Poor Data Quality, Weak Algorithms Lead to Patient Matching Issues

Hospitals and health information exchanges (HIEs) still struggle with patient matching issues, with many citing data quality problems and poor algorithms as top barriers to patient matching, according to a survey from eHealth Initiative (eHI) Foundation and NextGate.

Correctly linking patient data across organizations is a key element of value-based care, patient safety, and care coordination. Duplicate or mismatched records can result in privacy risks, claim denials, redundant medical tests or procedures, and reporting errors.

"As the number of players and organizations in the healthcare space continue to expand rapidly, patient matching is even more important," said Jennifer Covich Bordenick, chief executive officer of eHealth Initiative Foundation.

Researchers surveyed 118 leaders at provider and HIE organizations to assess the current state of patient matching among healthcare entities.

The results showed that 66 percent of respondents believe that data entry errors contribute significantly to duplicate records at their organizations. The survey authors noted that the absence of even a single medication can have detrimental effects on patient health, leading to ill-informed clinical decisions.

"Incomplete or inaccurate data in one's health record can be detrimental to patient safety and a significant barrier to delivering coordinated, patient-centric care," said Andy Aroditis, CEO of NextGate.

"This survey confirms that healthcare institutions must continue to invest in better approaches to facilitate a comprehensive and accurate record of care across the continuum. Measures that move identity management out of EHRs, make meaningful information more accessible and sharable, and improve data governance at the point of collection, are much needed steps to accelerate patient matching performance."

Forty-six percent cited inadequate matching algorithms as a top contributor, while 42 percent of participants said that poor system integration was to blame for patient matching issues. Thirty-five percent pointed to a lack of industry-wide standards.

Technology and insufficient data governance also significantly contribute to patient matching issues. Forty-one percent of participants said a lack of technology is among the biggest barriers to improving patient matching, and 38 percent named a lack of data governance as a major hurdle. Forty-one percent cited too many competing priorities as a top barrier to better patient matching.

Duplicate patient records still pose a large problem for organizations as well. When asked what percentage of all stored records at their organization are duplicates, 32 percent of all respondents said three to ten percent. Among HIEs, 27 percent reported less than three percent, and among providers, 36 percent said three to ten percent.

Duplicate patient records can have negative effects on patient safety and healthcare costs. A 2018 Black Book survey showed that duplicate medical records and repeated medical care cost an average of $1950 per patient per inpatient stay, and over $800 per ED visit.

Participants in the Black Book survey also estimated that 33 percent of all denied claims are a result of inaccurate patient identification or incorrect information. Claims denied for these reasons cost the US healthcare industry over $6 billion annually.

Poor patient identification and record mismatches are exacerbated by large amounts of data and organizations’ use of multiple applications, eHI and NextGate found. Thirty-eight percent of HIEs said they currently have 31 or more EHR and information systems in their IT environment.

Fifty-two percent of providers are running an average of one to five EHRs and information systems, while 24 percent are running six to ten systems and 19 percent are running eleven to 20 systems.

To overcome patient matching issues, most entities have employees or contractors to solve potential duplicates and mismatches, with 80 percent of HIEs and providers reporting that their organization has teams dedicated to these tasks. Fifty-four percent of all respondents said that these employees and contractors address potential mismatches daily.

The survey also asked respondents to rate various innovations they believed would impact patient matching efforts the most, on a one to eight scale with eight being the most important.

Overall, demographic data standardization (5.7) and biometrics (5.5) ranked highest, followed by third-party data (4.9) and machine learning (4.8). Participants also identified other innovations such as blockchain (4.0), smart cards (4.0), and smartphone-based approaches (3.5).

As the healthcare system continues to shift toward value-based care, patient matching and data standardization will play a large role in organizations’ success.

“Correctly linking patient data across EHR systems remains a significant challenge for health systems, hospitals, offices, and any facility where patients receive care,” the survey concluded.

“Patient matching and identification remain a top priority when it comes to lowering costs, enhancing clinical decision-making, improving patient safety and fostering care coordination— all critical components of value-based care.”