P-1311. Community-associated carbapenem-resistant gram-negative bacilli infection in Japan

TL;DR

This study examines the characteristics and outcomes of carbapenem-resistant Gram-negative bacilli infections in Japan, finding that hospital-onset cases have worse outcomes than community-associated cases.

Contribution

The study provides new insights into the clinical differences and outcomes of community-associated CR-GNB infections in Japan, a high-income country.

Findings

Hospital-onset CR-GNB infections are associated with higher in-hospital mortality and longer hospital stays compared to community-associated infections.

Aeromonas species are more prevalent in community- and healthcare-associated infections, while Stenotrophomonas maltophilia is more common in hospital-onset cases.

Thirty-day mortality is lowest in community-associated cases and highest in hospital-onset cases.

Abstract

Carbapenem-resistant Gram-negative bacilli (CR-GNB) infections are associated with high mortality and increased healthcare costs. While certain resistant pathogens, such as extended-spectrum β-lactamase-producing organisms, are increasingly identified in the community, reports of community-associated CR-GNB infections remain limited, particularly in high-income countries like Japan.Table 1.Demographic characteristics of patients with community-associated, healthcare-associated, and hospital-onset CR-GNB infections.CR-GNB, carbapenem-resistant Gram-negative bacilli; CA, community-associated; HA, healthcare-associated; HO, hospital-onset; IQR, interquartile range; IV, intravenous. a P values for comparisons among the three groups (CA, HA, HO) were calculated using the Kruskal–Wallis test for continuous variables and the chi-squared test for categorical variables. b Pairwise comparisons were performed using the Mann–Whitney U test for continuous variables and Fisher’s exact test for categorical variables. For multiple comparisons, a Bonferroni-corrected threshold of P <0.025 was considered statistically significant. Asterisks indicate statistical significance at the corrected threshold.Table 2.Causative pathogens and sites of infection among patients with community-associated, healthcare-associated, and hospital-onset CR-GNB infections.CR-GNB, carbapenem-resistant Gram-negative bacilli; CA, community-associated; HA, healthcare-associated; HO, hospital-onset. a P values for comparisons among the three groups (CA, HA, HO) were calculated using the Kruskal–Wallis test for continuous variables and the chi-squared test for categorical variables. b Pairwise comparisons were performed using the Mann–Whitney U test for continuous variables and Fisher’s exact test for categorical variables. For multiple comparisons, a Bonferroni-corrected threshold of P <0.025 was considered statistically significant. Asterisks indicate statistical significance at the corrected threshold.

Demographic characteristics of patients with community-associated, healthcare-associated, and hospital-onset CR-GNB infections.

CR-GNB, carbapenem-resistant Gram-negative bacilli; CA, community-associated; HA, healthcare-associated; HO, hospital-onset; IQR, interquartile range; IV, intravenous. a P values for comparisons among the three groups (CA, HA, HO) were calculated using the Kruskal–Wallis test for continuous variables and the chi-squared test for categorical variables. b Pairwise comparisons were performed using the Mann–Whitney U test for continuous variables and Fisher’s exact test for categorical variables. For multiple comparisons, a Bonferroni-corrected threshold of P <0.025 was considered statistically significant. Asterisks indicate statistical significance at the corrected threshold.

Causative pathogens and sites of infection among patients with community-associated, healthcare-associated, and hospital-onset CR-GNB infections.

CR-GNB, carbapenem-resistant Gram-negative bacilli; CA, community-associated; HA, healthcare-associated; HO, hospital-onset. a P values for comparisons among the three groups (CA, HA, HO) were calculated using the Kruskal–Wallis test for continuous variables and the chi-squared test for categorical variables. b Pairwise comparisons were performed using the Mann–Whitney U test for continuous variables and Fisher’s exact test for categorical variables. For multiple comparisons, a Bonferroni-corrected threshold of P <0.025 was considered statistically significant. Asterisks indicate statistical significance at the corrected threshold.

From April 2019 to March 2022, we prospectively enrolled patients with CR-GNB infections through the Multidrug-Resistant Organisms Clinical Research Network (MDR-net), comprising 13 tertiary care centers in Japan. We compared patient demographics, clinical characteristics, and outcomes across community-associated (CA), healthcare-associated (HA), and hospital-onset (HO) infections.Table 3.Clinical outcome of patients with community-associated, healthcare-associated, and hospital-onset CR-GNB infections.CR-GNB, carbapenem-resistant Gram-negative bacilli; CA, community-associated; HA, healthcare-associated; HO, hospital-onset. a P values for comparisons among the three groups (CA, HA, HO) were calculated using the Kruskal–Wallis test for continuous variables and the chi-squared test for categorical variables. b Pairwise comparisons were performed using the Mann–Whitney U test for continuous variables and Fisher’s exact test for categorical variables. For multiple comparisons, a Bonferroni-corrected threshold of P <0.025 was considered statistically significant. Asterisks indicate statistical significance at the corrected threshold. 95% confidence intervals for death-related outcomes by onset: Death discharge – CA: 11.6% (95% CI: 1.9–21.4), HA: 17.0% (7.3–26.6), HO: 35.8% (30.6–41.0); 30-day mortality – CA: 9.3% (0.5–18.1), HA: 10.2% (2.4–17.9), HO: 23.8% (19.1–28.4).Figure 1.Kaplan–Meier survival curves showing patient days from the first positive culture to death for patients with community-associated, healthcare-associated, and hospital-onset CR-GNB infections.CR-GNB, carbapenem-resistant Gram-negative bacilli

Clinical outcome of patients with community-associated, healthcare-associated, and hospital-onset CR-GNB infections.

