Global surveillance quality gaps in antimicrobial resistance monitoring: a composite-index and inequality analysis of 85 countries

Abstract

Background: Globally, antimicrobial resistance surveillance generates extensive data, yet the quality of the infrastructure producing it is rarely assessed systematically, even though this is essential for interpreting resistance trends and targeting improvements. Methods: We evaluated surveillance across 85 countries using WHO Global Antimicrobial Resistance and Use Surveillance System bloodstream infection data from 2018 to 2023. A composite Surveillance Quality Index combined six dimensions weighted by principal component analysis. We quantified testing-coverage inequality with the Gini coefficient, used a detection-probability model to assess relative ascertainment of resistance, applied a random forest to identify which dimensions drive the index and used multi-criteria optimization to rank countries for intervention. Results: = 0.23); the weak positive direction is consistent with, but does not establish, detection bias, whereby stronger systems ascertain more resistant cases. The detection model identified, in ordinal terms only, the weakest systems as least able to ascertain resistance. Because the index is a composite of its dimensions, we used the random forest descriptively: temporal consistency accounted for 66.4% of the index and sample adequacy for 22.6%. Multi-criteria optimization identified 30 countries in greatest need of strengthening. Conclusions: Global surveillance capacity is highly unequal, with the largest quality gaps in high-burden settings; targeted strengthening would address most of the disparity.