Hospital investigation
On Thursday April 24, the public health unit shared its findings with the hospital. Automated machines for disinfection are used after the scopes are manually brushed, cleaned and tested for leaks. While the “dirty” and “clean” areas are in the same room, a barrier exists between them (a counter) and policies and procedures are in place to attempt to keep the areas separate.
It was noted that there is limited storage for clean endoscopes (they were short one hook for hanging), therefore, one randomly chosen, processed endoscope is always stored in the automated machine (with the lid open).
To identify any other potential sources of enteric bacteria to patients other than the endoscope itself, a discussion was held regarding process flow of patients undergoing an endoscope procedure and all actions and equipment that occur before, during and after a procedure. Patients receive the same instructions for preparation; however, specific preparation products varied and were obtained through different community pharmacies, ruling out preparation products as a source of infection. No other items or procedures were identified at that time as a potential source of enteric bacteria.
The hospital performed 15 to 20 colonoscopies per weekday and had 10 endoscopes used for colonoscopies. Each scope is stored with a printout from the disinfectant machine confirming the process was completed and when the scope is used. A copy of the printout is subsequently attached to the appropriate patient’s chart.
The hospital agreed to review patient records of Cases A, B and C to identify if any commonalities with staff or equipment were present and to confirm the quality control verification check of the scope cleaning.
At 5:45 p.m. that same day, the health unit received an email from the hospital that their record review revealed the same endoscope was used on Case A, Case B and Case C. No other commonality was found between cases, including no common staff among all three cases. The public health unit and the hospital agreed to remove this endoscope, identified as endoscope #17, from use until an investigation was complete.
Public health investigation
At the request of the public health, the hospital generated
a list of patients who received a colonoscopy at the hospital within a defined time period. The suspect case definition used to generate this list included patients who had a colonoscopy procedure at the hospital with endoscope # 17 between
March 17 and April 3, 2014. The time period used in this case definition was informed by the longer than expected incubation period reported by Dwyer et al (1987) (8) for transmission through colonoscopy. This produced a list of 24 people, not including the three people identified as cases.
Results of the health unit and hospital investigation were shared with PHO. PHO provided a procedure from the Public Health Agency of Canada (PHAC) (9) for sampling an endoscope to test for bacterial contamination. Several limitations were identified in the methodology; however, public health and the hospital agreed to proceed with this testing.
On Tuesday April 29, confirmation was received from PHO lab that all three cases were of the same genotype (phage type 8) and that the PFGE patterns were identical. Based on this information, the public health unit moved forward with active case finding and officially declared a suspect outbreak. Initial calls were to be made to patients who were scoped the same day and on all the days in between the cases with endoscope #17. If interviews yielded suspect cases, then the calling would widen to include the original range of dates found within the case definition. The rationale for this was to lessen the number of patients who may become worried or anxious until additional information was discovered which would necessitate contacting all patients exposed during the period of concern. An initial list of 14 patients was identified to be contacted.
RESULTS
The endoscope was sampled on May 1. Public health and the hospital performed the sampling. Sample results were received on May 7. No salmonella species was detected.
The patients were contacted and asked if they experienced symptoms of gastrointestinal upset before or after their procedure. It was anticipated that some patients would
have been experiencing gastrointestinal symptoms before
the procedure and these symptoms were the reason for the procedure. Since patients were required to do a pre-procedure colonic cleansing, the questions were also carefully worded to ask about any changes in the type or intensity of symptoms after the procedure. Patients were also asked if they would be willing to submit a stool sample for salmonella testing.
Of the 14 patients contacted, eight submitted a stool sample. All submitted samples were negative for salmonella.
None of the patients experienced fever after the procedure, which is a common symptom of salmonella infection. Of those that experienced abdominal pain and diarrhea, which are other common symptoms of salmonella infection, it was neither severe nor different from what they were experiencing before the procedure.
Based on these results, no further patients were contacted.
DISCUSSION
This outbreak supports the suggestions of Dwyer et al that transmission of salmonella can occur during colonoscopy with contamination occurring directly to the lower gastrointestinal tract and that when this occurs, the incubation period may be longer than with the usual ingestion route.
No further evidence of symptomatic or non-symptomatic salmonella infection was found among colonoscopy patients from the hospital who had a procedure performed with endoscope #17 during the time frame of the affected individuals. In addition, eight patients were lab tested for salmonella, all with negative results. While lab testing of patients was performed some weeks after they would have been expected to have salmonella if infected, carriers can shed the bacteria for years and 5 % of patients recovering from non-typhoidal salmonellosis can shed the bacteria for 20 weeks (10).