Mortality risks associated with empirical antibiotic activity in E. coli bacteraemia: an analysis of electronic health records

Background Reported bacteraemia outcomes following inactive empirical antibiotics (as judged by in vitro testing) are conflicting, potentially reflecting heterogeneous effects of species, minimum inhibitory concentration (MIC) breakpoints defining resistance/susceptibility, and times to rescue therapy. Methods We investigated adult inpatients with Escherichia coli bacteraemia at Oxford University Hospitals, UK, from 04-February-2014 to 30-June-2021 receiving empirical amoxicillin-clavulanate with/without other antibiotics. We analysed 30-day all-cause mortality from index blood culture using Cox models by in vitro amoxicillin-clavulanate susceptibility (activity) using the EUCAST resistance breakpoint (>8/2mg/L), categorical MIC, and a higher resistance breakpoint (>32/2mg/L), adjusting for other antibiotic activity and multiple confounders including comorbidities, vital signs, and blood tests. Results 1720 E. coli bacteraemias (1626 patients) were treated with empirical amoxicillin-clavulanate. 30-day mortality was 193/1400 (14%) [any active baseline therapy] and 52/320 (16%) [inactive baseline therapy] (p=0.17). With EUCAST breakpoints, there was no evidence that mortality differed for inactive vs. active amoxicillin-clavulanate (adjusted HR, aHR=1.27 [95%CI 0.83-1.93;p=0.28]), nor of an association with other antibiotic activity (p>0.18). Considering categorical amoxicillin-clavulanate MIC, MICs>32/2 were associated with mortality (aHR=1.85 vs. MIC=2/2 [0.99-3.73;p=0.054]). Using the higher resistance breakpoint, MICs>32/2 were independently associated with higher mortality (aHR=1.82 [1.07-3.10;p=0.027]), as were MICs>32/2 with active baseline aminoglycoside (aHR=2.34 [1.40-3.89;p=0.001), but not MICs>32/2 with active baseline non-aminoglycoside antibiotic(s) (aHR=0.87 [0.40-1.89;p=0.72). Conclusions EUCAST-defined amoxicillin-clavulanate resistance was not associated with increased mortality, but a higher resistance breakpoint was. Additional active baseline non-aminoglycoside antibiotics prevented amoxicillin-clavulanate resistance-associated mortality, but active baseline aminoglycosides did not. Granular phenotyping and comparison with clinical outcomes may improve AMR breakpoints.


Introduction
Antimicrobial resistance (AMR) has received substantial attention for its current and projected threats to safe healthcare worldwide [1,2]. In high-income countries, Gram-negative bacteria, predominantly E. coli, are the 30-day all-cause mortality was determined using hospital records that are updated with national data on all within [-12,+24] hours of blood collection for each index E. coli positive culture. Data on antibiotics given in 123 the community prior to admission were not available. We excluded episodes where recorded inpatient 124 antibiotics were commenced >24 hours after the index culture.

131
We included two factors to account for other baseline antibiotics: additional active aminoglycosides and 132 additional active other antibiotics (i.e., neither amoxicillin-clavulanate nor aminoglycoside). We considered aminoglycosides separately as these were typically given as a single additional dose, whereas other antibiotics 134 were generally prescribed for longer. To allow for differing effects of additional antibiotics in patients 135 receiving active or inactive amoxicillin-clavulanate, we used four mutually exclusive categories: active 136 amoxicillin-clavulanate (regardless of other drugs), inactive amoxicillin-clavulanate alone, inactive 137 amoxicillin-clavulanate with active aminoglycoside only, inactive amoxicillin-clavulanate with other active 138 antibiotic. There were insufficient data to include these partial interactions in the categorical MIC model so 139 only main effects for additional antibiotics were included.

141
We also adjusted for additional baseline factors, including patient characteristics, vital signs, and blood tests,  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 23, 2022.   is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 183 antibiotic(s): 138/320 (43%) 24-48h from index blood culture and 230/320 (72%) by 72h (Figure 2)  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

