Dose-response model for Burkholderia pseudomallei (melioidosis)

Aims: The objective of this study was development of a dose-response model for exposure to Burkholderia pseudomallei in different animal hosts and analysis of the results. The data sets with which the model was developed were taken from the open literature. Methods and Results: All data sets were initially tested for a trend between dose and outcome using the Cochran-Armitage test. Only data showing a statistically significant trend were subjected to further analysis (fitting with parametric dose-response relationships). Dose-response relationships (exponential, beta-Poisson and log-probit) were fit to data using the method of maximum likelihood estimation. Conclusions: Dose-response analysis of BALB/c mice, C57BL/6 mice, guinea pigs and diabetic rats showed that BALB/c mice exposed intranasally (i.n.) and guinea pigs exposed intraperitoneally (i.p.) are significantly more sensitive to B. pseudomallei than C57BL/6 mice exposed i.n. and diabetic rats exposed i.p. Significance and Impact of the Study: The results confirmed the findings of a study of outbreak data that the diabetic population is more susceptible to infection with B. pseudomallei than the general population. The low dose prediction from best fit dose-response models can be used to draw guidelines for public health decision making processes, including consideration of sensitive subpopulations.