A capture-recapture analysis in a challenging environment: assessing the epidemiological situation of foot-and-mouth disease in Cambodia.

We performed a two-source capture-recapture analysis for estimating the true number of villages that experienced clinical cases of foot-and-mouth disease (FMD) in 2009 in Svay Rieng province, Cambodia, and assessing the completeness of the official case-reporting procedure. As a first source, we used the 2009 official dataset made up of the 15 FMD-infected villages that were reported to the provincial authorities, and enlarged this list by assuming that all the villages located at less than 4 km from one of these villages also experienced clinical cases in 2009. In addition, we created a retrospective detection protocol using participatory tools cross-checked against a serological survey that detected 13 infected villages. The capture-recapture analysis of these two detection sources led us to the conclusion that 315 (CI(95%) 117-514) villages experienced clinical cases of foot-and-mouth disease in Svay Rieng province in 2009, corresponding to a village-level annual prevalence rate of 0.46 (CI(95%) 0.17-0.74). The official reporting rate to provincial authorities could therefore be evaluated at 0.05 (CI(95%) 0.03-0.13). An analysis of the sensitivity of the estimation of the number of cases to the radius used for enlargement of Source 1 was performed, indicating its low influence. This study clearly highlights the highly enzootic situation of Cambodia regarding foot-and-mouth disease and the substantial underreporting of clinically affected villages to veterinary authorities. We propose explanations for this low notification rate, stress the importance of accurate reporting procedures and, finally, discuss the potential of capture-recapture techniques as a tool for the quantitative evaluation of animal disease surveillance systems.