A modelling approach to the quantification of the benefits of a national surveillance programme.

Abstract A simulation modelling approach was developed to investigate the effects of a reduction in funding for animal health surveillance on the probability of diagnosis of a currently-exotic species of Salmonella if it were introduced into New Zealand. Both severe and mild clinical-disease outbreaks were hypothesized. The probabilities of diagnosis for a single outbreak on a cattle farm under the current surveillance programme were 0.742 and 0.328 for severe and mild clinical-disease episodes respectively, reducing to 0.018 and 0.005 respectively under the reduced surveillance programme. The impacts of these probabilities on the likely time delays from introduction into New Zealand before a definitive diagnosis was reached and the resultant numbers of infected properties were estimated for three different epidemic scenarios. For the best-case epidemic, the median time to diagnosis under the current surveillance programme was 4 weeks (by which time there was still only the index property infected), but under the reduced surveillance programme, the median time to diagnosis was over 1.5 years. Under the most-likely epidemic scenario, the mean time to diagnosis under the current programme was 4 weeks (by which stage there were two infected properties), whereas the median time to diagnosis under the reduced programme was 40 weeks (by which time there were 88 infected herds). With the worst-case epidemic, the median time to diagnosis under the current surveillance programme was still 4 weeks with only two infected properties; however the median time to diagnosis under the reduced programme was 33 weeks by which stage there were 90 infected farms.