Publication summary

This paper illustrates limitations of current methods to estimate the global mortality from antimicrobial resistance (AMR). Two lead attempts have been made by the O’Neill review and the GBD 2016 study. This paper demonstrates that a direct comparison of these estimates is not possible due to the different methodological approaches, data sources and type of infections considered. To guide implementation of action plans and efficient resource allocation, it is crucial to capture deaths caused by AMR. This paper explores guiding principles and potential approaches for improving AMR estimates in the future.

Who this is for

  • Policy makers
  • Healthcare providers
  • International public and private organizations
  • Non-governmental organizations
  • Researchers

Key findings

  • Current methods are limited: the counterfactual approach is hard to measure and understand while the ICD principle does not count most hospital-acquired AMR infections and an unknown proportion of community-acquired AMR infection as cause of death.
  • There is a need to strengthen health systems to collect, analyse and share better quality data
  • Improved methodological approaches to estimate deaths from AMR are needed. New methods need to be robust, reliable, sustainable and plausible.