Data linkage is typically conducted using specialized software that employs algorithms to match records from different datasets. These algorithms can use various identifiers such as patient names, birthdates, and medical record numbers. Advanced techniques like probabilistic linkage can handle inconsistencies and errors in the data, improving the accuracy of the linkage process.