Patient matching has long been one of the most critical challenges in assembling healthcare documentation. Usually under a time crunch, HIM staff have to find key pieces of information like names and birthdates to ensure that the files they’re receiving aren’t matched to the wrong patient or that they’re creating a duplicate patient.
The importance of patient matching can’t be understated as it’s one of the easiest ways to end up either not delivering needed care or administering improper care. It’s no surprise, then, that many in the healthcare space are looking to automation, including Artificial Intelligence and Machine Learning, to improve the speed and accuracy of their patient matching.
A recent study published in JAMIA and examined by EHR Intelligence showed that when using algorithms to drive patient matching, a referential approach will yield greater accuracy than a probabilistic one.
Probabilistic programs look at individually weighted attributes, compare them to the record, and then assign a score that adds up to a threshold that indicates a match. Referential matches, on the other hand, incorporate additional pieces of data like previous addresses or phone numbers, or even more general types of information like common spellings of names.
The referential algorithm was able to identify many patients that the probabilistic one could not, with those cases showing missing data, multiple phone numbers, nicknames, and other data that the referential method is better able to account for.
While it sounds fairly cut and dry that healthcare organizations would lean toward the referential match, it’s important to note that the study was comprised only of Indiana health systems and the quality of the referential data was strong.
Different organizations are going to have different levels of access to this type of information, whether it’s a concern of cost or priorities. Regardless, some of these data points can be generated by the patient and hospitals themselves if the proper questions are asked. It’s also important to be familiar with the demographics of your patient base as people who are homeless or children won’t have the same data availability as other groups.
Patient matching has been a critical part of our automation software not just because we know how challenging it can be, but because of the potentially devastating impact the wrong match can have for patient care. So while at its core, the objective of our software is to deliver data, our intimate knowledge of healthcare workflows lets us use lookups to identify not only the correct patient for matching, but also order and encounter numbers or even verifying that a test result has all of the necessary components to be posted.
If you’re interested in learning more about how you can automate your patient, order, or encounter matching, please reach out and we’d be happy to discuss how we can help.
About the Author: Chris Mack
Chris is a Marketing Manager at Extract with experience in product development, data analysis, and both traditional and digital marketing. Chris received his bachelor’s degree in English from Bucknell University and has an MBA from the University of Notre Dame. A passionate marketer, Chris strives to make complex ideas more accessible to those around him in a compelling way.