204
We found no evidence of that aminoglycosides per se increased mortality when given with active amoxicillin-

223
In patients with E. coli bacteraemia, after accounting demographics, comorbidities, past hospital exposure, 224 and illness acuity, high-level resistance to baseline empirical amoxicillin-clavulanate (MIC >32/2 mg/L) was 225 associated with increased 30-day mortality. However, we found no evidence that amoxicillin-clavulanate 226 resistance defined using EUCAST breakpoints was associated with mortality (including MICs 16/2-32/2 227 . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 23, 2022. ; heterogeneity in previous studies investigating the impact of AMR in E. coli bacteraemia.
there was borderline evidence of increased mortality in patients with Enterobacterales bacteraemia who 232 received inactive empiric antibiotic therapy (odds ratio=1.23, [95%CI 1.00-1.52;p=0.054]). However, this 233 study did not examine the impact of MIC. Another study of Enterobacterales bacteraemia highlighted the 234 methodological differences between CLSI and EUCAST for determining amoxicillin-clavulanate MICs, and 235 found no association between MIC and mortality, by either method and using a variety of breakpoints, but had 236 only limited power with 202 E. coli cases. [11] 237 238 In addition to highlighting adverse outcomes from AMR, our findings suggest that breakpoints for 239 amoxicillin-clavulanate may be set too conservatively, at least for some bloodstream infections. Interestingly, 240 the EUCAST amoxicillin-clavulanate breakpoint for Enterobacterales causing uncomplicated urinary tract 241 infection is >32/2 mg/L, i.e., the level of resistance we found was associated with increased mortality. As 242 urinary tract infection was the most common presumed bacteraemia source in our study, it is possible that 243 urinary excretion of amoxicillin-clavulanate, coupled with high blood concentrations from intravenous 244 administration, were sufficient to treat infections with MICs 16/2-32/2 mg/L. Our approach highlights more 245 generally the benefit of large-scale electronic health records as a tool for reviewing and setting antibiotic 246 breakpoints, with a greater focus on patient outcomes than has previously been possible.

248
Additional active baseline empirical non-aminoglycoside antibiotics, predominantly cephalosporins and 249 carbapenems, negated the impact of high-level amoxicillin-clavulanate resistance. These additional antibiotics 250 were started prior to microbiology results becoming available, typically replacing amoxicillin-clavulanate, for 251 example, following senior medical review, or transfer from the emergency department to the admitting 252 speciality. There was evidence that these antibiotics as a group were beneficial, i.e., we did not assess 253 variation between cephalosporins vs. carbapenems. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 23, 2022. ; https://doi.org/10.1101/2022.01.22.22269642 doi: medRxiv preprint In contrast, higher mortality with high-level amoxicillin-clavulanate resistance was similar whether additional active aminoglycoside (majority single-dose gentamicin) was given or not. As local guidelines recommended guidelines recommended gentamicin doses of 3-5mg/kg to minimise toxicity, in contrast to higher doses, 261 ≥7mg/kg, that may be required to achieve adequate peak-concentrations, particularly among critically ill

270
A study strength is that we adjust for confounding more completely than many other studies, e.g., we account 271 for prior healthcare exposure and other factors that increase the risk of AMR and may also be associated with 272 adverse outcomes. We also used both vital signs and laboratory tests to ensure differences in illness severity at 273 initial presentation were robustly accounted for. Consequently, we found multiple other independent 274 associations with increased mortality: greater age, lower BMI (potentially reflecting lower physiological   is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 23, 2022. ; https://doi.org/10.1101/2022.01.22.22269642 doi: medRxiv preprint guidelines recommended piperacillin-tazobactam for this); therefore, most neutropenia observed, and associated with increased mortality, is likely to be a consequence of acute infection. Low monocytes were infections, e.g., eosinopenia and C. difficile [39], and basophils and eosinophils in COVID [40]. In particular, 288 the strong associations suggest that these test results could be included in prognostic scoring systems for

295
Other study strengths include its focus on the impact of AMR in E. coli bacteraemia, the most common Gram-296 negative pathogen, thereby highlighting organism-specific associations with mortality; this study is also one 297 of the largest to assess the clinical impact of EUCAST breakpoints vs. alternative definitions of "active" 298 empirical antibiotic therapy for this pathology. We mitigated some limitations of previous studies: 299 heterogeneity of empirical antibiotic choice by selecting patients administered at least baseline amoxicillin-300 clavulanate; ambiguity of "inappropriate" therapy by using in-vitro-susceptibility-based definitions to define 301 "active" vs "inactive" therapy, and assessing the MIC continuum of susceptibility.

303
Limitations include the lack of data on community-prescribed antibiotics prior to hospital admission, in 304 particular because AMR potentially contributes to failure of community treatment, need for hospitalisation 305 and illness severity at admission. Since some of our model variables (e.g. vital signs, blood tests) capture 306 illness severity at admission, we may have underestimated the overall association between MIC and mortality, 307 by adjusting for factors that may mediate the pre-hospital impact of AMR. We cannot exclude the possibility 308 that other MIC values (e.g., 16/2 mg/L) may have smaller associations with mortality, particularly for 309 infections with a non-urinary source, which our study may have been insufficiently powered to detect. This is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 23, 2022. ; https://doi.org/10.1101/2022.01.22.22269642 doi: medRxiv preprint homogeneous population, with implications for generalisability. Another limitation is that the study antibiotic is a combination of two drugs (amoxicillin and clavulanate) with variation between EUCAST and CLSI 313 approaches to susceptibility testing with the former using a fixed clavulanate concentration and the latter a 314 fixed ratio, which has implications for which isolates are reported as resistant [43]. Furthermore, in the UK,   is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 23, 2022.    is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint    493 †Including events up to one year before the index blood culture.

494
. CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 23, 2022.  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 23, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 23, 2022.  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint    is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 23, 2022.