For a lot of people pushing stuff like PSA screening, I think the underlying thought process is that there is no such thing as too high a cost to pay to save one life. They would say, who cares if the other eleven men whose cancer was detected have some urinary symptoms, who cares what the tests cost, who cares what the opportunity cost of doing this versus something else is, as long is one life is saved, it's worth it. This is the same mentality that lead to peanut-free schools. Or masking for COVID. It doesn't matter if the number needed to treat is a million, they'll put a million people through it. This is a very American mentality; most countries with public healthcare systems think of things in a more two-sided way.
I think that ultimately this is a values discussion. Potential harm vs potential benefit. I am not comfortable with not recommending PSA and am not comfortable with universally recommending. What I am comfortable with is presenting the arguments for both and letting the patients decide
Patients can understand. My estimate is that when trying to present this date neutrality, about 90% opt not to have screening.
It seems to me that the first thing to remember with screening, especially for prostate and breast cancer, is that the RCTs deal with populations and not individuals. Given that the incidence of both prostate cancer and breast cancer are relatively small (~10%) and may be detected at am advanced age, it is not surprising that the RCTs reveal no benefit. But the issue is not whether one is going to die, because everybody is going to die eventually, but that one dies at the appropriate time; i.e. not prematurely. In other words a breast cancer or prostate cancer death in somebody who is fit of mind and body is a tragedy; on the other hand, if you have somebody who is suffering from dementia or multiple other severe comorbidities such as a debilitating CVA, CHD, etc...., one is dealing with quite a different situation.
The second thing to bear in mind is family history. A man or a woman with a family history of prostate or breast cancer, respectively, needs to be carefully followed up over many years. After all, it's not a single PSA value that matters but the PSA velocity over time. And a rapid increase in PSA over a period of 6 months or a year is an indication of something bad going on. And the truth is that it is better to be incontinent and have ED than to be 6 ft under. It is also true that PCA surgery has improved extensively in the hands of surgeons who do virtually nothing but; on the other hand the outcomes are poor with surgeons who do few radical prostatectomies. This should come as no surprise given that a radical prostatectomy involves the same level or surgical skill as neurosurgery.
The third thing to remember is that we all know of friends or family where the man has died prematurely of prostate cancer (rather than with prostate cancer). So this is not some rare occurence involving a rare cancer.
Finally, let me provide the example of Francis Collins, the ex-director of the NIH. He wrote about this extensively in the Washington Post. Here is a guy who in his early 70s developed prostate cancer confirmed by biopsy. At the stage of diagnosis it was presumably low grade Gleason 3+3 as he opted for surveillance. However, this is a guy who also had a family history of PCA. About 5 years after diagnosis, he goes in for his yearly PSA and biopsy, and lo and behold from one year to the next he has advanced Gleason 9 PCA. Supposedly he didn't have extra-prostatic involvement, but one can certainly not exclude micrometastases for an aggressive Gleason 9 PCA. He then had a radical prostatectomy, and given his age and the fact that he had this done at the NIH where the doctors are great researchers but don't exactly do a ton of prostatectomies compared to the urological surgeons at Hopkins who specialize exclusively in radical prostatectomies, he had quite severe post-op complications include various infections. He recovered in the end, but this course is a hardly surprise and would have been avoided and his PCA definitively cured had he had a radical prostatectomy upon first discovery. This is an object lesson to learn from.
In 40 years of primary-care clinic practice, I always followed the blindfold rule: If you don't look for it, you'll never see it. The secret to "screening" is doing the quick, cheap and easy looks routinely, no matter how boring: removing the 20 year old plug of wax and "miraculously" restoring hearing to a deaf old woman, finding the prostatic nodule in the 38 year old pilot, and so on. I did rectal exams and was really good at finding nodules.
One patient was a plaintiff's attorney who at the time made a good living suing us for missing Ca prostate. Of course he wanted both a PSA and a DRE with his annual PE. One year his PSA had doubled from low-normal to medium-normal. He had no nodule.
I said, "You've researched this more carefully than I; what should we do?"
He said, "I need a biopsy."
This showed Gleason 9 Ca; prostatectomy showed that it had replaced 90% of his prostate.
Finding this was not luck; it was clear thinking and understanding that "screening' belongs to populations, not individuals.
