Colorectal cancer risk

The study by Ellis et al1 in the December 2005 issue of the BJGP, raises several issues ranging from failure to assess the importance of other risk factors associated with colorectal cancer to misleading statistics and incomplete investigations. When evaluating rectal bleeding, it is essential to look at other risk factors such as unexplained anaemia, family history of colorectal cancer and previous history of removal of adenomatous polyps. They mentioned no laboratory investigations.

Colonoscopy should be the diagnostic procedure of choice and gold standard in evaluating the possibility of cancer. There was no mention why 37 patients had barium enema.

In Table 1, page 952, I read in the first row sensitivity 100. This means that all patients presenting with rectal bleeding and change in bowel habit will have colorectal cancer! Page 954 stated that all patients with cancer had an associated change in bowel habit. This is their explanation of 100% sensitivity. Their analysis included 219 patients having flexible sigmoidoscopy and 47 patients filling in a questionnaire. The assessment of a questionnaire will neither exclude nor confirm colorectal pathology. Despite 53 patients declining participation, Tables 1–3 included them in quantifying the ratio of pathology. Their presentation is misleading.

Not every individual with rectal bleeding needs a colonoscopy. But if their aim is to evaluate the diagnostic power of symptoms in the assessment of cancer, then the gold standard colonoscopy should be used. The authors gave no details of duration of various symptoms and no explanation of selection criteria for performing barium enema or colonoscopy.

I read in the second column of page 953 in the last paragraph: ‘As it has been shown that flexible sigmoidoscopy’. Their statement is inaccurate because the flexible sigmoidoscopy to 60 cm would detect the majority of colorectal cancer. It will miss 20–30% of significant proximal neoplasms2 leading to missed diagnosis, false reassurance, progression of the disease, suffering and death. I am concerned about the method of recruiting patients to the study offering three options: flexible sigmoidoscopy; if not accepted postal questionnaire; or neither. A consultation should be offered without the obligation to participate in the study.

The authors stated that physical examination was carried out at the time of flexible sigmoidoscopy. This means that 47 patients filling the questionnaire and 53 declining flexible sigmoidoscopy and a questionnaire had no clinical examination. If patients decline a procedure, you still have the duty to offer them alternative options such as clinical examination, laboratory investigations and follow up after a reasonable time. This paper adds nothing new to the various guidelines on criteria for high risk of bowel cancer.

• © British Journal of General Practice, 2006.