$5 for 3 months. Save 83%.
$5 for 3 months. Save 83%.

Doc: What happens when test results are at odds?

Keith Roach
To Your Health

Dear Dr. Roach: I am an 85-year-old male in pretty good health. I recently had a stool test by the Veterans Affairs hospital. They came back and stated that I had blood in my stool, which indicates colon cancer, and that I should have a colonoscopy ASAP. My insurance company provided me with some eight possible signs and symptoms to consider. I had none, and no family history. I consulted my very qualified primary physician.

For the past 15 years, I have had a colonoscopy every five years. The last was in 2010. My results were always “A-OK,” with always one small polyp in the same place.

I had a ColoVantage (methylated Septin 9) test. The interpretation came back as negative. My physician informed me that the negative ColoVantage blood test was a good result. He believed there was no further need for a colonoscopy. May I have your opinion?


Dear S.M.: When two tests have completely different results, a wise physician, like your well-qualified primary physician, looks at the test characteristics. I’m sure he found that the ColoVantage test, a blood test looking for an abnormal gene associated with colon and other cancers, has a high likelihood of being correct when it says there is no abnormal gene present, and he knows that the fecal blood test has many false-positive results.

I have the luxury of taking things one step further, and I performed a formal decision analysis (based on Bayes’ theorem) using your likelihood of having colon cancer based on being an 85-year-old man, the results of your positive stool blood test and the results of your negative methylated Septin 9 test.

Putting all the information together, I find you have a 0.4 percent chance of having colon cancer. Since the risk of having an adverse event from a colonoscopy at your age is about the same, I would recommend against getting a colonoscopy, as your physician did.

Dear Dr. Roach: I have noticed that three of my extended episodes with A. fib in the past five years have been after periods of reduced oxygen in my lungs. Two times I have had a head cold, and one time I was at 9,000 feet elevation in Colorado.

Is there any possible correlation between A. fib and oxygen deprivation?

If this situation is logical, is it also logical that A. fib can be treated with oxygen in order to make the additional electrical impulses unnecessary? Has oxygen ever been used as a treatment for A. fib?


Dear D.T.: You are partly correct. Having low oxygen levels to the heart is, indeed, a risk factor for development of atrial fibrillation. Lung disease, including sleep apnea, certainly increases risk. Although it makes sense that high altitudes would worsen existing A. fib or increase its likelihood, that hasn’t been proven. Treatment for atrial fibrillation can include reversing the underlying cause, if possible, and although that isn’t always possible, treating chronic lung disease is one way of doing so. If people have low oxygen levels in the blood (easily checked now in the office), then oxygen may, indeed, be a useful treatment.

However, colds do not affect oxygen levels in the blood, although a very severe pneumonia might. People with normal levels of oxygen in the blood don’t need or benefit from supplemental oxygen, even with atrial fibrillation.

Email questions to