How to Improve Reprocessing of Flexible Endoscopes Nationwide ? Data from the German Colorectal Cancer Screening Program

Background and Aims. International studies revealed prevalences of around 50% of microbiological contaminations in reprocessed �exible endoscopes. In Germany a system was installed where the quali�cation for refund for colonoscopies was made conditional on successfully passing twice annually a microbiological surveillance test of reprocessed endoscopes. is study is an implementation and outcome evaluation as well as a general discussion of the quality assessment assurance in colonoscopy in Germany. Methods. German data from 2003–2008 were analysed: number of endoscopic units performing therapeutic and/or screening colonoscopies; results of all microbiological surveillance tests of reprocessing quality; number of failed surveillance tests and retests; number of quali�cations for refund from the public health system cancelled due to repeated failure of microbiological surveillance tests. Results. Aer the introduction of the quality assessment assurance, the percentage of failed microbiological surveillance tests dropped signi�cantly and steadily from close to 17% to below 5%. Conclusions. is study evidences (1st) the successful implementation of the quality assessment assurance in Germany and (2nd) a substantial improvement in the quality of reprocessing �exible endoscopes achieved by these measures.


Introduction
Flexible endoscopes invariably become contaminated with microorganisms during clinical use.With the more widespread use of �exible endoscopy 30 years ago, nosocomial infections associated with oesophago-gastroduodenoscopy, colonoscopy, endoscopic retrograde cholangiopancreaticography (ERCP), or bronchoscopy were observed and described in the literature [1][2][3][4][5][6][7][8][9].A survey in the 70ies yielded an estimated risk of 1 per 10,000 endoscopic examinations (oesophagogastroduodenoscopies und colonoscopies) [10].More recently, optimistic estimates suggest a lower risk of 1 in 1.8 million [11] or 1 in 5 million procedures [12] given millions of endoscopic examinations worldwide.Data from the late nineteen-seventies report an infection rate of 0.74 and 0.93% following endoscopic retrograde cholecystopancreaticoscopy (ERCP) [7,8,13].e risk for transmission of, for example, Helicobacter pylori has been considered to be likely around of 2-3 per 1000 oesophagogastroduodenoscopies [14][15][16].e problem of possible infections originating from endoscopic examinations has been controversially discussed by the media, that is, in Germany [17] and the US [18].In Germany, this culminated in the repeated statement of a notable hygienist that he himself refrained from endoscopic examinations due to the risk of infection [19].

ISRN Endoscopy
e situation in Germany is different from other countries.In 2002, based on the results of the HYGEA study [45] con�rming similar studies in other countries [22,46] and a comparison of international guidelines for reprocessing �exible endoscopes [27] the commission for hospital hygiene and disease prevention at the Robert Koch institute, the German analogue to the Centers for Disease Control (CDC) published an evidence-based guideline for reprocessing �exible endoscopes and endoscopic accessories [47].In addition, a national colorectal cancer screening program via colonoscopy was initiated in autumn 2002.Before nationwide implementation of this screening program a quality assurance agreement was established by the Kassenärztliche Bundesvereinigung (KBV) and the stakeholder of the public health insurances to ensure that screening colonoscopy will be performed safely and without risk for the screened patient.According to this quality assurance agreement [48] both quali�cation indicators for the performing gastroenterologist were de�ned (see Section 4) and monitoring of colonoscope reprocessing by microbiological cultures was implemented.Certi�cation to participation in the colorectal cancer screening program was given by the local KV only aer documented ful�lling of these quali�cation requirements and �nancial funding for screening colonoscopy was coupled to compliance with the quality indicators and negative surveillance cultures of reprocessed colonoscopies.Under coordination by the central KBV and supervision and control by local Kassenärztliche Vereinigungen (KV) in the different states of the Federal Republic of Germany a nationwide system was established with certi�cated microbiological laboratories performing two times per year microbiological surveillance cultures from swabs from the tip of the colonoscope, from �ushing solution of disinfected channels of the instrument, and from water bottle solution of the optic �ushing system.e present study is an implementation and outcome evaluation of the German quality assurance agreement with the speci�c aims to analyse surveillance culture data from 2002 to 2008 in detail and to assess whether linking reprocessing quality indicators to funding of colonoscopy will improve colonoscope reprocessing quality in the long run.

