Methods to Estimate the Between-Population Level Effective Reproductive Number for Infectious Disease Epidemics: Foot-And-Mouth Disease (FMD) in Vietnam

Abstract

Foot-and-mouth disease (FMD), which is endemic in 77% of countries globally, is a major threat to the global livestock industry. Knowledge of the reproductive number at the population level (i.e., farm level, herd level, or above) for FMD is important to estimate the magnitude of epidemics and design and implement effective control methods. Different methods, based on disparate assumptions and limitations, have been used interchangeably to compute and report reproductive numbers at the population level without a formal comparison between them. This study compares the results obtained when using alternative methods to compute between populations (R_bp) for FMD using one single dataset collected over 10 years (2007–2017) at the commune-level swine farms in Vietnam. Seven spatial–temporal clusters were identified in the country, and the value of R_bp was computed on each of them using different analytical approaches, namely, epidemic doubling time, nearest neighbor, time-dependent reproductive number (TDR), sequential Bayesian (SB), and birth–death skyline (BDSKY) analysis in Bayesian evolutionary analysis by sampling trees 2 (BEAST2). Estimated R_bp values were relatively similar across methods ranging from 1.25 to 1.61. For the first time, the results here provide a comparison of different methods used to compute R_bp for FMD. Despite differences in assumptions and limitations, results suggest that different methods produce relatively similar outputs. Additionally, the results here provide foundational knowledge to support the evaluation and control of FMD epidemics in a population.