Worst-case soiling levels for patient-used flexible endoscopes before and after cleaning

Abstract

Background: The soiling levels of patient-used narrow-lumened flexible endoscopes were assessed for bronchoscopes, duodenoscopes, and colonoscopes. The effect of cleaning on the soil composition and concentration was evaluated. Design: Suction channels from 10 each of bronchoscopes, duodenoscopes used for endoscopic retrograde cholangiopancreatography, and colonoscopes were assessed immediately after patient use for the levels of bilirubin, hemoglobin, protein, sodium ion, carbohydrate, endotoxin, and viable bacteria. Another 10 suction channels of each type of endoscope were evaluated for the same components after routine cleaning but before processing by high-level disinfection or sterilization for subsequent clinical use. Results: Recognizing that only soluble components could be quantified, the worst-case soil levels in the suction channels (the average surface area of these channels was 45.6 cm 2, 149.8 cm, 2 and 192.0 cm 2 for bronchoscopes, duodenoscopes, and colonoscopes, respectively) were protein 115 μg/cm 2, sodium ion 7.4 μmol/cm 2, hemoglobin 85 μg/cm 2, bilirubin 299 nmol/cm 2, carbohydrate 29.1 μg/cm 2, endotoxin 9852 endotoxin units/cm 2, and bacteria 7.1 (log 10) colony-forming units (CFU)/cm 2. Colonoscopes had 4 to 5 times greater soiling on average compared with the other endoscope types. Routine cleaning reduced the levels of bilirubin to below the limits of detection for all endoscopes evaluated (limits of detection were <1 nmol/mL). After cleaning, residual hemoglobin was detectable in bronchoscopes only. After cleaning, the levels of protein, endotoxin, and sodium ion all were reduced fivefold to tenfold for all types of endoscopes. Carbohydrate was reduced to lower than the limit of detection for all endoscopes after cleaning, except the duodenoscopes. The average load of viable bacteria was reduced from 3 log 10 to 5 log 10 CFU/cm 2 (which represents 5.9-9.5 log 10 CFU/endoscope channel) after patient use to approximately 2 log 10 CFU/cm 2 (which represents 3.2-5.3 log 10 CFU/endoscope channel) after cleaning. Conclusions: These data demonstrated that cleaning effectively reduced or eliminated many components of soil, but a substantial amount of viable bacteria and protein remained. Hemoglobin levels in before samples indicated that blood was not present in high concentrations in the suction channels of the majority of flexible endoscope samples. Soil that mimics the worst-case composition from patient-used endoscopes would be ideal for simulated-use studies for such medical devices. (AJIC Am J Infect Control 1999;27:392-401)