Antimicrobial resistance trends, predictors, and burden in England: a retrospective study using the Clinical Practice...

Evidence describing the burden of antimicrobial resistance (AMR) to health care systems is essential to inform health care policy to tackle AMR. This study aims to investigate trends, predictors, and clinical and economic burden of AMR within the English National Health Service.

We conducted a retrospective cohort study assessing 1,036,850 patients diagnosed with an infection as a hospital inpatient linked to the Clinical Practice Research Datalink within England between 2015 and 2021. Resistance was identified with specific International Classification of Diseases, 10th revision, codes. Predictors of AMR were determined using logistic regression and extreme gradient boosting trees. A 1:1 propensity score matched cohort of 78,153 patients with an AMR infection compared with those without was created. Incidence rate ratios were estimated between those 2 groups for hospital length of stay (from admission to discharge), rehospitalisations within 1 year of diagnosis of infection, and total hospital costs. Acceleration factors were estimated for time to all-cause mortality and first rehospitalisation.

Between 2015 and 2020, AMR had a proportional increase of 10%. Resistant infections are associated with an additional 9.2 days length of stay and an additional hospitalisation cost of £3441 per patient compared with infections not reported as resistant. Within 1, 3, and 5 years 3.2%, 6.3%, and 7.3% more patients with AMR infections died and 7.8%, 5.3%, and 3.6% more patients were rehospitalised, respectively. Acceleration factors indicated patients with AMR infections die 27% earlier and rehospitalised 26% earlier than those with non-AMR infections.

AMR is a significant and growing threat to health care; this analysis suggests that AMR is associated with significant excess of health care use and increased mortality and readmission.