CR-GNB, carbapenem-resistant Gram-negative bacilli; CA, community-associated; HA, healthcare-associated; HO, hospital-onset. a P values for comparisons among the three groups (CA, HA, HO) were calculated using the Kruskal–Wallis test for continuous variables and the chi-squared test for categorical variables. b Pairwise comparisons were performed using the Mann–Whitney U test for continuous variables and Fisher’s exact test for categorical variables. For multiple comparisons, a Bonferroni-corrected threshold of P <0.025 was considered statistically significant. Asterisks indicate statistical significance at the corrected threshold. 95% confidence intervals for death-related outcomes by onset: Death discharge – CA: 11.6% (95% CI: 1.9–21.4), HA: 17.0% (7.3–26.6), HO: 35.8% (30.6–41.0); 30-day mortality – CA: 9.3% (0.5–18.1), HA: 10.2% (2.4–17.9), HO: 23.8% (19.1–28.4).

Kaplan–Meier survival curves showing patient days from the first positive culture to death for patients with community-associated, healthcare-associated, and hospital-onset CR-GNB infections.

CR-GNB, carbapenem-resistant Gram-negative bacilli

Among 426 patients with CR-GNB infections (CA, n=43; HA, n=59; HO, n=324), most were elderly, with comparable Charlson Comorbidity Index scores across groups (median [IQR], 2 [1–4]; P = 0.17). Pseudomonas aeruginosa was the most frequently isolated pathogen in all groups. Aeromonas species were significantly more prevalent in CA and HA compared to HO (CA: 23.3%, HA: 18.6%, HO: 2.2%; P < 0.001 overall), whereas Stenotrophomonas maltophilia was predominantly isolated in HO (HO: 17.3%, HA: 5.1%, CA: 0%; P = 0.001 overall). Clinical outcomes differed significantly by infection type. Compared to CA, HO infections were associated with longer hospital stay (68 vs. 17 days, P < 0.001), lower discharge to home (33.6% vs. 86.1%, P < 0.001), and higher in-hospital mortality (35.8% vs. 11.6%, P = 0.001). Thirty-day mortality also differed, with Kaplan–Meier analysis showing the lowest survival in HO (log-rank P = 0.013). In contrast, HA outcomes were comparable to CA, with no significant differences in length of stay (23 vs. 17 days, P = 0.25), discharge to home (69.5% vs. 86.1%, P = 0.061), or in-hospital mortality (17.0% vs. 11.6%, P = 0.58). These findings suggest that poor outcomes in CR-GNB infections were mainly driven by HO cases.

The clinical features and outcomes of CR-GNB infections differ markedly by onset, with HO infections associated with poorer outcomes compared to CA and HA infections.

Yasufumi Matsumara, MD, PhD, Beckman Coulter: Research support for a collaborative project|Precision System Science: Research support for a collaborative project Takashi Matono, MD, PhD, FACP, Gilead Sciences: Honoraria|GSK: Honoraria|Meiji Seika Pharma: Honoraria|MSD: Honoraria|Pfizer: Honoraria Naoya Itoh, MD, DTM&H, PhD, Asahi Kasei Pharma Corporation: Honoraria|AstraZeneca K.K: Honoraria|BD Co, Ltd: Honoraria|bioMérieux Japan Ltd: Honoraria|Gilead Sciences Inc: Honoraria|GlaxoSmithKline: Honoraria|Meiji Seika Pharma Co, Ltd: Honoraria|MSD K.K: Honoraria|Pfizer;: Honoraria|shimadzu co ltd: Grant/Research Support|Shionogi Co, Ltd: Honoraria Tetsuya Suzuki, M.D., Ph.D., Meiji Seika Pharma: Honoraria David van Duin, MD, PhD, British Society for Antimicrobial Chemotherapy: Editor stipend|Merck: Advisor/Consultant|Merck: Grant/Research Support|Pfizer: Advisor/Consultant|Roche: Advisor/Consultant|Shionogi: Advisor/Consultant Yohei Doi, MD, PhD, GSK: Advisor/Consultant|Meiji Seika Pharma: Advisor/Consultant|Shionogi: Advisor/Consultant|Shionogi: Honoraria Sho Saito, Dr, Shionogi & Co., Ltd.: Grant/Research Support|SUNSTAR: Grant/Research Support

Figures