The Dirty Work Of Healthcare Information Technology: Developing A Lingua Franca Or The Devil Is In The Details

Main Category: IT / Internet / E-mail
Also Included In: Public Health
Article Date: 17 Dec 2007 - 1:00 PDT

email to a friend   printer friendly   view / write opinions   rate article

I may say soda and you may say pop, but what if the hospital lab you go to today says BMP and the lab you visit next week calls the same test SMA7? Your physician knows that these are the same tests, but the computer systems they use don't automatically know. How can we develop a national healthcare information network, or even regional healthcare information exchanges, if we don't have a lingua franca?

A study by researchers from the Regenstrief Institute, the Indiana University School of Medicine and the Indiana University School of Health and Rehabilitation Sciences, looked at hospital laboratories from five hospital systems in the Indiana Network for Patient Care and found that even in the same metropolitan area - Indianapolis and its suburbs - a variety of names were assigned to the same test - for example: complete blood count, hemogram and CBC.

Looking at 49 million lab results, the researchers found that of the approximately 4,000 lab tests administered throughout the metropolitan area, approximately 20 percent of those tests accounted for 99 percent of the results and covered all the tests performed for 99 percent of patients. Nearly 80 percent of the tests were seldom administered.

"Because so few tests are routinely administered, finding a common language won't be as hard as we had assumed. We'll only have to look at the "David Letterman Top 10," as it were, not an infinite number of tests," said study co-author JT Finnell, M.D., a Regenstrief research scientist and associate professor of emergency medicine at the IU School of Medicine.

More than ten years ago, the Regenstrief Institute developed and now maintains and updates the Logical Observation Identifiers Names and Codes (LOINC) database, a universal standard for identifying laboratory and other clinical observations to deal with this problem.

"Mapping local test names to a common language like LOINC provides a bridge across the many islands of health data that currently exist. Focusing on the most common tests is a manageable task and also comes with a huge payoff," said co-author Daniel Vreeman, PT, DPT, a Regenstrief research scientist and assistant research professor of physical therapy, IU School of Health and Rehabilitation Sciences.

The study, published in the 2007 Proceedings of the American Medical Informatics Association Symposium, has been awarded the 2007 AMIA Distinguished Paper Award, one of only three such awards made this year in a field of 175 papers. J. Marc Overhage, M.D., Ph.D. is also an author of the study. Dr. Overhage is director of medical informatics and research scientist at the Regenstrief Institute, Inc.; Regenstrief Professor of Medical Informatics at IU School of Medicine and president and CEO of the Indiana Health Information Exchange.

Indiana University School of Medicine