Multidrug resistant Clostridium difficile ribotype 027 in southwestern Virginia, 2007 to 2013.

The excess from fecal samples submitted to a centralized laboratory in Roanoke, Virginia for routine C. difficile testing was used for this research study. We tested all samples, including any formed samples usually not assayed in diagnostic laboratories. Our first aim was to rank ribotypes by their frequency. Between 2007 and 2013, fluoroquinolone resistant 027 (027FQR), a multi-drug resistant ribotype, was 32% of 3118 Clostridium difficile isolates and the most common of 128 ribotypes. 027FQR was in 45% of cytotoxin positive but only 17% of cytotoxin negative fecal samples (p = 0.001) and 34% of unformed but only 21% of formed stool samples (p = 0.001), strong associations with features of symptomatic infection. Conversely, 014/020 (10% of isolates, third most common ribotype) was more often in unformed than formed stools (14% versus 9%; p = 0.002) and in cytotoxin negative than cytotoxin positive samples (11% versus 8%, p = 0.01). Fecal lactoferrin levels, an indication of intestinal inflammation, were significantly higher with 027FQR than with 014/020 infections (median 308 versus 26 ng/mL, p = 0.02). 027FQR fecal bioburdens and toxin levels were significantly higher than their 014/020 equivalents (median 104.1 versus 103.2/g feces, p = 0.01; median TcdA 58.7 versus 1.3 ng/g feces, p = 0.04; median TcdB 43.4 versus 0.3 ng/g feces, p = 0.001). Binary toxin was present in 40% of 027FQR positive samples but none of the 014/020 or non-toxigenic C. difficile positive samples. 027FQR made no more TcdA/cell than did 014/020 (p = 0.7) but did make close to significantly more TcdB/cell (p = 0.08).

Clostridium difficile prevalence rates in a large healthcare system stratified according to patient population, age, gender, and specimen consistency.

We evaluated Clostridium difficile prevalence rates in 2,807 clinically indicated stool specimens stratified by inpatient (IP), nursing home patient (NH), outpatient (OP), age, gender, and specimen consistency using bacterial culture, toxin detection, and polymerase chain reaction (PCR) ribotyping. Rates were determined based on the detection of toxigenic C. difficile isolates. We identified significant differences in the rates between patient populations and with age. Specimens from NH had a higher rate (46%) for toxigenic C. difficile than specimens from IP (18%) and OP (17%). There were no gender-related differences in the rates. Liquid specimens had a lower rate (15%) than partially formed and soft specimens (25%) and formed specimens (18%) for the isolation of toxigenic C. difficile. The nontoxigenic rate was lowest for NH (4%) and highest for patients<20 years of age (23%). We identified 31 different toxigenic ribotypes from a sampling of 190 isolates that showed the lowest diversity in NH. Fluoroquinolone resistance was observed in 93% of the 027 isolates, all of the 053 isolates, and in four other ribotypes. We observed different rates for toxigenic C. difficile in stratified patient populations, with the highest rate for NH, a low overall nontoxigenic rate, and fluoroquinolone resistance.