Can anyone remind me of the best statistical test to use for intra-observer variability when interpreting clinical data? The data is interpreted by two sets of clinicians (from primary and secondary health care) who have no other information about the subjects. Many thanks.