Evaluating the effect of badger culling on TB incidence in cattle: a critique of Langton et al. 2022

The European Badger (Meles meles) is a wildlife host for Mycobacterium bovis, the causative agent of bovine tuberculosis (TB) in cattle. In England, large scale badger culling has been implemented since 2013 as part of a licenced Badger Control Policy (BCP). While several studies have shown an association between the BCP and disease benefits in cattle, a study by Langton et al. (2022) failed to identify any evidence of such benefits, raising questions about the effectiveness of badger culling as a tool for disease control. In this study I summarise the primary methodology used in the Langton et al. study and identify two significant issues with the approach used, which compares cattle TB incidence in culled areas of SW England to land without culling. Firstly, their analysis groups data together from cull areas into a combined culled vs unculled treatment, regardless of the time that culling has been applied. Secondly, there is evidence of a clear cull area selection bias, such that land enrolled into culling during the period analysed had a significantly higher baseline TB incidence than areas not culling. I then conducted a Monte Carlo analysis using simulated data under scenarios with differing cull effects, and matching the degree of grouping, selection bias and statistical methods used by Langton et al. (2022). The results indicate that even in scenarios with a relatively rapid and large effect of culling seeded into the simulation there is a significant probability that their analyses will fail to find an effect and may even wrongly conclude that culling is associated with an increase in TB. These results potentially explain the discrepancy between Langton et al. (2022) and other studies evaluating badger culling and highlight the importance of rigorous statistical analysis that accounts for temporal and spatial patterns in disease control data.