Absence of effects of widespread badger culling on tuberculosis in cattle

Abstract

AbstractGovernment policy in England aims for the elimination of bovine tuberculosis (bTB). This policy includes culling of European badger (Meles meles) to reduce cattle TB incidence. The rationale is based on a field trial, the Randomised Badger Culling Trial (RBCT) 1998–2005, which reported a substantial decrease in bTB herd incidence where badger culling had been implemented, in comparison to untreated control areas. The RBCT was undertaken because previous studies of reductions in badgers by culling, reported a possible association between bTB in badger and cattle, but none could directly show causation. The effect of intensive widespread (proactive) culling in the RBCT was reported in 2006 in the journal Nature. Analysis of an extensive badger removal programme in England since 2013 has raised concerns that culling has not reduced bTB herd incidence. The present study re-examined RBCT data using a range of statistical models. Most analytical options showed no evidence to support an effect of badger culling on bTB herd incidence ‘confirmed’ by visible lesions and/or bacterial culture post mortem following a comparative intradermal skin test (SICCT). However, the statistical model chosen by the RBCT study was one of the few models that showed an effect. Various criteria suggest that this was not an optimal model, compared to other analytical options available. The most likely explanation is that the RBCT proactive cull analysis over-fitted the data with a non-standard method to control for exposure giving it a poor predictive value. Fresh appraisal shows that there was insufficient evidence to conclude RBCT proactive badger culling affected bTB breakdown incidence. The RBCT found no evidence of an effect of culling on ‘total’ herd incidence rates. Total herd incidences include those confirmed as bTB at necropsy and those herds where there was at least one animal animal positive to the comparative intradermal skin test, the standard diagnostic test used for routine surveillance, but not confirmed at necropsy. This was also the case using the more suitable statistical models. Use only of ‘confirmed’ herd incidence data, together with a more recent (2013) published perception that RBCT data presented ‘a strong evidence base….with appropriate detailed statistical or other quantitative analysis’ should be reconsidered. The results of the present report are consistent with other analyses that were unable to detect any disease control benefits from badger culling in England (2013–2019). This study demonstrates one form of potential driver to the reproducibility crisis, in this case with disease control management in an increasingly intensified livestock industry.