ResearchDiscoveries, inquiries, and interests
Here are my most significant research pursuits.
The LOINC vocabulary standard is the major product of my scholarly work.
LOINC® is a universal code system for identifying laboratory and clinical observations that facilitates exchange and pooling of results for clinical care, research, outcomes management, and many other purposes. LOINC is a rich catalog (encyclopedia) of measurements, including laboratory tests, clinical measures like vital signs and anthropomorphic measures, standardized survey instruments, and more. LOINC’s universal observation identifiers are an essential ingredient for combining health data from many sources. With 78,500+ users from 175 countries, it has become ubiquitous in healthcare. LOINC is an official national standard in about 30 countries, including the U.S. There are hundreds of millions of patients with billions of discrete electronic health data elements coded with LOINC.
Since 2006 when I began leading its development, LOINC has expanded vigorously. Today, LOINC contains more than more than 80,000 terms, and we are adding 4,400 new terms per year. We developed a novel method that has enabled translations into 18 variants of 12 languages. LOINC has been the grist for many other scientific endeavors. Thus far, I have been the PI on 22 funded projects related to LOINC. We have worked with 15 U.S. federal agencies and NIH institutes that are using LOINC to catalog their common data elements. In addition to its global implementation success, LOINC has spawned numerous scientific papers (Google Scholar: ~5,000 ; PubMed: ~250) covering topics of both “basic science” informatics and “applied” informatics.
Novel models for representing clinical content
We continue extending LOINC into new clinical domains by creating novel representations of biomedical data, information, and knowledge.
As I have led LOINC’s ongoing growth over the past decade, we have extended its coverage into many new clinical domains. As healthcare moves to capture more and different kinds of data electronically, new structures and semantic elements are needed to allow information systems to understand and make use of these data. We partner with domain experts, industry, government agencies, and others to create optimal representations for electronic reporting and interpretation. In this “basic science” aspect of informatics, we have developed new models for representing results of genetic tests, newborn screening results, radiology reports, clinical document titles, patient-reported outcome measures and phenotyping. We continue to expand into new areas including behavioral, social, and psychological data. Ultimately, standardized data coded with LOINC becomes the grist for clinical decision support applications, quality measurement, public health case reporting, clinical research, and many other applications.
Practical approaches for using data standards to enable interoperable health information exchange
To accelerate interoperable health information exchange, I develop and evaluate techniques to reduce the effort required to accurately map local terms to standard vocabularies.
There is strong evidence that health information technology can improve healthcare quality and efficiency. The benefits are even more compelling when the health data within their purview are seamlessly shared and processed across different providers, settings, and institutions. Sadly, this is rarely the case. Coalescing electronic health data across systems is especially difficult because of the myriad idiosyncratic local conventions for representing clinical concepts; the resource-intensive process of mapping local terms to a common vocabulary is often the rate-limiting step. Investigators at Regenstrief’s Center for Biomedical Informatics have created the Indiana Network for Patient Care (INPC), the most comprehensive and longest tenured health information exchange in the U.S.A. The scale, speed, and sophistication of data integration in the INPC continues to demand cutting-edge biomedical informatics research and development, and this is an exhilarating aspect of my research. Through my research and mentorship of graduate students, we have authored papers about mapping to LOINC using clinical data from INPC systems, evaluating mapping approaches across institutions, and the novel aspects of our INPC terminology infrastructure and data integration methods. Some of the promising new mapping approaches studied includes automated test mapping based on statistical profiles of associated clinical data, methods with “fuzzy matching” for the test name, and leveraging the “wisdom of the crowd” that can be gleaned from large sets of existing mappings.
Advancing Rehabilitation with Biomedical Informatics
To advance the field of rehabilitation, I evaluate and apply the interdisciplinary techniques of biomedical informatics
In addition to my primary research efforts in data standards, I also work on secondary projects that advance the field of rehabilitation sciences through biomedical informatics. I helped establish the Indiana-Ohio Center for Traumatic Amputation Rehabilitation Research in a collaboration with the Ohio State University, first as a co-investigator and later as PI. The Center was funded by by the Department of Defense for outcomes research on this population. My most significant contribution was establishing the biomedical informatics infrastructure for the Center’s activities. Early in my career I published the first review article on physical therapist’s use of electronic health records. I was also senior author on a paper about the urgent need for competencies in biomedical informatics within physical therapy education. This paper earned the 2010 Stanford Award from the Education Section of the American Physical Therapy Association (APTA), given to the paper with the most influential ideas for physical therapist education. I also authored a paper that outlined how the vocabulary standards like LOINC, SNOMED CT, and ICF could fit together in health IT systems to achieve semantic interoperability of functional status and other data used by rehabilitation professionals.