Coding longitudinal data

[clarkenj]

For longitudinal data do we want to add a column, e.g. date_of_scan, to enable potential matching with other variables later? Or, age could be used a proxy?

[htwangtw]

What's the label in the original dataset? Age would be more relevant to the analysis so I think that would be fine. We can bring this up in the meeting

[clarkenj]

ADNI and CIMA-Q have both a date and a visit name coded (e.g. V10, bl). OASIS uses age_at_entry and then days from entry (e.g. d0129) for each visit. OK!

[htwangtw]

Ahhh yeah it's going to be hard to find a way to standardise this across dataset. We will need some brain storming

[clarkenj]

Adding notes from meeting here:

• We will use the age variable to code longitudinal data, based on data provided from study.
• Can iterate if does not seem to be effective.