Skip to main content

Northwell on Mission to Annihilate Duplicate Patient Records

Analysis  |  By Mandy Roth  
   June 07, 2018


Referential matching holds promise to significantly improve claims denials, reduce medical errors and waste, and improve the patient experience.

In a development that holds promise to significantly reduce claims denials, diminish unnecessary repeat tests, and improve the patient experience, Northwell Health has tested a novel approach to reducing patient record duplication, producing results that vastly improved its performance.

The pilot project assessed the effectiveness of referential matching, an innovation developed by Verato. New York State–based Northwell resolved 87% of a sample set of mismatched records after other matching processes were deployed. The 23-hospital system must address 300 mismatched records each day.

Northwell has 5 million records in its system and treats more than 2 million patients annually; typically, hospitals experience a 10% duplication rate, according to the American Health Information Management Association (AHIMA).

Following full implementation of the solution in June, this reduction will considerably lower the case load of records that staff specialists must manually address in a labor-intensive process.

"Duplicate medical records are the nature of the business," says Northwell's IT director Keely Aarnes, "but it has implications that run throughout the healthcare system."

Besides the cost of claims denials, which are time-consuming to rectify, consequences include potential medical errors because a patient's complete record is not available, unnecessary repeat testing when records can't be located, and reputational damage due to poor patient and clinician experiences. 

Complex Challenge; Simple Technology

For nearly two decades, health systems have primarily relied upon probabilistic matching to resolve patient record duplication. This approach compares patient records and assesses the probability of whether they belong to the same person.

This method is now built into many EMR solutions, and health systems also can implement third-party technologies, known as Enterprise Master Patient Index, or EMPI, products that also employ the same logic.

"Probabilistic matching has done as good a job as it could possibly do," says Mark LaRow, CEO of Verato, "but there's a basic mathematical limitation to this algorithmic approach."

The technology behind referential matching is surprisingly simple. It involves comparing a patient record to Verato's propriety reference database of roughly 350 million people.

The company combines publicly and commercially available information containing all known names, addresses, phone numbers, and other data for individuals spanning back three decades, providing a comprehensive view into the identity of most people living in the United States.

"We see the full range of both the correct and incorrect versions of a person as they represent themselves [over time]," says LaRow. Depending on the quality of data and the population's characteristics, these automated matches average 92% without human intervention, and can be as high as high as 98%.

Why Health Leaders Should Care

Duplicate medical records are a growing problem due to several factors:

  • The situation has been exacerbated as hospital systems consolidate and records from different organizations are combined.
  • More departments now access the EMPI. As the number of access points increases, more decisions are made based on the data available, and additional patient identities are contributed to the system. These variables increase the probability that something will go wrong.
  • As patients begin directly interacting with their records, more opportunities for problems will occur, including privacy concerns.

This issue creates financial havoc for health systems. In addition to labor costs and resources required to address this challenge, according to a study conducted by Black Book Market Research and reported in an April 10, 2018, news release:

  • Thirty-three percent of all denied claims result from inaccurate patient identification or information. In 2017, this cost the average hospital $1.5 million, totaling $6 billion annually for the U.S. healthcare system.
  • The cost of repeated medical care related to this duplicate records averages $1,950 per patient for inpatient stays and more than $800 for emergency department visits.


Northwell's Eye on the Future

Two years ago, Northwell began an aggressive program to reduce duplicate records. At that time, it had about 700 duplicate records daily and a backlog of more than 200,000 potential duplicates.

An outside company was contracted to manually address the bottleneck, and Northwell began using InterSystems' HealthShare Patient Index EMPI product to reduce duplicate records through probabilistic matching to attain its current rate of 300 tasks each day.

"It's a significant improvement, but not quite where we want to be," says Aarnes. Northwell then added Verato to the mix because it was offering referential matching in the healthcare sector, and it features a cloud-based solution, which Aarnes says will further reduce complications and administrative time. 

The potential for the future is what excites Aarnes most.

"We are getting really good at resolving potential duplicates," says Northwell's IT director, "but we want to prevent them from even happening."

As staff members are freed from duties related to resolving duplicate records, their responsibilities have shifted to training more than 2,000 registration personnel about how to properly create records and prevent problems. But Northwell has an even bigger vision in mind.

Another pilot program is underway in six of the system's practices, testing a biometric patient registration process that involves scanning the iris. Northwell and Verato are joining forces with RightPatient in this venture.

If successful, "that will be our matching mechanism to be able to link to the right patient," says Aarnes. It also could reduce the need for patients to constantly fill out forms.

"It's all part of our digital transformation," says Aarnes. "Patient matching and patient identification are critical to that success. These two initiatives go very much hand-in-hand."

Mandy Roth is the innovations editor at HealthLeaders.

Get the latest on healthcare leadership in your inbox.