Materials and Methods
e following nationwide data, per year and per association of statutory health insurance physicians (in German: Kassenärztliche Vereinigung (KV)), were made available through the framework of the quality assurance agreement for the time period of 2003 to 2008: the number of endoscopic units performing therapeutic and/or screening colonoscopies within each KV, the results of all microbiological surveillance tests of reprocessed endoscopic units (two per year per unit), the number of failed surveillance tests and failed retests, and the number of quali�cations for refund from the public health system cancelled due to repeated failure of microbiological surveillance tests.All surveillance tests were conducted by independent, accredited microbiological institutes/laboratories. e frequencies of actually performed hygiene surveillance tests per year were expressed as a percentage of the tests prescribed by the assurance system, that is, 100% are reached when two tests per year were conducted per endoscopic unit.e frequencies of failed microbiological surveillance tests per year (i.e., the number of necessary retests) were expressed as a percentage of the number of actually conducted routine surveillance tests.e frequencies of repeated failure of microbiological surveillance tests per year were also expressed as a percentage of the number of actually conducted initial routine surveillance tests.e frequencies of quali�cations for refund from the public health system cancelled due to repeated failure of microbiological surveillance tests were listed as raw numbers.
Analyses of variance for repeated measurements were used to statistically assess the time trends of the above detailed percentages.An overall alpha level of 5% was used.Pairwise comparisons between consecutive years were analysed by means of paired -tests where a Bonferroni correction for multiple testing was applied.

Results
e time trends from 2003 to 2008 of the following key microbiological performance parameters are detailed in Table 1: (1) e number of endoscopy units accredited in Germany for therapeutic and/or preventive colonoscopies.
(2) Percentages of actually performed surveillance tests (out of those prescribed by the quality assurance agreement).Since some associations of statutory health insurance physicians did not provide data for all years, two rows of data are presented: all available information (here the number of reporting associations may differ over the years) and complete information only (only those associations included that reported data for all years of the study period).e latter formed the basis for statistical tests.
(3) Percentages of failed microbiological surveillance tests out of all conducted tests.Again two rows are displayed: for all information available and for complete series only.
(4) Percentages of repeated failures of microbiological tests out of all conducted initial tests.Again two rows are displayed: for all information available and for complete series only.
(5) e number of quali�cations for refund from the public health system cancelled due to repeated failure of microbiological tests.
e time trend of percentages of conducted surveillance tests of all KV's with complete data over the whole study period (  ) are depicted in Figure 1.e overall trend is statistically signi�cant (  ).Consecutive pairwise comparisons only reveal a signi�cant change between 2003 and 2004.e time trend of percentages of failed tests (out of conducted tests) of all KV's with complete data (  ) is shown in Figure 2. e overall trend is statistically signi�cant (  ).Consecutive pairwise comparisons reveal

Discussion
Since colorectal cancer is one of the main cancers in Western societies accounting for more than 50.000deaths each year in the United States [49] and more than 30.000deaths each year in Germany [50], colonoscopy as the gold standard for early detection of adenomas and carcinomas in symptom-free individuals is considered as a mass screening tool to reduce incidence and mortality from colorectal cancer by detecting and removing potential precursor lesions.
Data on the process quality and the incidence of acute complications and data on the diagnostic yield of screening colonoscopy of the German colorectal cancer screening program were recently published [63,64].
In addition, Germany was the �rst country worldwide having realized continuous monitoring of colonoscope reprocessing by microbiological surveillance cultures of reprocessed endoscopes.

e German System: Implementation of a Quality Control Program for Endoscope Reprocessing.
In order to establish and maintain high-quality standards for screening colonoscopy the public health insurances and the KBV, the main players of the ambulance health system in Germany [65][66][67], arranged a quality assurance agreement before starting the German colorectal screening program by colonoscopy.According to this agreement [48] the following quality indicators for the performing gastroenterologist were de�ned as a marker for experience-a minimal number of more than 200 colonoscopies and 20 polypectomies per year-as a marker for performance-(electronic or videoprint) documentation of the ileocoecal valve or the terminal ileum and-as a marker for staff performance/equipment reprocessing quality-negative surveillance cultures of disinfected colonoscopes two times per year.e central issue in this quality management program consists in payment for performance and continuous documentation of the quality indicators for colonoscopy mentioned above.Each gastroenterologist, performing endoscopies in his office and ful�lling the quali�cation requirements, can participate in the German colorectal cancer screening program.For a nationwide colonoscopy screening program, availability of an experienced gastroenterologist as well as minimizing problems concerning infection risk are crucial steps for acceptance of the screening program.us, the German colorectal cancer screening program was coupled and coordinated with a quality assurance agreement implementing outcome monitoring of endoscope reprocessing.
e present analysis is based on the collected data of this quality control program.Completeness and quality of the data needs critical comments.In 2003, the �rst year of implementation of this control system, surveillance cultures were not performed twice a year but only once a year.For the years 2003 and 2004 in several federal states only limited data were available, collection was done in retrospective and completeness of the data was questionable.us, in general data quality and completeness are somewhat limited for the �rst two years, but become valid in 2005 and the following years.

