Analytical distributions and statistics for managing import pathway risk using the continuous sampling plan CSP-1

B. Barnes, G. Hood and F. Giannini

Overview

Australia imposes regulations on goods that arrive in the country and have the potential to introduce exotic pests and diseases.

Continuous sampling plans (CSP) are operational biosecurity systems commonly implemented at the border to manage risk and, simultaneously, keep regulatory inspection costs low. Currently, systems are underpinned by the results and proposed design-criteria in the classical work of Dodge (1943). In this paper, we extend that foundational work to include uncertainty and inspection sensitivity, both of which are highly relevant to Australian border operations where samples from arriving consignments, with low but variable levels of contamination, are selected and inspected.

We provide analytical distributions and statistics for all processes of the CSP-1 cycle, including expressions for variance, for the probability that leakage occurs, and for the volume of leakage conditional on occurrence. The effect of inspection sensitivity is included and explicit.

Results show that CSP design principles that assume a perfect detection process do not necessarily generalise to reduce risk   in biosecurity systems where detection may be imperfect,

and we outline how a combination of limiting constraints can be combined to design more reliable CSP systems when detection is imperfect.