Zap!

Fazel and colleagues use insurance claims data to generate population estimates of the radiation doses that people age 18-64 get from medical imaging procedures. You only get to read the abstract, but that's okay, the whole article is not exactly scintillating reading. Nevertheless, it does stimulate Michael Lauer to riff on it in a more interesting way. You aren't allowed to read that either, so let me give you the 4-1-1.

Radiation doses are measured in millisieverts, which are a composite unit intended to represent the biological effect of exposure to ionizing radiation, i.e. taking into account the total energy, how it was delivered over time and volume of tissue, the type of radiation, and the body part irradiated. OSHA doesn't allow people in the nuclear industry to get more than 50 mSv/year, but that doesn't mean lower exposures are totally safe: as far as we know there is no threshold effect to the biological effects of ionizing radiation, in other words a very low exposure still constitutes a risk, albeit also very low. So it's best to avoid exposure unless there are sufficient offsetting benefits.

Having a job is clearly one such benefit. Most jobs, from taxi driver to coal miner to medical sociologist, involve non-zero risk, but we tolerate it because a man, or whoever, has to eat, and that includes Homer Simpson (who as I assume everyone knows works in a nuclear power plant). But when you're talking about medical procedures given to millions of people, even a very small risk to each individual can add up to a meaningful population risk. Fazel et al find that in fact, quite a few people get pretty substantial doses from medical imaging procedures. In fact, 1.9 out of 1,000 people in their sample got more than 50 mSv in a year. (Which could potentially get Montgomery Burns in trouble if it happened to Homer Simpson although as we know Burns owns the authorities.) 18.6 out of 1,000 got 20-50 mSv, and almost 20% got lower doses. The biggest doses come from Myocardial Perfusion Imaging, commonly known as "stress tests."

These involve injecting radioactive material into the bloodstream. General Electric, which sells the equipment for stress tests, is pleased to tell you that:

The radioactive materials, MYOVIEW and Thallium, have been shown to be safe with low incidence of adverse reactions. This test and materials are used routinely worldwide, under physician supervision, for myocardial perfusion imaging.

Neither MYOVIEW nor Thallium are "dyes," and there should be no serious side effects from their injection.

Uh huh. Dr. Lauer tell us that it's commonplace for physicians to send middle-aged men with risk factors for heart disease, but no heart disease symptoms, for such tests, and often others as well such as CT angiography. However, "no large-scale, randomized trials have shown that imaging in these circumstances prolongs life, improves quality of life, prevents major clinical events, or reduces long-term medical costs." Furthermore, they expose people to radiation, which increases people's long-term risk for cancer. According to one estimate, as many as 2% of all cancers may be attributable to CT scanning. Cancers attributable to stress tests would be in addition to that.

So what do we need?

A) We need the fascistic, socialistic, ObamaNaziistic Comparative Effectiveness Research, to tell us what benefit there actually is from such imaging procedures under what circumstances, so we know whether they are worth the immediate monetary cost, and the long-term cancer risk.

B) We need rationing, in the form of practice guidelines that will tell physicians and patients when it's worth it and when it isn't to do these tests.

Guess who isn't going to like that? Radiologists, who are among the highest paid medical specialists. Oh yeah, GE isn't going to like it either.

C) We need universal, comprehensive, single payer national health care.