Risk factors for recurrent healthcare-facility associated Clostridioides difficile infection in a Swedish setting.

OBJECTIVE: The objectives were to determine the risk factors for recurrent healthcare facility-associated Clostridioides difficile infection (HCF-CDI) in a high CDI incidence, low antibiotic use setting and to determine if length of cefotaxime exposure is a risk factor for recurrent HCF-CDI. METHODS: The risk factors for recurrent HCF-CDI were evaluated with a retrospective nested case control study based on chart reading. The risk factors were evaluated univariately and multivariately. Length of risk antibiotic exposure was evaluated further in a subanalysis. RESULTS: Risk factors for recurrent HCF-CDI were renal insufficiency (25.4% of cases compared to 15.4% of controls p = 0.006) and metronidazole treatment of initial CDI episode (88.4% compared to 71.7% p = 0.01). Exposure to cefotaxime and risk for recurrent CDI showed a dose-dependent relationship (linear by linear p = 0.028). CONCLUSIONS: Renal insufficiency and metronidazole treatment were independent risk factors for recurrent HCF-CDI in our setting. The relationship between cefotaxime exposure and risk for recurrent HCF-CDI, dose-dependent, could be evaluated further in a setting with high cefotaxime use.

Clostridioides difficile incidence related to in-hospital cephalosporin use: a tale of two highly comparable hospitals.

Background: Antibiotic treatment is a well-known risk factor for healthcare facility-associated Clostridioides (Clostridium) difficile infection (HCF-CDI). Antibiotic stewardship programmes (ASPs) targeting high-risk antibiotics have been shown to decrease HCF-CDI incidence. HCF-CDI incidence is high in Nordic countries despite relatively low antibiotic use in hospital. Objectives: To determine if HCF-CDI incidence was modified by a hospital ASP that restricted cephalosporin use. Methods: The effects of an ASP on HCF-CDI incidence were evaluated in a two-centre setting using a retrospective design. We exploited a strategy of both individual case ascertainment based on chart reviews and aggregated data from the hospitals. Cases were attributed to the antibiotics given prior to disease onset, in proportion to the number of DDDs used. Three periods were studied: 2007 (before the ASP), 2012 and 2015. Results: At the ASP hospital, cephalosporin use decreased by 87% and the number of HCF-CDI/1000 hospital admissions decreased significantly from 2.25 (2007) to 1.16 (2015) (P = 0.0014). The corresponding results at the non-ASP hospital showed a non-significant increase from 2.09 to 2.38. A high number of cases could be attributed to cephalosporins at both hospitals. The increased use of other broad-spectrum antibiotics, e.g. piperacillin/tazobactam, at the ASP hospital was not associated with offsetting increases in attributable HCF-CDI cases. Conclusions: Decreased use of cephalosporins is an effective strategy to decrease HCF-CDI incidence over time in a setting with high incidence and low antibiotic use.

Clostridioides difficile outbreak detection: Evaluation by ribotyping and whole-genome sequencing of a surveillance algorithm based on ward-specific cutoffs.

OBJECTIVE: We evaluated the performance of an early-warning algorithm, based on ward-specific incidence cutoffs for detecting Clostridioides difficile transmission in hospitals. We also sought to determine the frequency of intrahospital Clostridioides difficile transmission in our setting. DESIGN: Diagnostic performance of the algorithm was tested with confirmed transmission events as the comparison criterion. Transmission events were identified by a combination of high-molecular-weight typing, ward history, ribotyping, and whole-genome sequencing (WGS). SETTING: The study was conducted in 2 major and 2 minor secondary-care hospitals with adjacent catchment areas in western Sweden, comprising a total population of ∼480,000 and ∼1,000 hospital beds. PATIENTS: All patients with a positive PCR test for Clostridioides difficile toxin B during 2020 and 2021. METHODS: We conducted culturing and high-molecular-weight typing of all positive clinical samples. Ward history was determined for each patient to find possible epidemiological links between patients with the same type. Transmission events were determined by PCR ribotyping followed by WGS. RESULTS: We identified 4 clusters comprising a total of 10 patients (1.5%) among 673 positive samples that were able to be cultured and then typed by high-molecular-weight typing. The early-warning algorithm performed no better than chance; patient diagnoses were made at wards other than those where the transmission events likely occurred. CONCLUSIONS: In surveillance of potential transmission, it is insufficient to consider only the ward where diagnosis is made, especially in settings with high strain diversity. Transmission within wards occurs sporadically in our setting.