In the study we explore the legal requirements for anonymization in Finland and the European union. We also cover the mathematical basis and function of the most popular anonymization methods. We present a path attribute -method for anonymization where one can pinpoint anonymization to identifying nodes of the care pathway instead of censoring the entire event or the entire care pathway as a part of the anonymization process. We experiment with the validity of the method with generated data and we find that the data anonymized using the path attribute -method correlates strongly with the original data.