Burden of Candida-related vascular graft infection: a nested-case control study.

PURPOSE: We aimed to assess risk factors of candida-related Vascular Graft Infections (VGIs). METHODS: We did a case-control study (1:4) matched by age and year of infection, nested in a cohort of patient with a history of VGIs. Cases were defined by a positive culture for Candida spp. in biological samples and controls were defined by a positive culture for bacterial strains only in biological samples. Risk factors for Candida-related VGIs were investigated using multivariate logistic regression. Mortality were compared using survival analysis. RESULTS: 16 Candida-related VGIs were matched to 64 bacterial-related VGIs. The two groups were comparable regarding medical history and clinical presentation. Candida-related VGIs were associated with bacterial strains in 88% (14/16). Gas/fluid-containing collection on abdominal CT scan and the presence of an aortic endoprosthesis were risk factors for Candida spp.-related VGIs [RRa 10.43 [1.81-60.21] p = 0.009 RRa and 6.46 [1.17-35.73] p = 0.03, respectively]. Candida-related VGIs were associated with a higher mortality when compared to bacterial-related VGIs (p = 0.002). CONCLUSIONS: Candida-related VGIs are severe. Early markers of Candida spp. infection are needed to improve their outcome. The suspicion of aortic endoprosthesis infection may necessitate probabilistic treatment with antifungal agents.

Severe secondary peritonitis: impact of inappropriate initial antimicrobial therapy on prognosis and potential carbapenem-sparing.

BACKGROUND: Initiation of antimicrobial therapy (IAT) with broad-spectrum antibiotics is usual in Intensive Care Unit (ICU) patients with secondary peritonitis. Carbapenems are widely proposed by recent guidelines contrasting with current antibiotic stewardship policies of carbapenem-sparing. However, prognosis of inappropriate IAT remains unclear in these patients and broad-spectrum antibiotics are probably overused. We aimed to assess the role of inappropriate IAT in ICU patients with secondary peritonitis and the use of carbapenems in our IAT regimens. METHODS: We performed a retrospective analysis during a six-year period including 131 ICU patients with secondary peritonitis. We collected data concerning comorbidities, source and severity of peritonitis, management of IAT, peritoneal samples and outcome. RESULTS: Forty-one patients presented with community acquired peritonitis (CAP) and 90 with postoperative peritonitis (POP). Thirty-seven (28.2%) patients died during ICU stay. IAT was inappropriate in 35 (26.7%) patients. Inappropriate IAT was not associated with reduced survival with respectively 26 (27%) deaths when IAT was adequate and 11 (31.4%) deaths when IAT was inadequate (P=0.87). Inappropriate IAT was not associated with the need of re-operation and duration of ICU stay. Carbapenems were delivered in 29 patients but were only necessary for eight patients without alternative treatment. CONCLUSIONS: In our study, inappropriate IAT was not associated with a worse prognosis and carbapenems were overused. Extensive delivery of carbapenems proposed by recent guidelines could be reconsidered in the management of these patients.

Continuous Infusion of High Doses of Cefepime in Intensive Care Unit: Assessment of Steady-State Plasma Level and Incidence on Neurotoxicity.

Continuous infusion (CI) with high doses of cefepime is recommended in the empirical antimicrobial regimen of critically ill patients with suspected Gram-negative sepsis. This study aimed to determine factors associated with cefepime overdosing and the incidence of cefepime-induced neurotoxicity (CIN) in these patients. We performed a retrospective study including all patients receiving cefepime treatment between January 2019 and May 2022. The plasma level of cefepime defining overdosing was over 35 mg/L. Neurotoxicity was defined according to strict criteria and correlated with concomitant steady-state concentration of cefepime. Seventy-eight courses of cefepime treatment were analyzed. The mean cefepime plasma level at steady state was 59.8 ± 29.3 mg/L, and overdosing occurred in 80% of patients. Renal failure and a daily dose > 5 g were independently associated with overdosing. CIN was present in 30% of patients. In multivariate analysis, factors associated with CIN were chronic renal failure and a cefepime plasma concentration ≥ 60 mg/L. CIN was not associated with mortality. Overdosing is frequent in patients receiving high doses of cefepime by CI. Steady-state levels are higher than targeted therapeutic pharmacokinetic/pharmacodynamic objectives. The risk of CIN is important when the plasma concentration is ≥60 mg/L.