Whereas one might presume that cross-sensitivity found in laboratory
settings would be translated to similar resistance patterns in the clinic,
the reality was different. Although cross-sensitivity was observed in
the lab, no pairs of antibiotics were found to have negatively associated
resistance probability in the clinic, both in a pairwise comparison and in
an ABN analysis. We further investigated the implications of crossresistance
and co-occurrence in the clinic using a mathematical model.
The results suggested that resistance co-occurrence is very influential
on resistance frequencies, and should be considered when designing a
treatment strategy.