A discussion on Twitter caught the eyes of my colleagues yesterday, and raised a very interesting question: should insurance companies be allowed to do PSA testing to detect prostate cancer on men as a condition of getting insurance?
What started the discussion was a blog post by a well-known and respected medical blogger who goes by the name "Skeptical Scalpel." In his blog he detailed the saga of a 56 year old man who had a pre-employment physical in order to be covered by his new company's health insurance plan. He was not informed that he was going to have a PSA test. It was just done as part of the process. No informed consent, no nothing, just stick out your arm, have blood drawn, and register your surprise that the test was done once the results come back.
The United States Preventive Services Task Force recommends against any man having the test to detect prostate cancer. Even among those who say the test is an option, -including the American Cancer Society-recommend that a man have a full, informed understanding of the pros and cons of PSA testing before getting the test, given the uncertainties of whether it really saves lives and the known frequency of side effects associated with treatment for prostate cancer. None of those recommendations were part of the consideration for this man when he was examined for his insurance plan.
You can imagine what happened: his PSA was slightly elevated at 5.9. He was "rated" by the insurer, charged an extra $200 a month for his health insurance, and may well have to have a number of additional studies. Not to mention that he may have prostate cancer, and may undergo more treatments-as a result of a test that is of uncertain value for most men. And, by the way, the odds are overwhelming that the gentleman in question does not have prostate cancer, but will have to go to considerable lengths (and some risk) to find out. Because he had not discussed the test, he did not have the opportunity to prepare for it correctly, so his reading may have been falsely high; we'll never know. And, if he does have prostate cancer, as reflected in the USPSTF recommendations, he has a significant chance of having long term side effects from the treatment, but little proven benefit in terms of saving his life.
Turns out if you Google the question about PSA testing for life insurance policies, you find a long-standing debate about the role of PSA testing for life insurance policies, and even some companies that will help you get insurance if your PSA is elevated. But that prompts the question: should insurance companies be doing routine PSA screening and rating men-whether it is for life insurance or health insurance-without the men knowing they are getting the test, or offering any form of informed consent? Do life insurance companies know something the rest of us don't, like PSA results correlate with other more typical measures of longevity, such as blood pressure, weight, waist circumference and cholesterol for what is probably a term insurance policy with a 10 or 15 year duration?
And then there is this oft ignored bit of information: men who participated in a large clinical trial done here in the United States to determine the value of PSA testing actually had more deaths in the group that had the PSA testing than the men who were assigned to the "no PSA" testing group? (Yes, I know there were many men in the "no-PSA" group that actually did get the test, but the data is the data. I also know that the numbers were not statistically significant, so all we can say really is that there was a "trend" but not statistical proof of the observation. But it prompts the question of whether or not getting a PSA test may actually diminish your survival. Sort of reverse logic: get the test and possibly die earlier than you would have otherwise.)
Health insurers, by the way, are not allowed to consider genetic testing data as part of their assessment of your risk of disease when you apply for your policy. But apparently they can consider the information supplied by a PSA test, even though that test is generally held to have a limited usefulness in everyday practice. What you may not know is that disability and life insurers are not bound by the same rules as health insurers when it comes to genetic tests and patient confidentiality. If you have an abnormality such as BRCA or Lynch syndrome found on a genetic test, you must declare it if asked.
So in this confusing world, I ask the same question that Skeptical Scalpel asked: should life and health insurers be bound by the same guidelines that everyone else who cares for patients are supposed to follow when it comes to the PSA test? Or should they just ignore the debate and do what they want to do?
It is clearly a dilemma, and a situation that-once again--is not being guided by the science. The question should be: what does the evidence show? If the evidence shows that there is considerable uncertainty about the value of PSA testing then no one should have to get a test that provides no useful or meaningful information without their full informed consent, and without coercion.