Identificação única de pacientes em fontes de dados distribuídas e heterogêneas

Abstract

Through its lifetime, patient's healthcare information is generated in several health care institutions and submitted to a series of procedures. The amount of stored information is incremental by nature, both in volume and diversity. There are also differences in the patient identification, bringing high expenses with duplicated procedures and contributing to inaccuracy in diagnosis and treatments. Therefore, the present work uses Record Linkage techniques during the creation of a MPI (Master Patient Index), combined with the specification of a PIX (Patient Identifier Cross-Referencing) integration profile, in order to establish a unique patient identification in different healthcare information systems, which include homogeneous and distributed data sources. Using those concepts and technologies, a project has been specified and it gave birth to an IHE (Integrating the Healthcare Enterprise) /PIX prototype. Experiments have been carried out in three sceneries with real data.