Co-resistance between oral antibiotics for pyelonephritis and those for cystitis—applying an escalation antibiogram model to local community data

Objectives We applied an escalation antibiogram to community urine data to assess how presumptive resistance to first-line antibiotics for cystitis affects resistance to antibiotics used to treat pyelonephritis. Methods We extracted susceptibility data from Escherichia coli isolates grown from urine samples from general practice during a 5 year period (2019–2023) in a region served by three NHS hospital trusts. Female patients over 18 years old were included, giving a total of 130 514 isolates. We applied a Bayesian model to estimate antibiotic resistance rates for oral pyelonephritis antibiotics, when presuming resistance to each of the first-line antibiotics used for cystitis. The model estimates the probability of resistance with 95% credible intervals and was applied to a variety of patient groups based on age and history of recurrent urinary tract infections. Results Resistance to cystitis antibiotics has a marked effect on the probability of resistance to oral antibiotics used to treat pyelonephritis. In particular, amoxicillin/clavulanate should be avoided for pyelonephritis if resistance to pivmecillinam is presumed, because predicted resistance rates exceed 50%. For patients with presumed resistance to nitrofurantoin or trimethoprim, the optimal pyelonephritis antibiotic depends on both age group and history of past infections. Conclusions Analysis using an escalation antibiogram informed by our Bayesian model is a useful tool to support empirical antibiotic prescribing for pyelonephritis. It provides an estimate of local resistance rates and a comparison of antibiotic options with a measure of the uncertainty in the data.