A fundamentally different approach to patient matching
No matter what patient matching technology you use, and no matter what vendor it is from, it uses fundamentally the same approach to patient matching as every other technology. This is true whether you use the built-in master patient index (MPI) module in your EHR from Epic®, Cerner®, or another vendor, or whether you use an enterprise master patient index (EMPI) product from a vendor like IBM®.
All conventional patient matching technologies use algorithms to compare the demographic data from two patient records to determine if those records match—in other words, if they belong to the same person. If the demographic data is the same or very close, the technology determines that the records match.
The most sophisticated of these algorithms are called “probabilistic” algorithms, and they use statistics, weights, thresholds, rules, and complicated math to calculate the probability that two patient records match. This lets them overcome minor data errors like misspelled names and mistyped birthdates. And it lets them understand that two records with the last name “Rumpelstiltskin” are more likely to belong to the same person than two with the last name “Smith.”
Probabilistic algorithms have actually been around since the 1970s, but they have seen little innovation since then. Yet they are the cornerstone technology of almost every MPI and EMPI on the market.
All conventional patient matching technologies are fundamentally limited
Because probabilistic algorithms directly compare the demographic data from two records, their accuracy is fundamentally limited by the quality and completeness of the underlying patient demographic data they are comparing. Yet patient demographic data is of notoriously low quality and completeness. In fact, typically 30% of patient demographic data in an organization’s systems is out-of-date, incomplete, or errored—making patient matching extremely challenging for even the most sophisticated probabilistic algorithms.
Consider these examples of matches that no algorithmic approach—no matter how advanced—could ever make:
- One record contains a patient’s old address and maiden name, and another contains a patient’s current address and married name
- One record contains very sparse demographic data, such as just a patient’s name and birthdate
- One record contains a patient’s name, address, and SSN, while another contains a patient’s name, phone number, and birthdate
Because of their fundamental limitation, the probabilistic algorithms found in conventional patient matching technologies typically miss 10-20% of matches—leaving EHRs riddled with duplicate records, preventing organizations from assembling complete patient histories, and leading to massive and costly inefficiencies in the revenue cycle.
To compensate, organizations are forced to invest in data quality initiatives, data cleanup exercises, data governance committees, and data stewardship efforts whereby health information management (HIM) staff review and remediate suspected duplicates their EHR or EMPI cannot automatically match.
Referential Matching technology is a fundamentally different approach
Fortunately, there is a powerful new technology called “Referential Matching” that is a fundamentally different approach to patient matching and that any organization can use to dramatically improve its patient matching—even if it already has an EHR or EMPI technology for patient matching.
Rather than directly comparing the demographic data from two patient records to see if they match, Referential Matching technology instead matches that demographic data to a comprehensive and continuously-updated reference database of identities. This proprietary database contains over 300 million identities spanning the entire U.S. population, and each identity contains a complete profile of demographic data spanning a 30-year history. It is essentially a pre-built answer key for patient demographic data.
By matching records to this database instead of to each other, Referential Matching technology can make matches that conventional patient matching technologies could never make—even patient records containing demographic data that is out-of-date, incomplete, incorrect, or different.
Referential Matching technology is so accurate and so powerful that providers, payers, and HIEs across the country are using it to improve the patient matching of their EHR and EMPI technologies, typically by using cloud-based plug-ins to find and remediate undiscovered duplicates. These plug-ins can even automatically remediate the suspected duplicates that the EHR or EMPI has flagged for manual remediation by HIM staff.
Referential Matching isn’t simply a better algorithm—it is a completely new approach that represents a quantum leap in patient matching technology and accuracy. It combines sophisticated probabilistic algorithms with big data and machine learning technologies, and it incorporates 50,000 person-hours of data science and engineering efforts to ensure the integrity and completeness of the reference data. And it does all of this in a highly scalable and secure cloud infrastructure—allowing any organization to instantly and dramatically improve its patient matching through simple integrations and modern APIs.
Simply put, Referential Matching is the new gold standard in patient matching technology.
You can also hear our Sharing @ SHIEC audio with Josh Firstenberg of Verato on the role of Verato in the HIE space as a database master patient indexing technology, unifying patient data across various sources. Josh also shares some unusual findings from recent projects that include aquarium mammals.