Can you change OpenClinica CRFs (add variables, alter categorical variabes) mid-trial? What happens to existing data?
You can add new versions of CRFs, but the old data will still be attached the old version of the CRF. You can't load the old data from the old CRF into the new CRF. You can export all versions of CRFs at the same time- the number of columns is approximately the number of questions in each CRF times the number of CRFs. If data is not collected for a particular version, that column is left blank.
What are OpenClinica's audit capabilities?
Auditing seems to be unimplemented, yet planned (and at a moderately high priority FWIW)
Can you use OpenClinica for non-patient data?
You probably aren't supposed to - a new subject needs an identifier, a gender, and a YOB date.
However, you could easily enter any gender or birthdate for any subject, regardless if it is human.
JeremyStephens - research the OpenClinica Java code that reads the CRF excel spreadsheat spec. Does it translate to the spreadsheet to an intermediate format? Is it easy to add new CRF import functionality?
There seems to be no intermediate format. OpenClinica makes use of a Apache Java class called HSSF (Horrible Spread Sheet Format) to convert Excel documents to useful Java objects.
Using the existing algorithm for validation and importing files, it wouldn't be too hard to make a new class to accept different kinds of input. The easiest way would be to copy the existing class and modify it.
CSVTable and supporting classes are installed in the OpenClinica framework on the test server; OpenClinica will accept CSV files for CRF input now (to upload, take a XLS CRF form and consolidate all sheets onto one; seperate the sections by one or more empty rows). Example CSV attached below.