Winning the Match Game: Strategies for Patient Identity Matching

This monthly blog highlights and discuss emerging trends and challenges related to healthcare data and its ever changing life cycle.

By Lorraine Fernandes, RHIA


It was a packed house October 9 when Michele O’Connor and I discussed the latest strategies for addressing patient matching at a professional practice roundtable at the AHIMA Annual Convention and Exhibit. Perhaps the most surprising element of the hour was the diversity of attendees, with payers, public and private healthcare providers, and vendors from all technology and consulting sectors of HIM practice. Three key themes emerged during the session.

Buzz on the National Scene

The activity at the national level is probably as high as it’s ever been, with the College of Health Information Management Executives (CHIME) sponsoring the National Patient ID Challenge and several research projects undertaken by the Regenstrief Institute. The Regenstrief Institute work is being funded by several different parties, including the Agency for healthcare Research and Quality and the Pew Charitable Trusts, further illustrating the breadth of interest in patient matching.

Meanwhile, the Office of National Coordinator (ONC) sponsored the Patient Matching Algorithm Challenge, with the winners just announced. The preliminary results indicated that 20 vendors scored within two percentage points of each other, 94-96 percent accuracy. Thus, it will be interesting to see how the final results are interpreted by researchers, vendors, and those outside healthcare, particularly since some might infer that patient matching must be easy if so many vendors scored so well. Alas, as practitioners we know that working with real data (and the inherent human errors) is far more difficult than working with the synthetic data represented in this challenge.

Innovation in Resolving Duplicates

Kathy Fitzgerald, HIM Director at St Jude’s in Orange County, CA (part of the Providence St. Joseph Health system), explained how her organization is applying machine learning to auto-adjudicate duplicate records. Fitzgerald shared that her cost of resolving a duplicate was reduced by 80 percent, and accurate patient data is now available on a timely basis to meet the analytic, research, and care coordination needs of the organization. Kathy discussed how her organization has incorporated this approach into data governance activities, and the importance of a solid, multi-stakeholder pilot program. I reminded attendees of the August Journal of AHIMA article “Applying Innovation to the Patient Identification Challenge” that discussed the various approaches to auto-stewarding, machine learning, and data enrichment.

Lack of Data Governance

Sadly, only two of the attendees shared that their organizations or firms are elevating the patient matching challenges and forward progress to a data or information governance discussion. O’Connor and I advised the group to “think more broadly” and advance the discussion to a governance issue, as accurate and consistent patient/person and provider matching is critical to creating consistent, trusted data that can support strategic and operational goals.

Data governance and improving patient matching should go hand in hand and help guide the discussion. Providers, payers, and vendors must ask several basic questions, such as:

  • Who can approve new sources for identity data?
  • Who can add or modify data?
  • Which quality measures should be applied to new data?
  • Which basic quality standards must apply to ensure data consistency and accuracy?

Moving the discussion to a governance issue should broaden the stakeholder community and create interest from C-level execs.

Continuing the patient matching dialogue, applying innovation, and elevating the discussion to a governance perspective are all vital aspects of healthcare transformation. With so many data uses (including analytics, care coordination, population health, consumer engagement, and care delivery), we must elevate the patient matching discussion at every level.


Lorraine Fernandes is principal at Fernandes Healthcare Insights and president-elect at the International Federation of Health Information Management Associations (IFHIMA).

Submit a Comment

Your email address will not be published. Required fields are marked *

Share This

Share This

Share this post with your friends!