From the Journals

Postcolonoscopy cancer rates persist despite quality protocols

Publish date: January 29, 2018
By
Heidi Splete
MDedge News

View on the News

What is holding back ‘quality’?

Postcolonoscopy colorectal cancers (PCCRCs) are those cancers that occur between 6 and 36 months after a complete colonoscopy. For cancers diagnosed less than 6 months from exam, it is presumed that the exam itself was diagnostic. Most of these cancers grow from cancers or near cancers missed or incompletely resected during the baseline colonoscopy. Clinical researchers have published extensively about reasons for missed lesions and we know that age, female sex, and proximal location of cancers increase rates of PCCRC. GI societies worldwide have developed training initiatives, performance metrics (adenoma detection rate or ADR, withdrawal time, and prep quality documentation), and postcolonoscopy guidelines, all intended to mitigate risk of PCCRCs. It would be nice to know whether such efforts have made a difference.

Murthy and colleagues studied PCCRC rates in Ottawa, Canada during three different time periods to determine whether quality and educational efforts impacted PCCRC rates. More than 99% of this population has health care covered under a single public payer system where all encounters are carefully tracked. Using population-level health data derived from over 1 million people (screen eligible people, 50-74 years of age with low to moderate CRC risk) they identified cancers diagnosed within 36 months of a colonoscopy and compared three 5-year periods (1996-2001, 2001-2006, and 2006-2010).

Their method of calculating PCCRC rates essentially says, “If I am destined to develop CRC in the next 3 years, what is my chance of a false-negative colonoscopy?” There are five published methods of calculating PCCRC rates (summarized in Gut 2015;64:1248-56) and each method uses different inclusion criteria and denominators. The question posed above yields “rates” that would terrify patients (4%-10%) without a detailed explanation (it took me about an hour of focused attention to finally understand this methodology). In essence, if we could, a priori, identify and examine only patients who have a prevalent cancer or near cancer, how close can we come to 100% accuracy with a colonoscopy? Turns out, that rate is somewhere between 90% and 96% and really hasn’t changed over time. Thus, these studies speak to the impact of our efforts around colonoscopy quality.

Dr. John I. Allen, professor of medicine, department of gastroenterology and hepatology, University of Michigan, Ann Arbor, and Editor in Chief of GI & Hepatology News

Dr. John I. Allen

The discouraging conclusion from Murthy’s analysis is that despite substantial efforts, false-negative colonoscopy rates have remained around 8% (in Ottawa) since 1996. Of note, this contrasts with studies out of England, where a national, focused quality improvement effort has been ongoing for over a decade and has made a dent (although slight) in PCCRC rates. This is a provocative study that deserves your attention.

John I. Allen MD, MBA, AGAF, professor of medicine, department of gastroenterology and hepatology, University of Michigan, Ann Arbor, and Editor in Chief of GI & Hepatology News.


 

FROM GASTROINTESTINAL ENDOSCOPY

The number of colorectal cancers diagnosed after a colonoscopy remained consistent at approximately 8% over a 15-year period despite the introduction of quality improvement measures, according to data from a population-based cohort study of more than 1 million individuals in Canada.

“It is believed that the majority of PCCRCs [postcolonoscopy colorectal cancers] arise due to cancers or near cancers that were either missed or incompletely treated during colonoscopy,” wrote Sanjay K. Murthy, MD, of the University of Ottawa, and colleagues.

Established quality improvement measures included adenoma detection rate, cecal intubation rate, colonoscopy withdrawal time, and endoscopy training standards, but how well the measures have been implemented remains uncertain, the researchers said. In a study published in Gastrointestinal Endoscopy (2018 Jan 6. doi: 10.1016/j.gie.2017.12.027), the researchers assessed data from 1,093,658 eligible adults aged 50-74 years over a 15-year period. The time period was divided into three sections: July 1, 1996, to June 30, 2001; July 1, 2001, to June 30, 2006; and July 1, 2006, to Dec. 31, 2010.

Overall, the number of colonoscopy procedures increased during the study period, from 305 per 10,000 people in 1996-1997 to 870 per 10,000 people in 2010-2011, and the percentage of individuals who underwent complete colonoscopies increased from 67% in the 1996-2001 period to 88% in the 2006-2010 period. “There was a considerable increase in the proportion of colonoscopies performed in younger age groups and community clinics in successive study periods,” the researchers noted.

Comparing the 2006-2010 and 1996-2001 time periods yielded adjusted odds of PCCRC, distal PCCRC, and proximal PCCRC of 1.14, 1.11, and 1.14, respectively; the trends were not affected by endoscopist specialty or institutional setting.

“Our findings are concerning for lack of improvement in colonoscopy practice quality in Ontario, particularly in the wake of greater emphasis having been placed on colonoscopy quality metrics during the study period,” the researchers said. The findings contrast with the decline in PCCRC rates in the United Kingdom reported in a previous study of a similar time period, they noted.

The study findings were limited by several factors, including possible patient and outcome misclassification, an unvalidated definition for PCCRC, and unmeasured confounders such as changes in practice or changes in the definition of PCCRC. Although more research is needed in other jurisdictions to confirm, the results “call for increased population-based practice audit as well as endoscopy educational programs and certification requirements.”

The study was supported by a research grant to Dr. Murthy from the University of Ottawa. The researchers had no financial conflicts to disclose.

Next Article:

Higher BMI linked to problems for IBD patients