One Thousand Colonoscopies:
What Does This Procedure Really Achieve?

We are all familiar with the multicentre trials published in our medical journals, but auditing our own practice performance can be very useful and informative. Dr. Robert Sers, a surgeon at St. Martha's Regional Hospital did just that with regard to his colonoscopies. The following is a report that would likely be representative of the procedure in most small regional hospitals. - Editor

Introduction:
The purpose of this study was to answer the question posed in the subtitle. What are we actually achieving with this test? Certainly there are a lot of colonoscopies being done, and given our present demographic trend, the demand for this procedure will increase greatly in the next few years. Patient selection is therefore critical, as is the need to demonstrate what if anything useful is gained by doing this examination. With this in mind it was decided to keep a list of colonoscopies, the results, and the indications. These were all done by the author, and the indications were strictly observed. Because only one Surgeon is involved, the criteria for doing them, and the assessment of the results is consistent. The intent was to evaluate several things, including, the results of screening for family history, and surveillance in cases of inflammatory bowel disease. Also, it was planned to see if any thing was actually discovered when "follow up scopes" were done.

Method:
The cases are consecutive cases from October 1997, to early March 2001. Almost all were done with the videoendoscope, and most have pictures as well. Names and unit numbers are recorded to allow these charts to be identified for further study if needed.

Where the scope was done for screening or follow up on polyps or cancer, the results were listed as either normal, polyp, or cancer. Because the intent was to screen for cancer, other diagnoses were not relevant. When there was an indication such as bleeding or some other clinical symptom, then all diagnoses were included.

Results:
The list of indications is shown in Fig. 1. Screening for bleeding and family history, and follow up on polyps are the major indications, with smaller numbers for inflammatory bowel disease and follow up on cancer. Abdominal pain and "other" make up only 14% of the total. The scopes done in the category "other" were done for reasons such as barium enema findings requiring further confirmation, diarrhea, suspected strictures, and on one occasion because of a suspected solitary rectal ulcer.

Screening in the IBD category, was done both to evaluate the extent and activity of the disease, and to look for associated malignancy. No cancers, or definite signs of dysplasia were found.

Using these criteria for choosing patients for colonoscopy, the overall results showed a normal rate of 47%, with 53% showing some form of pathology (Fig. 2). Many of the normals are in the follow up for polyps or cancers.

In the subgroup of patients with rectal bleeding or anemia, 32% have polyps, while 12.4% have carcinomas already. When people over 70 are selected from this subgroup, the rate of carcinoma rises to 23% with 28% polyps and only 16% normal. In Fig. 3, one can see a comparison of the rates of cancer and polyp detection in patients in the under 70, and 70 or over groups. As expected the total percentage of patients with polyps or cancers goes up with age, and it is noted that this increase is largely in the number of cancers. Incidentally, the average age of people with polyps was 62.2, and for those with cancer it was 71.0 years old.

Figure 3

There is a slight increase in the percentage of cancers and polyps in males with anemia or bleeding. (Figure 4).

A total of 84 scopes were listed as limited, that is the ileo-cecal valve was not reached. In 21 cases this was planned leaving 63 (6.3%) scopes that could not be completed as planned. Pain, loop formation and stricture were the most common reasons. In most cases where pain limited the study, the patient had significant diverticulosis.

When patients have follow up scopes for cancer, 38.75% have polyps and 1.25% are found to have new cancers, When the follow up is done for polyps, 47% have polyps. So far there have been no cancers diagnosed in the latter group.

Figure 4

The use of colonoscopy for abdominal pain should be restricted to cases where the diagnosis is difficult to achieve, and where there is genuine reason to suspect serious disease. This remains a somewhat vague indication but nevertheless yielded a polyp detection rate of 27%. Crohn's Disease, diverticulosis and cancer were also found in this group. The normal rate was the same as for the whole group 47%. There is clearly a role for this test in certain cases of difficult to diagnose pain.

Discussion:
As the population ages the demand for colonoscopy will inevitably increase, and this will put a lot of pressure on the existing capacity to do this examination. For this reason the criteria used in selecting patients for colonoscopy should be clear and consistent. When the criteria (indications) that were mentioned above are used, there is a high rate of pathology found. In fact, overall only 47% of scopes were normal. Many of these were found among the "follow up" cases where one would want to find them.

Certain subgroups are particularly important, especially those with risk factors for bowel cancer such as a family history of cancer, rectal bleeding or anemia. The rates of polyp and cancer detection are high in these groups, with the total of polyp and cancers patients adding up to 46 - 48% consistently. With advancing age this number increases almost solely due to increasing numbers of cancers. In fact when the patient is over 70, there is a 23% risk of cancer, and a 28% risk of polyp detection. The rates in the younger age group show around 12% cancers and 32% polyps.

Screening for family history (with no further specification such as number and degree of relatives) yields a polyp detection rate of 32%, and in this group one cancer.

Surveillance for inflammatory bowel disease is done both for assessment of the disease status, and for detection of dysplasia, a precursor of IBD associated cancer. No cancers were detected, but the assessment of the disease represents a sufficient reason to scope.

In the future, data could be added to assess the location of the polyps and cancers, as well as the type of polyp, and finally there could be more family data in the family history category. It would be possible to gain some idea as to the risk of developing cancer in polyp patients. This is a high risk group, and if the risk is lowered to or below that of the general population then the value of screening could be even more clearly demonstrated.

In summary colonoscopy is a valuable tool, and when the patients are selected carefully, there is a high rate of detection of significant illness.

- Robert Sers, MD FRCSC

Thanks to Dr. J. Kennedy MD FRCSC for reviewing this article, and for her comments.