After the USPSTF recommendation came out my thought was that PSA screening should be applied and evaluated using clinical judgement just like everything else. A significantly elevated PSA in a 55 year old man should be approached differently than one in a 70 year old with other medical problems, or perhaps differently that in a 70 year old who is fit and healthy. (Like Dr Cifu I had a couple or younger men in my practice whose lives I'm sure were saved by screening - and in addition they seemed to tolerate treatment much better.) Take the patient's age and general health status into account as to who to screen. Also the level of the value. Often simple observation of the test value, and nowadays MRI, to determine the course of the finding is sufficient. Late stage prostate cancer is on the rise in recent years, which may actually be attributed to reduction in PSA screening because of the USPSTF recommendation. Late stage prostate cancer is a killer - second only to lung cancer in men. I wouldn't want it. So are we to wait until we find a prostate nodule on exam or symptomatic metastases to diagnose this condition?
I'd love to see data on percent of physicians who get PSA screening. Physicians see the patients who have been "saved" from aggressive PC by early screening. They see the patients who have been left with incontinence or impotence from curative prostate surgery. They see the patients who have been diagnosed with metastatic PC and died slowly from it. They get first hand accounts of the anxiety patients feel from getting high PSA scores and having to undergo multiple diagnostic procedures. They know how painful prostate biopsies can be.
My sense is the vast majority of physicians under age 80 years are getting PSA tests.
I’d love to see stats on sensible doctors ;) I’d like to know who in Medicine has a big interest in these difficult questions. I do not do Medicine in this field, but I follow the topic and it seems like a very difficult topic to me. (maybe i’m just dumb !)
BRAVE discussion, gentlemen, thank you. The "secret ethics of cancer screening," as you wisely put it, will one day be seen as an embarrassment of our era. Meanwhile, incredibly, the editorial accompanying the new paper called it "an abdication of responsibility" to let individuals make the decision, rather than make it for them! (https://www.nejm.org/doi/full/10.1056/NEJMe2509793). An honest study report would have plainly stated all-cause mortality, and reported harms (increases in biopsies, surgeries, impotence, incontinence, etc.). They did not do that, for obvious reasons. While we wait, perhaps vainly, for their harm data, here is a best-available-data infographic: (https://researchtranslation.substack.com/p/psa-screening-for-prostate-cancer). Thank you again for this important discussion.
Final paragraph, for your potential chuckling enjoyment: “The ERSPC was a monumental effort, a 30-year trial that involved hundreds of researchers and more than 150,000 patients. To best use the results of that research to improve the lives of our patients, we have to do better than abdicate the responsibility for PSA decision making to individual patients and instead formulate policies on PSA screening that maximize benefits and minimize harms.” Don’t you see? We should respect the researchers’ effort by scientifically misinterpreting their results in the shadow of our ideological obsessions, and forcing them on hapless elderly men. Obvi.
Good Fornight episode, thanks. The J-shaped phenomenon is an artefact driven by frailty. Primarily observational studies, and observational studies based on response to treatment in RCTs cannot exclude it (DOI: 10.1038/jhh.2013.78). Only by keeping the groups created by randomization and testing the interaction between treatment and BP levels can the existence of a J-shaped phenomenon be documented or ruled out, as shown by the Blood Pressure Trialists for systolic BP (DOI: https://doi.org/10.1016/s0140-6736(21)00590-0) and Andrew for diastolic BP. There is a new MA, with three additional studies, showing that there is no interaction by baseline diastolic BP for the incidence of the composite CV outcomes and all-cause mortality (DOI: https://doi.org/10.1681/asn.0000000539). 100 to live 100.
How do modern surgical methods (Da Vinci) or center experience in prostate resection influence the downsides/side effects of prostatectomy? I would - probably naively - expect that side effects go way down with above mentioned points. Any insights in this?
Of course the side effects go down with increased surgical expertise. And surgical expertise for radical prostatectomies basically comes by doing nothing but and many of them. The results of Da Vinci versus open surgery are not different providing the level of expertise of the 2 surgeons is the same. In some ways open surgery is better because on can actually feel things. But of course, the pain at the incision site for 2 days post-op is a lot worse for open surgery than laparoscopic/robotic surgery.
could you add a transcription of a video each time it is included. When I watch a video, I have no time to reflect and nothing to look back and review. Just saying. Thank you.
