Traditional Patient Matching is History: How to Link Records Despite Out-of-Date Data
Federal institutions like ONC and private collaboratives like The Sequoia Project have made it a priority to solve the challenge of consolidating and harmonizing patient identities within health systems. ONC has even set a milestone for healthcare organizations to achieve a 0.5% patient record duplicate rate by 2020.1 But organizations have been addressing this challenge for over a decade, and the average duplicate rate is still 10-20%.2 Clearly, current state-of-the-art technologies and approaches are not adequate to hit this milestone.
Verato approaches patient matching in an innovative new way—one that can get you to a 2% patient record duplicate rate in just 2 months. Each blog in this three-part series briefly examines a key differentiator of the Verato platform over traditional matching technologies. To learn more about our platform, reach out to us.
The Problem: Out-of-Date Data Makes Patient Matching a Challenge
Matching engines—like those found in Master Patient Index (MPI) tools—are the foundational technology responsible for linking and deduplicating patient records. These engines use patients’ identity data as the key to making a match. For example, if a hospital has the following two records, then its matching engine will determine that both of those records belong to the same person and will link them:
But matching engines are only as accurate as the data they are using, and patient identity data is notoriously inaccurate. In fact, 10-20% of patient identity data in any given database is mistyped, misspelled, incomplete, or incorrect. Most importantly, 1% of patient identity data becomes out-of-date each month.3 For example, if Rebecca Jones changes her last name and moves (e.g. after a marriage), her old patient records suddenly contain out-of-date data. Next time she goes to the hospital, a duplicate record will be created because the matching engine can’t match her old information to her new information:
This problem is very common. In fact, in our experience over 10% of records that should match to each other have different addresses—making it a challenge for matching engines to accurately link them.
The Solution: Use an Identity Data “Answer Key” as a Reference During Matching
Verato takes a different approach to patient matching that can utilize historical data to make a match. In other words, Verato can make a match even if one or both patient records have an out-of-date name and address.
Verato leverages its extensive self-learning graph database of US identities as a universal “answer key” for identity data. This database, called CARBON™, uses sophisticated algorithms to combine billions of identity data fragments from credit, telco, and government records into unique identities. Importantly, these data fragments include out-of-date names and addresses—so a CARBON identity will contain a person’s address and name history.
Instead of matching two patient records directly to each other, Verato matches them to CARBON identities. The image below shows how both of Rebecca’s records match to the same CARBON identity even though they have different addresses and names.
Because of the amount of out-of-date, incorrect, and incomplete patient data in organizations’ databases, typical matching technologies only achieve a 70% match accuracy rate. This leads to an excess of duplicates—and therefore to an incomplete view of patients’ medical histories. The Verato platform, on the other hand, achieves a 98% match accuracy rate. And because it is cloud-based, Verato is 6x faster to implement and 2.4x more cost effective than typical matching technologies.
To learn more about Verato, watch our overview video or read why Gartner named us a Cool Vendor for 2016. Then, reach out to us to learn how we can get you to a 2% internal duplicate rate in just 2 months.