for one patient to a single umbrella record.

MPI Benefits

According to AHIMA, the prevalence of duplicate records in most hospitals is between five to 10 percent of all stored records. However, health systems that have multiple facilities can see duplicate rates of up to 20 percent.

Duplicate medical records have major consequences for care coordination and clinical decision support. For instance, if a patient's lab results are stored in a different record than the one a provider is accessing, the patient may have to complete unnecessary testing.

While this inconveniences the patient, it also financially strains health systems. Experts estimate that in 2015, the healthcare industry spent $65 billion on lab testing alone, with 20 to 30 percent of that used to conduct unnecessary duplicate tests.

Not to mention, duplicate medical records pose serious safety risks. According to a 2019 survey, 38 percent of healthcare providers have incurred an adverse event in the last two years due to a patient matching issue.

For instance, if a patient's complete medication history is unavailable at the point of care, a provider may prescribe a drug that reacts dangerously with the patient's current medications.

MPI Challenges

While MPIs can improve patient matching, they're not the silver bullet to longitudinal health records. In fact, one in five patient records within the same healthcare system are duplicates, and 50 percent of records are mismatched during transfers, according to ONC.

"Reasons that duplicate records continue to plague healthcare systems include varying methods of matching patient records; departmental system silos; lack of data standardization; lack of policies, procedures, and data ownership; frequently changing demographic data; multiple required data points needed for record matching; and default and null values in key identifying fields," researchers wrote in a 2016 report.

The study used a multisite data set of 398,939 patient records and found that the MPI field that caused the greatest proportion of mismatches was middle name, accounting for 58.30 percent of mismatches. In these instances, one record used the patient's middle initial, and one used the patient's complete middle name.

Patient Social Security number was the second most frequent reason for record mismatch, occurring in 53.54 percent of the duplicate records. Duplicate pairs with Social Security number discrepancies often had a blank entry for one record and a default value for the other.

"The use of more sophisticated technologies is critical to improving patient matching," the researchers pointed out. "However, no amount of advanced technology or increased data capture will completely eliminate human errors. Thus, the establishment of policies and procedures (such as standard naming conventions or search routines) for front-end and back-end staff to follow is foundational for the overall data integrity process."

A 2018 qualitative study mirrors these MPI challenges rooted in organizational processes.

The researchers interviewed health information management (HIM) professionals at an urban nonprofit healthcare facility about challenges related to the MPI that hindered organizational workflows.

First, participants noted that the MPI system's lack of standardized data entry procedures caused inaccurate and incomplete records. They also noted that the organization did not have a consistent data collection process for constantly changing patient demographic data, such as name, address, and phone number.

Additionally, participants reported that the MPI records did not have enough matching data points, which resulted in overlaid records, where one record was attributed to two patients.

The Debate of a National Patient Identifier

Over the past two decades, the House of Representatives and the Department of Health and Human Services (HHS) have debated an NPI that would assign each American citizen a unique number at birth for use across the healthcare system, similar to a Social Security number.

While HIPAA mandated the implementation of a nationwide unique patient identifier, concerns about patient privacy and security prompted the barring of any funding for this endeavor in 1999.

Fundamentally, an NPI would be useful for advancing interoperability and patient matching across the care continuum. For example, a physician may be less likely to prescribe incorrect medicine because two patients share the same name if there was an NPI. Health systems may also incur fewer costs adopting an NPI due to reduced duplicative testing and procedures.

Opponents of an NPI argue that if a patient's identifier number is compromised, the patient's entire health record would be easily accessible. HHS would have to implement strict security measures to ensure the number is secure.

Opponents are also worried about the implementation and cost of an NPI system. While there are no concrete cost estimates, opponents of an NPI system say it could cost an insurmountable sum. The project would require the development of a new health IT system and the issuance of NPIs.

Stakeholders have emphasized that serious medical issues could occur if two patients accidentally received the same number. Additionally, if the provider were to type in the incorrect number or add a medical record to the wrong number, then medications or testing would be inaccurate, and the result could be fatal.

However, after pressure from organizations such as HIMSS and AHIMA, Congress voted to end the ban that bars HHS from funding NPIs in June 2019.

But, in March 2022, Congress reinserted rider language in the Fiscal Year 2023 Labor, Health and Human Services, Education, and Related Agencies (Labor-HHS) Appropriations bill prohibiting HHS from spending federal dollars to adopt an NPI.

Members of Patient ID Now, a coalition of healthcare organizations working to advance a national patient identification strategy, wrote a letter to House and Senate appropriations committee leaders criticizing this decision.

"Despite bipartisan support—led by Representative Bill Foster (D-IL) and Representative Mike Kelly (R-PA)—for the removal of Section 510 in the Labor-HHS appropriations bill, the rider once again was included in the federal budget, continuing its suppression of progress towards addressing the dire issue of patient misidentification within the US healthcare system," coalition members wrote in the letter.

"A narrow interpretation of Section 510 over the past two decades has hindered the ability to create a national strategy on patient identification and matching, endangering patient safety and privacy, and increasing the cost burden to both patients and the healthcare system," they continued.

Advancing MPI and EMPI Tools

While the federal government has not made any further advancements toward creating an NPI, health IT companies continue to advance their MPI and EMPI offerings.

In 2022, Surescripts introduced new patient-matching features through its MPI to enhance the patient and provider experience. The patient matching features supplement the MPI's algorithm with patient reference records, including demographic data such as name, name changes, date of birth, and current and prior addresses.

According to the vendor, the enhancements helped identify additional data for 400,000 patients in one day.

Additionally, nonprofit healthcare industry alliance DirectTrust launched an initiative in 2022 to develop a data standard for a voluntary nationwide patient matching ecosystem.

The Privacy-Enhancing Health Record Locator Service (PEHRLS) Ecosystem Consensus Body aims to profile existing data standards and create new standards for a privacy-enhancing record locator.

The initiative aims to define a model that the private sector could deploy voluntarily or with the support of government funding or encouragement in the future to enable longitudinal health records and cut infrastructure requirements and computing costs.