BACKGROUND: Evaluating the selective pressure of antimicrobials on bacteria is important for promoting antimicrobial stewardship programs (ASPs). The aim of this study was to assess the selective pressure of antimicrobials by evaluating their use (carbapenem [CBP] and CBP-sparing therapy) over time and the detection status of CBP-resistant organisms using multicenter data.
METHODS: Among the facilities whose data were registered in the Japan Surveillance for Infection Prevention and Healthcare Epidemiology from 2017 to 2020, those that had data on the use of CBP and CBP-sparing therapy (fluoroquinolones [FQs], cefmetazole [CMZ], piperacillin-tazobactam [PIP/TAZ], ampicillin-sulbactam [ABPC/SBT], ceftriaxone/cefotaxime [CTRX/CTX], CAZ (ceftazidime), cefepime [CFPM], and aminoglycosides [AGs]) as well as on CBP-resistant Enterobacterales (CRE) and CBP-resistant Pseudomonas aeruginosa (CRPA) detection were included. Alcohol-based hand rubbing (ABHR) usage was also analyzed. Regression analyses, including multivariable regression analysis, were performed to evaluate trends. The association of antimicrobial use density (AUD) with CRE and CRPA detection rates was evaluated.
RESULTS: In 28 facilities nationwide, CBP, FQ, CAZ, AG, and PIP/TAZ use decreased over the 3-year period, whereas the use of CMZ, ABPC/SBT, CTRX/CTX, CFPM, and ABHR as well as the rates of CRE and CRPA detection increased. The average AUD did not significantly correlate with CRE and CRPA detection rates. The multivariable regression analysis did not reveal any significant correlation between each AUD or ABHR and CRE or CRPA detection.
CONCLUSIONS: CBP and ABHR use showed a decreasing and an increasing trend, respectively, while CRPA and CRE detection rates exhibited a gradual increase. The considerably low CRE and CRPA detection rates suggest that slight differences in numbers may have been observed as excessive trend changes. Further investigation is warranted to evaluate selective pressure while considering the characteristics of ASP and the mechanisms underlying resistance.

OBJECTIVE: The Japan Surveillance for Infection Prevention and Healthcare Epidemiology (J-SIPHE) system aggregates information related to antimicrobial resistance (AMR) measures. We aimed to investigate the correlation between antibiotic use and AMR at a university hospital from 2013 to 2021 in a time series analysis using the J-SIPHE system. We also studied this correlation in each ward (inter-ward analysis).
METHODS: Data on antibiotic use and resistance rates were collected from the J-SIPHE system, except for the resistance rate in each ward, which was calculated from the source data prepared for this system.
RESULTS: Piperacillin/tazobactam use was positively correlated with piperacillin/tazobactam resistance in Escherichia coli and Klebsiella pneumoniae in the inter-ward analysis, and in Pseudomonas aeruginosa in both analyses. Carbapenem use was positively correlated with meropenem resistance in Enterobacter cloacae in the time series analysis and in P. aeruginosa in both analyses, and imipenem/cilastatin resistance in P. aeruginosa in inter-ward analysis. Quinolone use was positively correlated with levofloxacin resistance in E. coli in both analyses, and in K. pneumoniae in inter-ward analysis.
CONCLUSIONS: This is the first study to investigate the correlation between antibiotic use and AMR at a single hospital in time series and inter-ward analyses using the J-SIPHE system and data prepared for this system, suggesting that this system may be useful for promoting AMR measures.

During the coronavirus disease 2019 (COVID-19) pandemic, a change in the trend of infections was observed. However, there are few reports comprehensively assessing the impact of the early phase of COVID-19 on the trend of bacteria isolated.
We extracted the number of positive cultures of hospitalized patients for approximately 200 institutions using the Japanese national database. The outcome was the ratio of 10 species isolated in comparison to the total isolates for each month. Interrupted time-series analyses were conducted between 13 (from Jan-2019 to Jan-2020) and 8 (from May-2020 to Dec-2020) monthly data points.
A total of 369,210 isolates were involved. Differences in the level change for Streptococcus pneumoniae, Haemophilus influenzae, and Streptococcus pyogenes decreased significantly by 0.272 (95% confidence interval [CI]:0.192-0.352), 0.244 (95%CI:0.174-0.314), and 0.324 (95%CI:0.06-0.589), respectively. Bacteria transmitted by contact infection, such as Staphylococcus aureus, did not decrease. Differences in slope change were not significant in all species.
The ratios of isolated bacteria transmitted by droplet infection decreased immediately after the early phase of COVID-19 and maintained the same level. The awareness and behavioral changes toward increased COVID-19 prevention might have a substantial impact on the prevention of bacterial infections, especially droplet infections.

The clinical evidence for the effect of narrow-spectrum antimicrobial shortages on bacterial susceptibility is limited. Our purposes were to determine the affects of the disruption of most of the cefazolin (CEZ) supply in Japan on the susceptibility of pathogens and to analyze how long these changes persisted after the shortage of CEZ.
We performed an interrupted time series analysis using the Japanese Infectious Disease Nationwide database. We analyzed each pathogen before and after CEZ shortage in 52 university hospitals from 2018 to 2020. May to November 2019 was designated as the implementation term for CEZ shortage. The primary outcome was the susceptibility to CEZ and other antimicrobial agents. Amongst all pathogens isolated from facilities, we identified pathogens that were tested for susceptibility to CEZ.
Of the 26 pathogens identified, analysis was performed on a total of 36,346 isolates of five pathogens (Escherichia coli, Klebsiella pneumoniae, Klebsiella oxytoca, Proteus mirabilis, and Staphylococcus aureus). Amongst four Gram-negative pathogens with low susceptibility, there were no significant immediate changes after the CEZ shortage; however, the slope change significantly increased by 1.29 to 2.69% per month and continued to improve one year after the shortage. Regarding S. aureus, which was highly susceptible at the baseline, neither immediate change nor slope was significant.
This quasi-experimental analysis using a nationwide-large database revealed that restriction of use because of narrow-spectrum antimicrobial shortages may lead to improved susceptibility over the subsequent year. The results suggest that temporary switching of antimicrobial agents on a national scale could be effective.