|Home | About | Journals | Submit | Contact Us | Français|
Primary care clinicians initiate and oversee colorectal screening for their patients, but colonoscopy, a central component of screening programs, is usually performed by consultants. The accuracy and safety of colonoscopy varies among endoscopists, even those with mainstream training and certification. Therefore, it is a primary care responsibility to choose the best available colonoscopy services. A working group of the National Colorectal Cancer Roundtable identified a set of indicators that primary care clinicians can use to assess the quality of colonoscopy services. Quality measures are of actual performance, not training, specialty, or experience alone. The main elements of quality are a complete report, technical competence, and a safe setting for the procedure. We provide explicit criteria that primary care physicians can use when choosing a colonoscopist. Information on quality indicators will be increasingly available with quality improvement efforts within the colonoscopy community and growth in the use of electronic medical records.
Screening for colorectal cancer can prevent cancer incidence and death and is recommended in clinical practice guidelines1,2. Primary care clinicians initiate and oversee colorectal cancer screening programs for their patients3. However, colonoscopy, which is usually not performed by the primary care clinicians themselves, is a central part of such programs—for screening, for diagnostic evaluation of positive test results regardless of the initial screening test, and for surveillance of patients at increased risk. Most colonoscopies in the U.S. are performed by gastroenterologists and colorectal surgeons, although other specialists also do this procedure4. Thus, primary care clinicians and referral endoscopists share responsibility for successful colorectal cancer screening.
The accuracy and safety of colonoscopy vary from clinician to clinician, even among those with extensive experience5,6. In a study of 12 full-time, board-certified gastroenterologists with formal training, each of whom had performed at least 3,000 colonoscopies, adenoma detection rates ranged from 9.4% to 32.7%5. Withdrawal times ranged from 3.1 to 16.8 minutes in patients in whom no polyps were found and endoscopists with longer withdrawal times had higher adenoma detection rates. Rates of cancer detection and perforation also vary among examiners7.
Given this variation in performance of colonoscopy, primary care clinicians should be in a position to identify endoscopists who meet high standards for quality. We have identified a set of indicators that primary care clinicians can use to assess quality of colonoscopy procedures. This work is part of a larger effort to improve the quality of colonoscopy. In 2002, the U.S. Multi-Society Task Force on Colorectal Cancer published recommendations for indicators that should be measured by colonoscopy practices as part of ongoing quality improvement8. The Quality Assurance Task Group of the National Colorectal Cancer Roundtable then developed a standardized reporting system (CO-RADS) that specifies the elements that should be included in colonoscopy reports and a standard method for reporting them8. That documentation system should also serve to improve communication with referring clinicians and patients by providing reports that use standard terms and include specific, evidence-based recommendations for follow-up.
These recommendations were developed by members of the National Colorectal Cancer Roundtable, a coalition of nearly 60 public, private, and voluntary organizations who work together, with support from the American Cancer Society and the Centers for Disease Control and Prevention, to advance colorectal cancer control through screening9. The Quality Assurance Task Group, a working group of the Roundtable, identified the importance of colonoscopy quality from the primary care perspective at a meeting on November 5, 2007. A panel of the Task Group, including three primary care physicians (two family physicians and one general internist), one preventive medicine specialist, six gastroenterologists, two epidemiologists, and one nurse volunteered to prepare this statement. This panel met by conference call and exchanged views by email and telephone. Drafts of the statement were revised and circulated until all members of the Panel approved the recommendations.
We based our recommendations on existing research evidence and clinical practice guidelines. While there is not strong, direct research evidence that these recommendations change outcomes, all reflect the kinds of performance measures—such as completeness of examination, adenoma detection rates, and safety—that matter to patients and have been shown to be achievable. They also reflect the information that clinicians need to have to make evidence-based decisions about screening in individual patients. For example, recommendations for surveillance intervals depend on the number, size, and histology of adenomas found on the preceding examination10.
Recommendations for the quality of colonoscopy services are summarized in Table 1. The following describes the basis for these recommendations.
1. Elements of the Colonoscopy Report We listed six basic descriptors of what was done and what was found during the colonoscopy based on earlier work by Task Group8. Each of these descriptors is necessary to interpret the clinical significance of findings and plans for follow-up.
Physicians in practice say they often choose shorter surveillance intervals11 than recommended in clinical practice guidelines. Because some risks and substantial costs (measured both in financial and human terms) are at stake, and the clinical benefit of short-interval colonoscopy surveillance after initial polypectomy is low for most patients12, the surveillance interval is an important decision for everyone involved with the patient’s careFamily history information gathered at the time of screening as well as some colonoscopy findings (such as a malignancy or advanced adenomas at a young age) may suggest that other family members are at increased risk and should have earlier than usual screening with colonoscopy. In the case of Lynch Syndrome, they are at increased risk for other cancers as well. The endoscopy report should include recommendations for colorectal cancer screening in family members when appropriate. The endoscopist and the primary care clinician have a collective responsibility for encouraging patients to notify family members if they are at increased risk, and for recommending that these family members talk with their own primary care clinician about colorectal cancer screening.
