Patient-level data matching, a fundamental and critical success factor for the longitudinal alignment of anonymized health information, can be a daunting challenge given the complexity, sensitivity, volume, and heterogeneity of patient healthcare data, as well as the use of primitive technologies.
A comprehensive picture of the anonymized patient requires accurate matching of individual patients to their health records across settings, however, patient matching rates vary widely. Matching records to the correct anonymized person becomes increasingly complicated as organizations share records electronically using different systems.
False positives – when two non-matching records are linked – and false negatives – when two matching records are not linked – are the two primary types of errors that occur in patient matching attempts. False positives are more difficult to correct. False negatives are considered the lower risk error but can result in a fragmented view of the individual, leading to inaccurate conclusions and/or strategies.
MSA’s secondary matching and/or data steward processes resolve false negatives by using source data patient identifiers to identify patients with a different token value for the same internal patient identifier. False positives are resolved by identifying patients with the same MSA Patient ID and a different source data patient identifier.
MSA’s patented De-Identification (De-ID) Engine and Data@Factory processes anonymize and connect patient data from any number of disparate datasets, providing the highest patient matching rates in the industry.