The Regenstrief Institute announced this week that some of its researchers have developed an eight-point framework designed to evaluate the validity and accuracy of algorithms for matching patient medical records. .
WHY IT IMPORTANT
Accurate patient matching throughout the continuum of care is essential for quality and safety. It is also key to reducing healthcare costs by reducing the ordering of duplicate medical tests. But without a national patient identifier – Regenstrief points out that the US is the only developed country that does not use a patient identifier – that goal has been difficult to achieve for many years.
Due to the lack of national patient IDs, in the United States linking patient records often relies on algorithms designed by technology providers or healthcare researchers.
Regenstrief vice president of data and analytics, Dr Shaun Grannis led a team of research scientists to create an eight-point framework to evaluate the performance of algorithms tailored to that patient. .
“Individuals increasingly receive care from multiple sources. Although patient matching is complex, the exchange of health information is crucial,” the Regenstrief researchers explained when the study was published. framework father.
“Is William Jones seen at a health care system the same person as William, Will or Willy Jones or perhaps Bill or Billy Jones being cared for at other facilities? Is Elizabeth Smith’s name correct? shows up at different medical offices or perhaps at a physical therapy facility or a dialysis facility like Liz or Beth? Juan J. Gomez has the different test results to which facility?
“Spelling errors, missing information and other data errors as well as typical variations add to the complexity,” they added.
Regenstrief’s research, supported by the Agency for Health Research and Quality, created a standard for matching the datasets needed for record association. The new framework covers technical areas including data preprocessing, blocking, records adjudication, link evaluation, and reviewer characterization.
“Our eight-pronged approach covers the waterfront of what needs to be assessed,” Grannis said in a statement. “Setting a framework and specifying tasks and activities to complete will greatly help in standardizing patient matching.”
The Regenstrief researchers say the framework is intended to help “provide necessary transparency” when building and validating patient-matching algorithms – helping to support both “internal and external validity of patient matching algorithms.” documenting link studies and improving the robustness of new profile linking strategies.”
TREND TO BIGGER WOMAN
Efforts towards a national tailored strategy have been underway for many years in both the public and private sectors. There has been some political progress towards national patient IDs, but frustration over the roadblocks continues to remain.
Meanwhile, research on new matching strategies continues on several fronts. Some see biometrics as a potentially transformative approach. Other researchers have developed frameworks that achieve accuracy as high as 99.5%.
Among other recent work at Regenstrief, Grannis and his teams have worked on developing HIE-trained AI models to predict an individual’s risk of COVID-19 hospitalization and build a system. population-based monitoring using EHR data.
ON PROFILE
“The value of data normalization is well recognized,” Grannis said in a statement about the patient-fit framework. “Have a National Healthcare Provider ID. Have a Facility ID and Subject Identifier. Have a Billing Code. Have a Standard Vocabulary for Healthcare Lab Test Results and medical observation – such as the LOINC at Regenstrief.
“Patient identity is the ultimate flaw in our medical infrastructure,” he added. “We are providing a framework for evaluating the accuracy of patient-matching algorithms.”
Grannis says the matching algorithms “come in a variety of flavors, shapes, and sizes.” “To be able to compare the performance of one with the other, or even to understand how they might interact with each other, we must have a standard way of measuring it.
“We’ve created a new, powerful framework for consistent and repeatable evaluation. Simply put, the framework we’ve developed at Regenstrief provides a ‘metric’ for performance. of patient-appropriate tools.”
Email the writer: [email protected]
Healthcare IT News is a publication of HIMSS.
