Introduction

Beginning at age 50, the American Cancer Society recommends that both men and women at average risk should have one of the following screening measures to detect possible colorectal (CR) cancer:

• Fecal occult blood test (FOBT), every year
• Flexible sigmoidoscopy, every 5 years
• Both an annual fecal occult blood test plus flexible sigmoidoscopy every 5 years (of the first three options, ACS prefers this one)
• Double-contrast barium enema, every 5 years
• Colonoscopy, every 10 years

The case

Andy B, now 56 years old, has been advised by his physician to have a fecal occult blood test (FOBT) as a screening test for colorectal cancer. Andy is a healthy, non-smoking, normal weight lawyer with a healthy lifestyle, who exercises regularly and follows a mostly Mediterranean diet. His uncle died at 75 because of colorectal cancer; this went undetected for years before symptoms of abdominal colic brought him to the emergency unit of the local hospital 2 weeks before the final event. Andy's first-degree relatives (father, mother and sisters) did not have any cancer history.

Andy agreed to take the FOBT. The result came back positive. Andy and his wife went through a terrible period of despair, with the certainty of colon cancer looming.

The question

In your opinion, does Andy have colon cancer? 'Yes', or 'undecided'?

If you answered yes, you violated Franklin's law (...nothing can be said to be certain but death and taxes); you were probably swayed by the family history and the positive FOBT. Obviously you need to be more careful before delivering such bad news. More information is required to assess the risk that Andy has a colorectal cancer.

As suggested in the previous article in this series (see first link below), a good way to approach this is to ask, "Which group does Andy belong to? Where can I best place him to figure out his risk?" We know that Andy is over 50 and that he has a positive FOBT. So we need to find the normal frequencies for colorectal cancer and positive FOBT in people older than 50.

The average risk in natural frequencies

The first question that comes to mind, which you should ask your doctor, is "Does anyone know the frequency of colorectal cancer in men over 50?"
In the general population, the frequency of colorectal cancer is 30 out of every 10,000 people older than 50. Since no first-degree relative is known to have been affected with cancer, Andy belongs to the general group of people over 50 years of age. If no FOBT had been performed, Andy's risk for colorectal cancer would then be 3 in a thousand (30/10,000). That is the average risk. However, now that a test was performed and was positive, you may conclude that Andy's risk of colorectal cancer is no longer an average risk, but a higher one. But what is his risk?

Franklin's law tells us that the results of the FOBT cannot be taken as a certainty. This means it's not 100% sure that a positive test means the test subject has colorectal cancer. To be able to figure out which group Andy belongs to, we need to consider how frequently the test is wrong in those who have no cancer, and how often it's negative in those who do have cancer. The FOBT has been carefully evaluated in many people, and it's known that it will be 'positive' in 3 out of 100 people who don't have any colorectal cancer. Moreover, it will be 'positive' in only about 50% of patients who do have colorectal cancer (i.e. approximately 1 time out of 2). Now we can figure out Andy's actual risk for cancer.

To do this, we must imagine a group of people all having both the known features that feature in Andy's case: older than 50 and positive for FOB. If we take 10,000 people over 50, 30 will have colorectal cancer, but only 15 of these will test positive for FOB. The remaining 9,970 people without cancer will, however, turn up 300 of them positive for FOBT. So, in total, there are 315 positive FOB in the 10,000 people but only 15 of these 315 actually have colorectal cancer. Now it is easy to figure out Andy's risk: it is only 15/315, i.e. around 5 chances in a hundred. One may also say that there are more than 95 chances in 100 that Andy has not got colorectal cancer. We can simplify this further to 1 chance in 20, or 5 in 100 (5%).

The pie chart below helps you visualize the risk size for colorectal cancer in red, versus no such cancer in blue, the chart representing all the apparently healthy people sharing with Andy the same characteristics regarding age and positive FOBT result.

On the other hand, if his FOBT had been negative, do you think it's certain that Andy has not got colorectal cancer? Of course not, you now know Franklin's law; if the test was negative, then there would still be 15 out of 9,700 chances that a cancer is present. So the risk would be very small (less than 2 chances in 1,000), but not non-existent.

Conclusions

What is the lesson learned here? If Andy had enquired about the risks before agreeing to the test, the physician would have explained what was to be expected and why. As shown above, a good explanation of the actual risk for colorectal cancer in the event of a positive test would have mitigated the worry and emotional burden on Andy and his family. A course of action might have been determined, which could look like this: "If the FOBT is positive, you still have 95 chances out of 100 that you don't have colorectal cancer. However, a flexible sigmoidoscopy, as recommended by the American Cancer Society, will be done to see if you aren't one of the 5 out of 100 cases who do indeed have colorectal cancer."