2. Cecal Intubation Rate Clinically-important adenomas and cancers occur throughout the colon and will be missed to the extent that the entire colon is not examined. Reports of consecutive screening colonoscopies have established that cecal intubation rates of over 90% are achievable13–18, especially in patients without clinical reasons for incomplete colonoscopies such as severe colitis, poor preparation, severe diverticulosis, vital sign instability during the procedure17. Several expert groups have set a quality target of 90% or higher for cecal intubation rate7,19. Screening guidelines recommend that if the cecum cannot be reached other imaging procedures (computed tomographic colonography or double contrast barium enema) should be used to complete the examination1.We concluded that an average cecal intubation rate of at least 90% was achievable after excluding examinations that were terminated for clinical reasons and those for which full colonoscopy was not the original intent. Rates are lower in some settings6,20,21, but there is evidence that rates can be improved by quality improvement programs21. We recommend that all endoscopists should aim to meet this target, regardless of specialty, training, or experience.
3. Adenoma Detection Rate The prevalence of adenomas at age 50 years is estimated to be 15% in women and 25% in men, increases with age in both sexes7, and the majority of adenomas are detected by colonoscopy22,23. It is not feasible to measure the proportion of adenomas found for individual colonoscopists against research standards, (a second colonoscopy and computed tomographic colonography) so we chose adenoma detection rates as a crude metric for the proportion of adenomas found at colonoscopy. Adenoma detection rate during screening colonoscopy has been shown to be inversely related to the risk of interval cancers24. We confined this quality measure to first colonoscopies because prior polypectomies can change the prevalence of adenomas, making polyp prevalence lower and detection rates more difficult to interpret.
4. Safe Setting Colonoscopy can cause clinically-important complications such as bleeding, perforation, and cardiovascular events during bowel preparation or endoscopy. The procedure can also spread infection if equipment is improperly cleaned and disinfected. Although the great majority of colonoscopies occur without incident, complication rate is an important aspect of quality. However, we chose not to include complication rates in the quality measures because events occur too infrequently to allow stable estimates of rates for individual colonoscopists, unless he or she has performed an unusually high volume of procedures.Instead, the panel recommended that safety be assessed by a surrogate measure, characteristics of the setting in which procedures are done. Among these are adequate cleaning and disinfection of equipment, well-maintained equipment, well-trained endoscopist and staff, and the ability to react to emergencies that might arise during the procedure25.While some states and professional societies have guidelines for safe settings and requirements for accreditation, primary care clinicians cannot always rely on external review to decide whether the colonoscopy setting is safe. At this time, there is no single, overall mechanism for credentialing that applies to all specialties that do colonoscopy and in all settings in which it is done.
We have identified a set of quality indicators that primary care clinicians can use when deciding where to send their patients for colonoscopy services. The need for a more rigorous approach to choosing among endoscopists is highlighted by reports of substantial variation in the quality of colonoscopy, even among clinicians generally considered to have appropriate training, experience, and credentials. The elements of quality are based on the information needed to follow screening and surveillance guidelines. The targets we set have been shown to be achievable in community settings.
We have recommended that primary care clinicians rely on the actual performance of colonoscopy, not just the background and specialty of those who do the procedures. Specialty societies have set standards for performance of colonoscopy but they have not been universally adopted or enforced. In one study of gastrointestinal endoscopy centers in the US, 60% required no minimum number of procedures before granting privileges, and two-thirds required no minimum number of procedures per year26. Also, training, experience, and specialty do not in themselves assure high quality colonoscopy5.
Is it feasible for primary care clinicians to implement these recommendations? We would argue that primary care clinicians have always considered it a core responsibility to choose the best available consultants, and now, at least for colonoscopy, they are in a better position to meet that responsibility. They are also better able to care for their patients if the report of the procedure includes the information they need to explain plans for colorectal cancer screening and surveillance to their patients and judge the appropriateness of colonoscopy intervals for themselves. In the absence of performance information on individual endoscopists, it seems likely that consultants are chosen based on other considerations, such as availability, reputation, and congeniality. New models for how primary care can be organized and financed, which are now being tested as part of the patient-centered medical home movement27, might foster, among many other improvements in care, ways to identify consultants by performance.
Will individual endoscopists be able and willing to provide information on their performance? Standardized reporting is being promoted within the endoscopy community as part of a larger set of quality improvement activities. We encourage clinicians to refer patients to endoscopists who use a standardized reporting mechanism such as the CO-RADS lexicon8, and to communicate with endoscopists that they value receiving reports that follow this standard. If electronic medical records become more widely implemented, information on quality of colonoscopies may be easier to monitor and share.
There has been a lively debate in the literature about the kind of training needed to perform high quality colonoscopy28. The published literature is not yet robust enough to answer this question with confidence. Whatever the answer, we thought that the performance of each endoscopist should stand on its own merits, as reflected in measurable quality indicators such as the quality of the colonoscopy report and cecal intubation and polyp detection rates. Choosing endoscopy services according to actual performance is becoming possible for the first time, as a result of efforts in the colonoscopy community to set measurable quality standards and the growing use of electronic medical records. With a better understanding of quality standards for colonoscopy, primary care clinicians can be better advocates for their patients.
This work was supported by the Division of Cancer Prevention and Control at the Centers for Disease Control and Prevention, the American Cancer Society and the VA Puget Sound Health Care System, Department of Veterans Affairs. Dr. Dominitz was supported by an American Society for Gastrointestinal Endoscopy Career Development Award.
The views expressed in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention or the Department of Veterans Affairs.
Conflict of Interest None of the authors has a financial tie to entities that would stand to gain financially from implementation of the recommendations in this statement.