Determining SARS sub-clinical infection: a longitudinal seroepidemiological study in recovered SARS patients and controls after an outbreak in a general hospital.

A cohort of 67 confirmed SARS patients were prospectively followed for 16 months and were compared with a control population. Serum samples taken at various times were tested for IgG and IgM; dynamic serological changes in these antibodies were described. The positive responses of IgM and IgG antibodies in sera against SARS virus from the first week to the sixth week after onset of the illness in patients with SARS were measured. The ELISA test of IgG antibody was negative in 200 community controls. The positive rate in the SARS high-risk population was 0.61% tested by ELISA and 0.21% by IFA. The high-risk population in this study was defined as those who provided health care and other services to SARS patients during the outbreak. IgG antibody in convalescent serum of patients with SARS revealed an increasing trend, peaking at the 22nd week after onset of illness followed by a slow decline. IgM appeared earlier than IgG and can be better used for early detection. IgG remained at a high level for a much longer period, serving as a good indicator for follow-up and for assessing past exposure. Our results also suggest that sub-clinical infection, if it exists, is very rare.