Quality of colonoscopy: the real need is for universal quality measurement and common standards for certification and credentialing

The study by Leyden and colleagues [1] has provided us with more evidence that the quality of colonoscopy is not carefully controlled. This is true throughout the world. Although common training standards would likely improve colonoscopy quality, I would argue that the most important control over quality comes at the stages of certification and/or credentialing. Those two steps determine who is allowed to perform colonoscopies. In the United States, certification by the American Board of Surgery or the gastroenterology subspecialty board of the American Board of Internal Medicine determines which trainees have successfully completed their training. Credentialing, which occurs at the local level of hospitals and ambulatory surgery centers, represents the final decision as to which providers are allowed to perform colonoscopy. Perhaps more control at the training level would work in Europe because certification and credentialing rules are different from those in the United States. However, common training standards would not address the problem of the older clinician who is no longer performing quality work. Competence should not be considered to be a static entity during training or throughout a career [2]. Continuous measurement of quality outcomes for all endoscopists, regardless of specialty or training, and the use of these quality measures for granting of endoscopic privileges is the only real solution.

Most of the gastroenterology literature on colonoscopy training shows that performance of 250 – 500 supervised colonoscopies is needed to reach competence [3] [4]. Despite this data, the American Society for Gastrointestinal Endoscopy guideline on credentialing suggests that at least 140 be performed prior to assessment of competence [5]. Unfortunately, “competence assessment” has traditionally been vague, usually simply a statement of adequacy by the head of the training program. Recently, more rigorous training measurements have been suggested although not widely adopted, partly because of their complexity [6] [7]. The evidence supplied by the study of Leyden et al. helps to confirm that threshold numbers are insufficient measures of training. The surgical registrars had colonoscopy volumes similar to those of the gastroenterology registrars in the 2 years prior to study entry. Although this was not clearly delineated in Leyden et al.’s paper, by my calculations the surgical registrars had already performed 500 – 900 colonoscopies per trainee (250 – 300 per year for 2 – 3 years) vs. 300 – 900 for the gastroenterology registrars (150 – 300 per year, with six of the seven having already trained for 3 years and one having had 2 years’ training). Therefore both types of trainees had performed more than enough procedures to be well trained. This raises the question of why the surgical registrars did not perform as well in either cecal intubation or polyp detection as the gastroenterology registrars. One explanation could be a lack of emphasis for the surgical registrars on cecal intubation and careful inspection on withdrawal. The faster withdrawal rates of the surgical registrars support that contention, as such rates have been shown to be a crude marker for lower adenoma detection rates (ADRs). Another explanation could be a difference in the level of supervision initially available to the two types of registrar. It is unclear whether different numbers of “continuous direct supervision” examinations are required for each group before they are allowed to independently perform colonoscopy. Indeed, what is that number in Ireland?

Since the performance difference between gastroenterology and surgical registrars cannot be explained by insufficient procedure volume (which is a problem for surgical residents in the United States), the lower performance of the surgical registrars must be due to a lack of emphasis on good technique and/or poor initial training. In the United States, all colonoscopies by gastroenterology fellows or surgical residents have “continuous direct supervision,” since this is required by the major payers (with the possible exception of training at Veterans Administration hospitals). However, in the US, the certifying surgical board only requires 50 colonoscopies during training whereas most gastroenterology fellows perform more than 500 colonoscopies during their training. It is unclear how many American surgeons go on to seek privileges in colonoscopy with this small amount of training, but undoubtedly some of them do. Given the pressure on credentialing boards to approve all procedural privileges since procedures are financially beneficial to the hospital or ambulatory surgery center, it is likely that some of these undertrained American surgeons are credentialed and perform colonoscopy.

It is not clear from the study by Leyden et al. whether their ADRs are low compared with what is reported in the literature because they chose the wrong denominator. They had an ADR in patients over the age of 50 of only 12 % (14 % for gastroenterology registrars and 9 % for surgical registrars). The published recommended rate is greater than 25 % in men and greater than 15 % in women. The accepted definition of ADR is the percent of patients found to have one (or more) adenomas amongst those who are undergoing the procedure for colorectal cancer screening or surveillance indications, which is a polyp-enriched population. That would exclude patients undergoing colonoscopy for other indications such as inflammatory bowel disease, abdominal pain, or diarrhea. Therefore the denominator for ADR should be a lower number than it is for cecal intubation (which should include all colonoscopies that are not aborted due to poor preparation, patient intolerance, obstructing lesion, or severe colitis). If the Leyden et al. study did use the appropriate denominator, I would argue that neither group of registrars hit adequate quality benchmarks. Of course, real-life situations as reported in this retrospective review of colonoscopies may not match the high quality outcomes seen in prospective trials. An example of that is the population-based study from Canada that showed a cecal intubation rate of only 87 % [8].

In conclusion, I would like to congratulate Leyden and co-authors for their focus on colonoscopy training. Too little data are available on this important topic. In the study of Leyden and colleagues, the number of procedures previously performed did not predict quality outcomes. The reasons for the differences in performance between the surgical and gastroenterology registrars need further study. That would allow us to develop common international training standards, a laudable goal of the authors. However, I continue to argue that the real “control valve” on endoscopic quality is at the certification and credentialing steps. The international community owes it to our patients to ensure that only proficient endoscopists, as measured by quality outcomes, are allowed colonoscopy privileges.

References

• 1 Leyden J E, Doherty G A, Hanley A et al. Quality of colonoscopy performance among gastroenterology and surgical trainees: a need for common training standards for all trainees?.  Endoscopy. 2012;  43 935-940
• 2 Vargo J J. North of 100 and south of 500: where does the “sweet spot“ of colonoscopic competence lie?.  Gastrointest Endosc. 2010;  71 325-326
• 3 Spier B J, Benson M, Pfau P R et al. Colonoscopy training in gastroenterology fellowships: determining competence.  Gastrointest Endosc. 2010;  71 319-324
• 4 Lee S H, Il–Kwun C, Kim S J et al. An adequate level of training for technical competence in screening and diagnostic colonoscopy: a prospective multicenter evaluation of the learning curve.  Gastrointest Endosc. 2008;  67 683-689
• 5 Eisen G M, Baron T H, Dominitz J A et al. ASGE Guideline: Methods of granting hospital privileges to perform gastrointestinal endoscopy.  Gastrointest Endosc. 2008;  55 780-783 [reapproved 11/08]
• 6 Sedlack R E. The Mayo colonoscopy skills assessment tool: validation of a unique instrument to assess colonoscopy skills in trainees.  Gastrointest Endosc. 2010;  72 1125-1133
• 7 Gupta S, Anderson J, Bhandari P et al. Development and validation of a novel method for assessing competency in polypectomy: direct observation of polypectomy skills.  Gastrointest Endosc. 2011;  73 1232-1239
• 8 Shah H A, Paszat L F, Saskin R et al. Factors associated with incomplete colonoscopy: a population–based study.  Gastroenterology. 2007;  132 2297-2303