205 - Large Multi-site Clinical Field Study Characterizing Contamination Levels in Patient Used Endoscopes After Manual Cleaning

Background: Multiple outbreaks multi-drug resistant organisms (MDROs), have been associated with flexible endoscopes resulting in unacceptable patient mortality and morbidity. Evidence highlights the importance of effective cleaning to achieve effective high-level disinfection (HLD). This study presents an analysis of more than 700,000 measurements of adenosine-triphosphate (ATP) contamination levels found in flexible endoscopes after manual cleaning.

Method: This 2018-2019 study consists of 702,768 measurements of ATP levels found in the suction/biopsy channel of patient-used: gastroscopes (267,533 measurements from 223 sites), duodenoscopes (123,697 measurements from 161 sites), colonoscopes (252,249 measurements from 229 sites), and bronchoscopes (59,289 measurements from 107 sites) after manual cleaning. Sites were located across the US and employed protocols that included routine cleaning verification performed by the reprocessing technicians using a handheld luminometer and the associated ATP water test (3M™ Clean-Trace™).

Results: Figure 1 shows a boxplot analysis of the ATP levels by endoscope type. Upper gastrointestinal (GI) endoscopes (gastroscopes and duodenoscopes) show a significantly (p-value < 0 .005) greater level of ATP contamination after manual cleaning. The pairwise mean differences are all significant (p-values < 0 .005) except for colonoscopes when compared to bronchoscopes (p-value=0.203). Also shown on Figure 1 is a literature supported adequate cleanliness value of 200RLUs (=2.3LOG(RLUs)) (M.J. Alfa, et al.; American Journal of Infection Control; 41; (2013); 245-248 and 249-253; ANSI/AAMI ST91:2015). A 95% confidence tolerance interval analysis performed against this literature value (Table 1) showed that a high number of gastroscopes (12%) and duodenoscopes (10%) are not adequately clean. Figure 2 shows a boxplot analysis of the data set by endoscope type and by site. There is significant (p-value < 0 .005) site to site variability for all endoscope types as demonstrated by variation in mean values, box size, and many outliers.

Conclusions: This study highlights the importance of using a quantitative cleaning verification method to better understand process capability and provide more robust quality assurance for manual cleaning. There were significant differences in the level of cleanliness between upper GI scopes and lower GI scopes and bronchoscopes. When compared to a literature supported level for adequate cleanliness, upper GI scopes resulted in fail rates in excess of 10%. Furthermore, there was significant site to site variability and many outliers falling well beyond the normal process envelope, representing significant cleaning lapses. Root causes to these concerning findings could range from inadequate execution of the cleaning protocol, to device design, age and existing damage that could prevent achieving adequate cleaning and possibly impairing the effectiveness of HLD.