Colonoscopy remains an effective method of colorectal cancer screening, but one important limitation is the wide variability in adenoma detection rates (ARDs). Higher rates are associated with improved outcomes and better cancer prevention, and lower rates with a greater risk of interval cancer.
Although many groups have tried to improve adenoma detection among individual endoscopists, their efforts have been largely unsuccessful, says Michael B. Wallace, M.D., of Mayo Clinic in Florida."Mandating longer procedural and withdrawal times hasn't helped. Making ADRs publicly available at department meetings and even offering salary incentives didn't work. Some groups tried monitoring people or letting them know they were being videotaped, which had only a small benefit."
Dr. Wallace thought an educational approach — having endoscopists complete a relatively simple but intensive Endoscopic Quality Improvement Program (EQUIP) — would be more effective at improving ADRs. So he and his colleagues designed a study involving 15 endoscopists who routinely perform screening, surveillance and diagnostic colonoscopies in Mayo Clinic's Florida ambulatory practice.
The study had three phases: baseline, randomization and training, and post-training. In the first phase, lasting approximately three months, ADR — defined as the proportion of patients in which at least one adenoma was detected — was measured in all endoscopists. A standard procedure sheet was completed for each colonoscopy, including patient information, procedure start, withdrawal start and end times, and adequacy of bowel prep. Polyps were categorized according to size, location and predicted pathology.
At the end of the baseline phase, the endoscopists received feedback on their ADR and withdrawal times. Half were then randomly assigned to undergo EQUIP training, which consisted of two one- to two-hour didactic sessions that reviewed techniques to increase ADR and especially flat adenoma detection. A narrow band imaging module also was included.
After completion of the training, ADR was again measured for approximately three months and data collected for each colonoscopy. In all, 1,200 procedures were completed in each of the two study phases. During this period, intervention endoscopists received private monthly feedback on their ADRs.
"There was a very active monitoring process; the endoscopists were told they were being monitored and whether they were improving or not," Dr. Wallace explains. "More important was that we created a culture of quality, and everyone was very focused on that."
Overall, the training program led to significantly improved adenoma and polyp detection rates. At baseline, the ADR for all endoscopists was 36 percent. In the post-training phase, the ADR for the EQUIP-trained group increased to 47 percent but remained virtually unchanged for untrained endoscopists. Improvements occurred for all indications, including increases from 35 to 44 percent for screening and 47 to 59 percent for surveillance. The polyp detection rate also increased significantly among the trained endoscopists — from 51 to 65 percent — but from just 53 to 54 percent among the untrained doctors.
"Every doctor was already above the national benchmark for quality, which is 25 percent for male patients and 15 percent for female patients. Women generally have fewer polyps so the benchmark is lower. The highest performing endoscopists in our group are more than 60 to 70 percent, so it wasn't a matter of taking poorly performing endoscopists and making them good; it was taking people from good to great," Dr. Wallace points out.
"Why the training succeeded when so many other ADR improvement efforts have failed is probably multifactorial," he explains. "Rather than focusing on withdrawal time, we emphasized improving inspection technique in the training. At the same time, the current average withdrawal time is seven minutes, but in our practice it averages 13 minutes and some doctors average more than 20 minutes. So in our view, withdrawal time remains an indirect measure of quality and surrogate marker for more careful looking."
More careful looking — and knowing what to look for — seem to be key outcomes of the training. "Detection of nonpolypoid, or flat, lesions is especially challenging and research shows that these adenomas are frequently missed. If you know you're missing something, then you start to look for it. So we provided plenty of pictures and videos of what difficult polyps look like," Dr. Wallace says.
A follow-up study lasting five months was initiated a few weeks after the end of phase II. It included 1,200 patients undergoing routine colonoscopy and the same two groups of trained and untrained endoscopists. Results were presented at Digestive Disease Week 2011.
Given EQUIP's success, Dr. Wallace was awarded a grant to perform the same intervention nationally and is still looking for participating sites. "There is a broad consensus that ADR is a good way of measuring quality," he says. "It's a surrogate marker that is very tightly linked with prevention of colon cancer, and we know that patients of doctors with higher adenoma detection have a lower cancer risk within the next five to 10 years."
He adds, "There is a huge emphasis on expensive technology, but focusing on inexpensive interventions like technique and education change has a greater impact than the most advanced technologies. Doctors inherently want to improve quality, and if there is evidence that we can improve quality, then we are quite willing to do it."
Coe SG, et al. An endoscopic quality improvement program improves detection of colorectal adenomas. The American Journal of Gastroenterology. 2013;108:219.