|Sign in London Underground car|
During my hiatus, Chris Snowdon asked for my opinion about a commentary by the Australian public health community’s own keystone kop, Simon Chapman, which argued that screening for prostate cancer is a bad idea. Even though it was Chapman, the basic advice turns out to be good — after all, a stopped clock is still right twice a day. (For more on why I am being so negative about someone who got the right answer, the above links go to Snowdon’s and my analyses of some of Chapman’s other “contributions”.)
To see why it was good advice, we need only look to something that came out a few weeks ago, in which a US panel led by Virginia Moyer, a former colleague of mine whose clock keeps perfect time, recommended that the common practice of PSA screening tests cease. Here is Moyer briefly arguing for it in her own words. An even briefer version: Because most cases of prostate cancer are unlikely to grow fast enough to cause a problem before someone dies of something else, and perhaps the cases that will be deadly are unlikely to be prevented by screening, and because the treatment that results from screening kills some patients, it turns out there is no detectable mortality benefit from screening. Meanwhile, the treatment often causes nasty side effects (impotence, incontinence).
Following this recommendation, there was the usual outcry against any recommended curtailment in wasteful medical spending. It was the same as the reaction to scale back mammography that I wrote about before. As always, the most remarkable comments are the breathless testimony from cancer survivors who are absolutely sure that the screening saved their lives, when obviously they have no way of knowing that, and indeed the statistics show that it is almost certainly not true.
But what is a bit different about this case is how clear it is that the treatment, not the screening, is the real problem. In contrast with a mammogram, a PSA test is not harmful and will never cause the disease it is meant to prevent. Of course, the problem is that once someone screens positive, he is very likely to demand treatment even though it turns out to not be a good idea. Still, the last chance to avoid the cost comes not from the decision to test, but from the decision that is made afterward. Of course, if we agree that treatment after screening positive is a bad idea — and obviously the same is true for treatment after screening negative — then there is no value in screening other than to just know.
Those observations, in turn, led to a spate of occasionally interesting articles about whether it is better or worse to know. A lot of the discussion came back to how difficult it is to just live with the knowledge rather than acting on it, even if acting is not beneficial. The decision to treat is what actually causes the harm, but since that apparently bad decision is inevitable following a positive screen, the only way to avoid the mistake is to avoid the triggering knowledge. It makes sense, but it is still interesting that people cannot resist acting.
The above-pictured “In an emergency” sign from the London Underground is really amusing, with its major point of advice being “Do not take any risks”. What it is really trying to say (I assume) is to favor inaction over action until advised to do otherwise, which is quite often terrible advice and obviously entails some risks. But someone apparently calculated that it is good advice, on average, for subway emergencies, and it turns out to be the right advice for treating the average detected case of prostate cancer.
To finish the story I began with, Snowdon’s incredulity about Chapman’s advice did not focus on the overall merits of the advice itself, but rather on how Chapman tried to support the claim: He made a wandering argument in which he suggested that screening and treatment does have mortality benefits, but since prostate cancer kills mostly old men who have already had a good life and might die of something else shortly anyway, they should favor quality of life (avoiding the damage from the treatment) over longevity. Snowdon pointed out the strangeness of this argument that men should not have the option of going for the supposed longevity gains, coming as it was from someone who is best known demanding the opposite choice. Or, more pointedly, known for his willingness to pervert science, abuse the social contract, and do pretty much anything he can think of that might keep people from choosing to smoke or even choose to be in smoky environments. He will ferociously fight to deprive people of the choice to take that risk, however much their lives might benefit from the choice, but then demands that men accept another mortality risk because he judges that their lives will benefit.
The funny thing is that, according to Moyer et al., it is reasonable to deprive men of this choice because there seems to be no upside. But Chapman claimed there is an upside to treatment, but that the choice to pursue it should still be denied. A slow clock, rather than being right twice per day, might only manage once per month.