Quantitative analysis of a presence/absence microbiological assay: the hard surface carrier test of disinfectant...

The Hard Surface Carrier Test (HSCT) is a presence/absence assay for determining the extent to which chemical disinfectants kill bacteria. The disinfectants are marketed for cleaning surfaces in hospitals, restaurants, the home, etc. In the HSCT, a presence response (or equivalently, a positive carrier) is recorded if one or more bacteria survive among M bacteria attached to a glass carrier and exposed to the chemical disinfectant; otherwise, the response is absence. The presence or absence response is observed for each of many (usually 60) glass carriers and the disinfectant is considered effective if few (e.g., < or = 5%) of the carriers are positive. It would be more satisfactory to microbiologists if effectiveness were based on the fraction of bacteria that survive exposure; denote the expected survival fraction by phi. An equivalent parameter commonly used in antimicrobial research is the log-reduction, denoted by psi, where psi = -log10(phi). Although the number M of bacteria on each exposed carrier is not known, this paper shows it is possible to estimate psi. The suggested estimator is based on the assumption that M follows a gamma distribution. The gamma parameters are estimated using counts of bacteria on the HSCT nonexposed (control) carriers. This paper provides a formula for the standard error of the estimate, a computer simulation technique for calculating the lower confidence limit, and a computer simulation study of these statistical methods.