Data corruption in a multilingual database environment: how to keep the data safe

The data entered in the Greek Cardiac Pacemaker Registry include some items which are entered in the Greek language and others which use letters from the Latin alphabet. Entering a record in the conventional way thus requires constant switching of the computer keyboard between languages and involves a high risk of data corruption. Rather than relying on checks and annoying error warnings, the solution chosen was to create an algorithm which "forces" the data entered in specific fields into the desired alphabet and case. Keyboard switching is thus no longer necessary, saving time, and all errors of this type are eliminated in a manner transparent to the user. A similar approach could be used in any context which involves a language based on a non-Latin alphabet.