Speak to an Expert
If you are looking to ground an important decision in robust analysis and evidence, let’s start the conversation.
A systematic review exploring the relationship between antimicrobial use (AMU) in cattle and antimicrobial resistance (AMR) in Salmonella and E. coli. Using strict criteria, we screened 13,738 articles and included 36 for analysis.
Despite a well-defined scope, we found large heterogeneity between the studies (i.e. differences not explained by randomness alone). And following international guidance for assessments of different dimensions in the study, we found poor certainty of evidence. Both findings suggest that conclusions about AMU and AMR in cattle should be interpreted very cautiously, and pooling of studies to come up with AMU-AMR estimates is not recommended.
But since most authors pool studies anyway, we wanted to explore how AMU-AMR association estimates would be affected by explicitly modeling the heterogeneity between studies using a statistical output called a “prediction interval”, and the results were shocking! While pooling showed a statistical association between use of tetracycline and third-generation cephalosporin antibiotics and AMR in cattle, using prediction intervals yielded no consistent association for all tested resistance outcomes.
In practice, this means that we should be extremely cautious when interpreting meta-analyses that pool AMU-AMR evidence in livestock, and at a minimum, those estimates should incorporate prediction intervals.
If you are looking to ground an important decision in robust analysis and evidence, let’s start the conversation.