Very interesting as always, but with respect it is not doctors business to make preference judgments, as you imply when discussing what are ‘acceptable’ NNs (for whatever) in the patient , not public health ,context..The individual patient is not giving their informed consent to PSA screening when they are not told (at the very least) the NNBenefit in terms of avoiding (i) dying from PC AND (ii) dying from Any Cause, from being diagnosed and treated for PC; as well as being told the NNHarm from (at least) (i) incontinence and (ii) erectile dysfunction. A person’s choice, if informed by these numbers, may be to be screened or not, but neither can then represent either 'over' or 'under' diagnosis. These are purely population-level phenomena. I personally believe ODOT in cancer (and many other conditions) at the population level is mainly produced by condition-specific pressure groups (and healthcare systems!) that, for various reasons, do not wish to have people required to give truly informed consent to screening. Since you all agree it’s the patient’s decision, to be based on their informed preferences, isn’t it your ethical duty, not just legal obligation, to always reveal the 4 basic (‘secret’) NNs? And to push back against the huge pressures mentioned by Andrew .
For a lot of people pushing stuff like PSA screening, I think the underlying thought process is that there is no such thing as too high a cost to pay to save one life. They would say, who cares if the other eleven men whose cancer was detected have some urinary symptoms, who cares what the tests cost, who cares what the opportunity cost of doing this versus something else is, as long is one life is saved, it's worth it. This is the same mentality that lead to peanut-free schools. Or masking for COVID. It doesn't matter if the number needed to treat is a million, they'll put a million people through it. This is a very American mentality; most countries with public healthcare systems think of things in a more two-sided way.
I think that ultimately this is a values discussion. Potential harm vs potential benefit. I am not comfortable with not recommending PSA and am not comfortable with universally recommending. What I am comfortable with is presenting the arguments for both and letting the patients decide
Patients can understand. My estimate is that when trying to present this date neutrality, about 90% opt not to have screening.
How do you measure diastolic BP: office visit; patient records: 24 hour ambulatory?
It seems to me that the first thing to remember with screening, especially for prostate and breast cancer, is that the RCTs deal with populations and not individuals. Given that the incidence of both prostate cancer and breast cancer are relatively small (~10%) and may be detected at am advanced age, it is not surprising that the RCTs reveal no benefit. But the issue is not whether one is going to die, because everybody is going to die eventually, but that one dies at the appropriate time; i.e. not prematurely. In other words a breast cancer or prostate cancer death in somebody who is fit of mind and body is a tragedy; on the other hand, if you have somebody who is suffering from dementia or multiple other severe comorbidities such as a debilitating CVA, CHD, etc...., one is dealing with quite a different situation.
The second thing to bear in mind is family history. A man or a woman with a family history of prostate or breast cancer, respectively, needs to be carefully followed up over many years. After all, it's not a single PSA value that matters but the PSA velocity over time. And a rapid increase in PSA over a period of 6 months or a year is an indication of something bad going on. And the truth is that it is better to be incontinent and have ED than to be 6 ft under. It is also true that PCA surgery has improved extensively in the hands of surgeons who do virtually nothing but; on the other hand the outcomes are poor with surgeons who do few radical prostatectomies. This should come as no surprise given that a radical prostatectomy involves the same level or surgical skill as neurosurgery.
The third thing to remember is that we all know of friends or family where the man has died prematurely of prostate cancer (rather than with prostate cancer). So this is not some rare occurence involving a rare cancer.
Finally, let me provide the example of Francis Collins, the ex-director of the NIH. He wrote about this extensively in the Washington Post. Here is a guy who in his early 70s developed prostate cancer confirmed by biopsy. At the stage of diagnosis it was presumably low grade Gleason 3+3 as he opted for surveillance. However, this is a guy who also had a family history of PCA. About 5 years after diagnosis, he goes in for his yearly PSA and biopsy, and lo and behold from one year to the next he has advanced Gleason 9 PCA. Supposedly he didn't have extra-prostatic involvement, but one can certainly not exclude micrometastases for an aggressive Gleason 9 PCA. He then had a radical prostatectomy, and given his age and the fact that he had this done at the NIH where the doctors are great researchers but don't exactly do a ton of prostatectomies compared to the urological surgeons at Hopkins who specialize exclusively in radical prostatectomies, he had quite severe post-op complications include various infections. He recovered in the end, but this course is a hardly surprise and would have been avoided and his PCA definitively cured had he had a radical prostatectomy upon first discovery. This is an object lesson to learn from.
