Minimum inhibitory concentrations (MIC) indicate the concentration of a specific antibiotic required to prevent the growth of a specific bacteria in vitro. To determine whether a bacteria is labelled as antibiotic-resistant or sensitive, MIC values are then compared to clinical breakpoints, such as the ones provided by EUCAST. The use of clinical breakpoints is currently common practice for epidemiological analyses of AMR in humans. However, these breakpoints can depend on the treatment regimen and vary over time. Hence, their use can introduce artificial trends in regional and temporal resistance levels.
Instead, EUCAST encourages the use of ecological cutoff values (ECOFFs) for AMR surveillance. ECOFFs distinguish between wild-type bacteria and bacteria with acquired resistance to a specific antibiotic, do not vary over time once established, and are independent from the origin of isolates (human or animal). In this project, we aim to investigate whether the use of ECOFFs instead of clinical breakpoints can affect AMR trends and resistance monitoring.
Industry-generated AMR datasets provide the opportunity to answer this question by presenting MIC values for multiple bacteria, antibiotics, countries and years. Using our team’s expertise in analysing and combining these datasets, we will focus on eight pathogens of public health interest (Streptococcus pneumoniae, Staphylococcus aureus, Enterococcus faecalis, Enterococcus faecium, Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, and Acinetobacter spp). We will use either ECOFFs or clinical breakpoints to convert MICs to susceptibility labels, and apply statistical tests and regression methods to compare resulting AMR trends.
This will reinforce public health practice by providing a clear example of the usefulness of ECOFFs, alongside an open-source analysis to facilitate their use. We further hope that, in the context of One Health approaches, this work will serve as a basis to improve comparability between different health systems, since ECOFFs are applicable to both humans and animals.