Patterns of antibiotic cross-resistance by bacterial sample source: a retrospective cohort study

Background and objectivesAntimicrobial resistance is a major healthcare burden, aggravated when it extends to multiple drugs. While cross-resistance is well-studied experimentally, it is not the case in clinical settings, and especially not while considering confounding variables. In addition, bacteria from different sample sources may have undergone different evolutionary trajectories, therefore examining cross-resistance across sources is desirable. MethodologyWe employed additive Bayesian network (ABN) modelling to examine antibiotic cross-resistance in five major bacterial species, obtained from different sources (urine, wound, blood, and sputum) in a clinical setting, collected in a large hospital in Israel over a 4-year period. ABN modelling allowed for examination of the relationship between resistance to different drugs while controlling for major confounding variables. ResultsPatterns of cross-resistance differed across sample sources. All identified links between resistance to different antibiotics were positive, and most were present in several culture sources. However, in 15 of 18 instances, the magnitudes of the links were significantly different between sources compared. For example, E coli exhibited adjusted odds ratios of gentamicin-ofloxacin cross-resistance ranging from 3.0 (95%CI [2.3,4.0]) in urine samples to 11.0 (95%CI [5.2,26.1]) in blood samples. Conclusions and implicationsOur results highlight the importance of considering sample sources when assessing likelihood of antibiotic cross-resistance and determining antibiotic treatment regimens and policies. Abstract ImportanceWe examine the patterns of antibiotic resistance of a given bacterial species, obtained from different clinical infection locations, while accounting for potentially relevant clinical variables. We find that such patterns of cross-resistance between pairs of antibiotics vary between culture sources (e.g., urine vs blood samples), indicating different selective pressures. These findings have implications on prescription policies aiming to minimize collateral resistance.