Abstract: Representing clinical data accurately and consistently across diverse sites requires mapping potentially idiosyncratic, ambiguous institution-specific data identifiers to a standardized terminology. This study assessed LOINC code assignment accuracy for 52 common laboratory tests by examining the key LOINC attributes of selected codes. Optimizing LOINC code accuracy at our institution will require reducing redundancy in the mapping schema, examining the role of the technician and the laboratory device in generating mapping features, and improving concept coverage.
Learning Objective 1: Systematically evaluate potential sources of error in laboratory-generated LOINC code mapping.
Sharidan Parr (Presenter)
DEPARTMENT OF VETERANS AFFAIRS, TENNESSEE VALLEY HEALTHCARE SYSTEM
Shuanghui Luo, Vanderbilt University Medical Center
Sina Madani, Vanderbilt University School of Medicine
S Rosenbloom, Vanderbilt University School of Medicine