Research studies often use "participant linking maps" to depersonalize patient responses in a secure manner and accomplish blinding of the investigators. Concretely, this usually takes the form of a spreadsheet with a row for each participant containing the patient's name/ID and a sequence number. The sequence number is then recorded, usually in handwriting, at the top of each paper questionnaire completed by the patient. The dataset contains only the sequence number, enabling unbiased and anonymous analysis of the result set(s).
There are two drawbacks of this process, which is as old as blinded research:
- It takes time and energy to create as well as look up a sequence number from the map every time a participant is going to complete a response or be assessed for a data point.
- Somebody needs access to this "master" list and therefore can "unmask" the participants.
Ocean Studies provides a better, simpler way of mapping participating patients to result data. It is done with "participant key", which is generated automatically from stable, personally identifying information using a one-way hashing algorithm. An Ocean participant key may look ugly, but it provides an irreversible token so that you cannot figure out any identifying information about the patient from the key. The concept is gaining momentum in health research and is often called pseudonymization.
In Ocean, the participant key is constructed automatically from the site number, the patient birthdate, sex, surname and, if available, a few digits from the middle of the health number. The resulting participant key is attached to every Ocean Study submission. As a result, every submission from a given patient will have the same participant key for a given study.
For patient data security, Ocean will generate different participant keys for different studies, even if the same patient is participating in two different studies. If you are doing a multi-form study, you can use participant key "alignment" on the Ocean Study definitions to have multiple studies produce the same participant key for the same patient. For more information about aligning participant keys, please refer to this article.