De-identified Data
For reasons of data privacy, we are committing ourselves to de-identifying the data that we receive as much as possible. We use automated processes to detect any identifiable information such as names, places, numbers, addresses, etc. We then replace those identifiable pieces of information either by a code (e.g. “[LastName]” instead of “Potter”), or by a pseudonym (e.g. “Ron” instead of “Harry”).
However, we are aware that there may be other content that would allow someone to identify someone’s identity, for example a nickname, the mention of an illness, or other personal elements. To circumvent this, we provide everyone who has participated in the chat with the opportunity of giving us terms that they know are in the chat that might make anyone identifiable. When giving consent for the messages to be used, a box will appear where they can add any of the content. We will then go through the data in addition to our automated processes and make sure that they are also replaced.
Here is an example of a chat that is not de-identified (“Before”), and once we have finished replacing the content, it should look something like the second example (“After”):