German and International Guidelines in Endoscope Reprocessing: Do
We Need Surveillance Cultures?e German guideline for reprocessing �exible endoscopes published in 2002 [26] was the �rst evidence-based guideline for scope reprocessing and was based on a profound literature review comparing international guidelines for endoscope reprocessing prepared by a working group [27] and the results of a large study on hygiene in endoscopy performed in the area of Munich in 2000 [45].One and a half year later, several gastroenterological, microbiological, and hygienic societies in the United States published a similar evidence-based multisociety guideline [68].
One of the most important differences of these two guidelines is the answer to the problem whether microbiological surveillance cultures of reprocessed endoscopes should be recommended or not.It is absolutely correct, that no data exist demonstrating that regular microbiological surveillance cultures of endoscope reprocessing decrease the risk for infection transmission during endoscopy.us, the interpretation of the multisociety guideline [68] and other authors [69] that the value of regular microbiological surveillance cultures of reprocessing is an "unresolved issue" is principally correct.In accordance with this statement microbiological controls of endoscope reprocessing quality are not performed in the United States [49].In recent years, however, usefulness of surveillance cultures is mentioned more oen in the United States [37,40,44,69].
In contrast to the multisociety guideline [68] and considering the poor compliance with hygienic recommendations and variability in implementation [70,71] the German guideline recommended regular surveillance cultures of endoscope reprocessing [26].e recent European guideline for endoscope reprocessing, published in 2007 [24], also recommended microbiological controls of endoscope reprocessing quality.
e experience of smaller studies from Germany [72,73], the results of the former nationwide study [74], and the present update reinforce the hypothesis underlying the German RKI guideline that measurement and feedback of reprocessing quality will help to implement reprocessing procedures in accordance with present guidelines.In other words and from a quality management point of view: feedback of reprocessing outcome quality enhances the process quality and compliance with reprocessing guidelines which in turn vice versa affects outcome quality.

Process Quality or Outcome Quality of Reprocessing Complex Medical
Devices? e question whether we should focus on process quality or outcome quality of reprocessing �exible endoscopes is still a matter of debate.Although standardized reprocessing is possible in special washer-disinfectors using automatic single channel cleaning with enzymatic cleaners and programmed disinfection with glutaraldehyde at higher temperature [75], only microbiological controls of reprocessing quality are able to detect equipment defects of �exible endoscopes (such as perforation of internal channels) which affect the outcome of the reprocessing process and which would be without surveillance cultures only recognized in working ups of nosocomial infection outbreaks [76].

Summary and Conclusions
. e quality assurance agreement for screening colonoscopy in Germany in 2002 [58] improved colonoscope reprocessing quality signi�cantly.Aer implementation a quality control system for colonoscope reprocessing in ambulance offices the rate of positive surveillance cultures from reprocessed endoscopes dropped from 50% of failed tests observed before the introduction in 2000-2001 to 17,7% in 2003 and to 10,1% in 2004 and reached a plateau of under 5% in 2008.us, the ethical principle that a screening procedure, such as a screening colonoscopy should be without risk of infection transmission, seems nearly ful�lled.e question whether monitoring endoscope reprocessing by regular microbiological surveillance cultures will reduce the infection transmission during endoscopy cannot yet be answered by the present study.However, our data strongly favor regular surveillance cultures as an important catalyzer for implementation and adherence to national endoscope reprocessing guidelines and for improvement of reprocessing quality.So in order to reduce infection transmission by endoscopy, we recommend that other countries considering implementation of screening colonoscopy should also consider implementation of outcome monitoring of endoscope reprocessing.
T 1: Time trends of key microbiological performance data in endoscopy in Germany from 2003 to 2008.Stated are the raw numbers and percentages as indicated.In brackets the numbers of the reporting associations of statutory health insurance physicians are given.
signi�cant changes between 2003 and 2004 and between 2004 and 2005.e overall time trend of repeated failures (  ) is signi�cant (  ).Consecutive pairwise comparisons revealed only between 2004 and 2005 a signi�cant result.