Unless a study falls within the exemption at 21 CFR 812.2(c), specific record requirements are listed in 21 CFR 812.140. In general, the records that are needed are those that provide the data for testing the study hypotheses. Records should contain sufficient detail to allow the study to be reproduced when the same protocol is followed. We recommend that investigators maintain detailed records because a review of the study may indicate the need for other analyses of the collected data.
We also recommend that investigators:
a. Maintain records of all data elements captured in the study, including raw measurements and subject co-variables in the form of demographic and morbidity factors;
b. Link every observation recorded to the subject and that person’s covariable data;
c. Preserve information obtained for all subjects enrolled and for all specimens collected.
Additionally, electronic spreadsheets of study data are useful. Given the possible
need to review or analyze study data at the most detailed level, electronic
spreadsheets may help to minimize review time.
