Record Linkage of Health Care Insurance

Design: The procedure for record linkage has three major components—data standardization, weight estimation, and matching. The proposed method was designed to incorporate a combination of exact and probabilistic matching techniques. Measurements: The procedure was validated using convergent, divergent, and criterion validity measures. Results: The output of the process achieved a sensitivity of 92 percent and a specificity that approached 100 percent. Conclusions: The procedure is a first step in addressing the current trend toward larger and more complex databases. ■ J Am Med Inform Assoc. 2001;8:281–288. TIMOTHY W. VICTOR, PHD, ROBERTINO M. MERA, MD, PHD quality that are prone to error. Another problem is the enormous size of the data sets to be linked. Linking must occur both within and between data sets. The unreliability of the identifying information contained in successive records of the same subject presents a significant challenge. Specifically, limitations of either matching technique include the difficulty in handling large and heterogeneous data sets, the need to provide a priori weights for probabilistic matching, and the need to generate unique identifiers. Furthermore, exact matching has low sensitivity, and probabilistic matching is computationally expensive in large data sets. This paper describes a method developed for linking records of individual patients and health care providers across time and geography. The method was designed to incorporate a combination of exact and probabilistic matching techniques. Finally, this method provides a static unique identifier for every patient and health care provider.