Well, no, not really

Screening men for breast cancer may be even more effective than for women


Screening men for breast cancer may be even more effective than for women after testing picked up nearly four times the rate of tumours.

Not what’s being said:

Breast cancer in men is rare, with around 1.5 cases for every 100,000 men in Britain, equating to 390 men diagnosed each year compared to 54,800 cases in women.

But around one in 1000 men carry gene mutations that put them at greater risk.

Now a study has shown that when 1,869 high risk men were screened over 12 years, cancer was detected in 18 of every 1,000 exams, compared to five in 1,000 for women of both high and average risk.

Selective screening of men appears more effective than unselective screening of women.

Something that seems fair and reasonable when we think about it.

12 thoughts on “Well, no, not really”

  1. There’s also a difference between being effective and being cost-effective. Mass screening for rare conditions is unlikely* to be cost-effective, no matter how effective the actual detection process is.

    *Assuming we’re talking about some sort of physical examination – if it was one more blood test that can be added to a series that are already being done, then fair enough, but presumably this is the male equivalent of mammography.

  2. I’ve yet to see any persuasive evidence that unselective breast cancer screening of women is worth it. Or unselective screening of men for prostate cancer. Further I don’t think I’ve ever seen evidence that an annual “health check”, as beloved by Americans and kinda, sorta offered by the NHS, is worth it either.

    I conclude that the lesson is that if you detect what might be a nasty symptom get to the doctor quickly. Otherwise, relax.

  3. There’s only scant evidence that breast screening has any benefit. Before we even get to the risk and cost side of the equation.

    An “annual check up” would probably have picked up the thing that put me in hospital (with sequelae) this time last year, but only if done rigorously and annually, and I’ll admit it was rather odd. Where it does help a lot is with chronic and very common stuff, particularly hypertension and diabetes. Some two-figure percentage of people walking into the doctor’s office are borderline for one or both of those, and if they leave the office with the knowledge of that, some small proportion will be sufficiently motivated to make the lifestyle changes that, while they may not ultimately prevent full-blown disease, can delay it substantially.

  4. There is no unselective breast screening in women.

    Screening is selective – mammogram (usually) every two years – determined by age group, family history of breast cancer, previous breast cancer.

    Given the number of false positives and trauma thus caused, and false negatives, plus radiation dose, the value of breast screening is disputed.

  5. @BiG “Where it does help …” But where is the evidence of that helping? The Times, yesterday, backed up things I’ve seen written again and again.

    “The authors looked at evidence from around the world, including one review of 17 trials involving more than 250,000 people. Their conclusion? There is no convincing evidence that health checks significantly reduce your risk of dying [within some finite period] — whether it be from a heart attack, stroke or cancer. They may pick up undiagnosed high blood pressure, kidney disease and type 2 diabetes in some people, but this doesn’t appear to translate into worthwhile benefits.”

  6. All-cause mortality with short duration of follow-up in the general population is not a useful endpoint. It’s an easy endpoint to use, for sure, because it’s kinda indisputable if someone is dead, alive, or lost to follow-up, but it isn’t going to tell you much in this circumstance.

    What would be more likely to show a change over a few years follow-up is morbidity, particularly cardiovascular and renal, but then you have a whole pallete of potential endpoints, so it’s harder to measure reliably.

    I picked diabetes and hypertension because diabetes is, what, 5% of the population, hypertension closer to 50%. We know (without any shadow of a doubt) that untreated diabetes and untreated hypertension cause significant morbidity and mortality, and that if you discover them earlier rather than later, you can do non-pharmacological interventions that slow the progression to overt disease that requires pharmacological intervention. Even if only 10% of people actually follow through with it, that’s a substantial healthcare saving, and substantial improvement in quality of life for those patients.

  7. I thought the annual check up was not as effective as it should be due to the worried well taking the offer up, and almost no one else.

  8. Chris, the question is moot, because whether you measure reduction in terms of achieving a threshold, or control relative to pre-treatment, reducing your bp always comes with an increased opportunity to die later, and of something else.

    The epidemiology on this is solid, at least as good as the smoking/lung cancer (though the association there is stronger).

  9. “What’s the cut-off point for hypertension – 140/90?”

    Who can tell?


    Or consider: “The trial found improved outcomes in high risk patients treated to a target of of 120 mm Hg instead of 140 mm Hg. … They estimate that treating 1,000 people for 3.2 years with more aggressive treatment would result, on average, in a beneficial effect in 16 persons, serious harm in 22, and no benefit or harm in 962.”


    Or there again

    But do not ignore
    Table 3 Errors in measurement of blood pressure

    If this field were considered part of Science rather than Medicine then expressions like “bloody disgrace” would be muttered.

  10. “What’s the cut-off point for hypertension – 140/90?”

    Hard to say. One big trial of which much was made, SPRINT, turned out to have involved bad measurements. So any recommendations based on it are pretty dubious.

    As a non-medic I found this Canadian discussion calm and persuasive (though I’m taking their numbers on trust).


    For anyone with even a little experience with statistical analysis this little gem might give pause.


    Why, you might wonder, don’t more doctors ask themselves such a fundamental question?

  11. There isn’t data for targets below 120/80 because no one would think of trying to target anything lower. Doing so will cause a lot of patients to stop treatment altogether and then you have first a dead trial, and second more dead patients.

    Diuretics versus beta blockers is very much oldhat (sic), since ACEIs are pretty established first-line therapy (or ARBs), usually alongside a thiazide, and have even better cardiac outcomes (and fewer side effects), and the ARBs also seem to preserve renal function better than other drugs (you see, it’s never been a secret that outcomes differ even for the same achieved BP). A bigger controversy is actually whether HCT should give way to chlortalidone or indapamide as the diuretic of choice (depending on renal function) – that’s how finely balanced the discussion is.

    Beta blockers only come into play if that isn’t enough, and, really, CCBs are higher up the list. Some guidelines even have CCB in first line. So a competently managed patient will these days only be offered a beta blocker as their fourth drug, three being insufficient. Then there various types of vasodilators and some more exotic diuretics, direct renin inhibitors, centrally-acting stuff, alpha agonists, even implants and renal denervation if you have truly uncontrollable hypertension.

    I don’t expect treatment lines to change much for a while, or the guideline to target at most 135/85, optimally 120/80, provided the patient can tolerate that and stay compliant.

Leave a Reply

Your email address will not be published. Required fields are marked *