During the study period, 3188 stool samples from 2189 ICU patients were sent to the microbiology laboratory (Fig. 1). Patients tested represented 6.0% of all patients admitted (2010–2015: 41.415 admissions of 36.477 patients). Ninety-four samples (2.9%) were rejected by the laboratory due to several reasons (solid stool, scarcity of material).
Overall, 5.747 tests were performed; only six were positive (0.1%).
Only four out of 2500 Campylobacter cultures yielded a positive result (0.16% of all tests performed, 0.01% of all patients admitted); three samples were positive for Campylobacter jejuni, and one for Campylobacter coli. Symptoms in all four patients with Campylobacter infection had started less than 48 h after hospital admission implying that these patients had community acquired diarrhea. None of these patients received specific antibiotic therapy as symptoms had already resolved spontaneously. Length of ICU stay after diagnosis of Campylobacter enteritis was not increased, since all four patients were transferred to the normal ward the next day. None of the patients with Campylobacter enteritis died during their hospital stay.
Testing for Salmonella and Shigella (Salmonella–Shigella culture) was done for 2488 samples (1678 patients), two of these were positive for Salmonella enterica (0.08% of all tests performed, 0.005% of all patients), and none were positive for Shigella. Salmonella enterica was detected in the accidental submission of stool from two chronic shedders 20 and 23 days after hospital admission. The patients were not treated antimicrobially due to lack of symptoms. Detection of infection did not delay hospital dismissal or further transfer to another hospital. The main diagnoses in patients with Campylobacter enteritis were pneumonia (patient no. 1), postoperative due to thymectomy (patient no. 2) and vascular graft (patient no. 3), as well as epileptical attack (patient no. 4). Underlying comorbidities in these patients included Korsakov syndrome (patient no. 1), myasthenia gravis, arterial hypertension (patient no. 2), chronic obstructive pulmonary disease, reflux disease, sigma diverticulitis (patient no. 3), and cerebral infarction and arterial hypertension (patient no. 4). Patients with Salmonella detection were not on any immunosuppressive medication; however, one patient was suffering from head and neck cancer requiring multiple surgeries. The other patient with Salmonella infection was admitted due to subarachnoid hemorrhage.
None of the 759 samples tested for enteropathogenic Yersinia spp. yielded positive results (Fig. 1).
Regarding ICU characteristics, all six patients tested positive for Salmonella or Campylobacter required vasopressor support (mean 5.7 days ± 6.1) during their stay. Only one patient received vasopressor infusion on the day of the diagnosis, and the charts do not indicate that the dosage of vasopressor infusion was increased in patients due to diarrhea. Five out of six patients were in need of mechanical ventilation, although only one patient was on mechanical ventilation at the time of diagnosis. Length of mechanical ventilation was < 3 days in all three patients with Campylobacter enteritis. None of the patients required renal replacement therapy.
In contrast, C. difficile toxin testing yielded positive results in 242 (GDH antigen only) and 179 (GDH antigen and toxin A/B) of 2209 samples from 1654 patients (11.0 and 8.1% of all tests, 0.7 and 0.5% of all patients), the latter group fulfilling criteria for clinically relevant C. difficile infection in symptomatic patients [9]. 52/144 (36.1%) episodes of CDI occurred within 48 h of ICU admission, and 108/144 (75%) were antibiotic-associated.
Regarding stool tests in the overall hospital community, results between 2010 and 2015 in all departments of our hospital are illustrated in Fig. 2. Patients presenting to the emergency room (ER) with diarrhea showed the highest rates of infection with Campylobacter spp. (11.6% of all analyzed specimens) and Salmonella or Shigella spp. (2.4%). Also, patients presenting to the infectious diseases department showed higher infection rates with Campylobacter (4.6%) and Salmonella or Shigella spp. (2.9%). Infection rates with Yersinia were overall very low with a maximum rate of infection in general outpatients (0.2% of samples). The lowest rate of bacterial infection in patients presenting with diarrhea was observed in patients treated at the ICU and bone marrow transplant unit.
Economic impact was estimated based on national medical fee schedules in Germany. Considering usual local discounts, we conferred costs of 20 € for enteric bacteria stool culture and 10 € for C. difficile testing adding up to 50,000 € (enteric bacteria stool culture) and 22,090 € (C. difficile), total 72,090 €.