The burden of antimicrobial-resistant bacterial infections: a causal perspective

Estimates of preventable antimicrobial resistance (AMR) burden are important to inform local, national, regional, and global policies, targets and research priorities. Such estimates rely heavily on model assumptions and several analytical approaches have been used. In this perspective article, we outline key conceptual and practical challenges in estimating AMR burden, and propose strategies for building on existing work to obtain more policy-relevant burden estimates. We highlight how new approaches taking an explicitly causal perspective are tackling these problems and have the potential to improve the way results are combined from individual studies to estimate national and regional AMR burden. Estimating preventable antimicrobial resistance (AMR) burden is vital for guiding policy and research, but current methods rely on complex assumptions. In this Perspective, authors outline the challenges and pitfalls in estimating AMR burden, and propose their strategies for reducing bias and improving generalisibility of estimates.