Measurement of incidence of the human liver fluke, Opisthorchis viverrini, in northeast Thailand.

A survey of a community in northeast Thailand where Opisthorchis viverrini is endemic permitted comparison of two methods of measuring incidence: direct determination of the rate at which an uninfected group became infected between two surveys a year apart, and estimation of incidence from age-specific prevalence data using a logarithmic regression method. Both methods revealed that incidence increased with age in young children, and estimation from age-prevalence data showed that the increase was roughly linear from near birth to about age 5 years, beyond which no clear trend in incidence was evident. A catalytic infection model incorporating an infection rate increasing from birth to age 5, and remaining constant thereafter, gave an excellent fit to age-prevalence profiles. Both methods of determining incidence are sensitive to errors in diagnosis, but the direct determination method is more sensitive to the presence of false negatives. A method for correcting this bias is given. The regression method, which was less sensitive to yearly variation in incidence and is easier to use, is recommended for preliminary surveys to identify villages with high transmission intensity.