Limitations of this study include potential bias associated with the retrospective nature of self-report in the OCGAS study (particularly for scarlet fever and RF, which require specific diagnostic criteria) and the absence of a control group. If present, however, the recollection bias appears to have affected probands and relatives for overlapping conditions (encephalitis or meningitis, scarlet fever, and RF), making it less likely that it is a random occurrence. Moreover, previous data suggest that the presence of an autoimmune disease based on self-report cannot be verified by checking medical records in up to 30% of cases (Broadley et al. 2000), and verification was not performed in this study. Similar error rates may be present in this study; although no data are available on recall bias specifically in patients with OCD, this population may overestimate personal risk of disease.
