Henry J. Lowe, MD

A Simple Heuristic for Blindfolded Record Linkage

Weber SC, Lowe HJ, Seto T, Olson G, Ferris T, Das AK, Kurian A, Olson C, Pragati K. J AM Med Inform Assoc. 2012; 19 (e1): e157-e161

A simple but effective algorithm for matching adult patients seen at more than one site in a multi-site de-identified registry is described. In a data set of 19,000 records a derived match variable consisting of a 2-character prefix from both first and last name combined with date of birth has a 97% sensitivity; by contrast, an anonymized identifier based on the patients full names and date of birth has sensitivity of only 87%.