In 40 years of primary-care clinic practice, I always followed the blindfold rule: If you don't look for it, you'll never see it. The secret to "screening" is doing the quick, cheap and easy looks routinely, no matter how boring: removing the 20 year old plug of wax and "miraculously" restoring hearing to a deaf old woman, finding the prostatic nodule in the 38 year old pilot, and so on. I did rectal exams and was really good at finding nodules.
One patient was a plaintiff's attorney who at the time made a good living suing us for missing Ca prostate. Of course he wanted both a PSA and a DRE with his annual PE. One year his PSA had doubled from low-normal to medium-normal. He had no nodule.
I said, "You've researched this more carefully than I; what should we do?"
He said, "I need a biopsy."
This showed Gleason 9 Ca; prostatectomy showed that it had replaced 90% of his prostate.
Finding this was not luck; it was clear thinking and understanding that "screening' belongs to populations, not individuals.
After the USPSTF recommendation came out my thought was that PSA screening should be applied and evaluated using clinical judgement just like everything else. A significantly elevated PSA in a 55 year old man should be approached differently than one in a 70 year old with other medical problems, or perhaps differently that in a 70 year old who is fit and healthy. (Like Dr Cifu I had a couple or younger men in my practice whose lives I'm sure were saved by screening - and in addition they seemed to tolerate treatment much better.) Take the patient's age and general health status into account as to who to screen. Also the level of the value. Often simple observation of the test value, and nowadays MRI, to determine the course of the finding is sufficient. Late stage prostate cancer is on the rise in recent years, which may actually be attributed to reduction in PSA screening because of the USPSTF recommendation. Late stage prostate cancer is a killer - second only to lung cancer in men. I wouldn't want it. So are we to wait until we find a prostate nodule on exam or symptomatic metastases to diagnose this condition?
I'd love to see data on percent of physicians who get PSA screening. Physicians see the patients who have been "saved" from aggressive PC by early screening. They see the patients who have been left with incontinence or impotence from curative prostate surgery. They see the patients who have been diagnosed with metastatic PC and died slowly from it. They get first hand accounts of the anxiety patients feel from getting high PSA scores and having to undergo multiple diagnostic procedures. They know how painful prostate biopsies can be.
My sense is the vast majority of physicians under age 80 years are getting PSA tests.
I’d love to see stats on sensible doctors ;) I’d like to know who in Medicine has a big interest in these difficult questions. I do not do Medicine in this field, but I follow the topic and it seems like a very difficult topic to me. (maybe i’m just dumb !)
Love this!
BRAVE discussion, gentlemen, thank you. The "secret ethics of cancer screening," as you wisely put it, will one day be seen as an embarrassment of our era. Meanwhile, incredibly, the editorial accompanying the new paper called it "an abdication of responsibility" to let individuals make the decision, rather than make it for them! (https://www.nejm.org/doi/full/10.1056/NEJMe2509793). An honest study report would have plainly stated all-cause mortality, and reported harms (increases in biopsies, surgeries, impotence, incontinence, etc.). They did not do that, for obvious reasons. While we wait, perhaps vainly, for their harm data, here is a best-available-data infographic: (https://researchtranslation.substack.com/p/psa-screening-for-prostate-cancer). Thank you again for this important discussion.
The editorial is paywalled. Too bad because I needed a chuckle. Fish or Cut Bait—I would say the latter.
Final paragraph, for your potential chuckling enjoyment: “The ERSPC was a monumental effort, a 30-year trial that involved hundreds of researchers and more than 150,000 patients. To best use the results of that research to improve the lives of our patients, we have to do better than abdicate the responsibility for PSA decision making to individual patients and instead formulate policies on PSA screening that maximize benefits and minimize harms.” Don’t you see? We should respect the researchers’ effort by scientifically misinterpreting their results in the shadow of our ideological obsessions, and forcing them on hapless elderly men. Obvi.
Kudos x XIII for excellent Fortnight presentations: keep them coming please.
Thank you, George.
Good Fornight episode, thanks. The J-shaped phenomenon is an artefact driven by frailty. Primarily observational studies, and observational studies based on response to treatment in RCTs cannot exclude it (DOI: 10.1038/jhh.2013.78). Only by keeping the groups created by randomization and testing the interaction between treatment and BP levels can the existence of a J-shaped phenomenon be documented or ruled out, as shown by the Blood Pressure Trialists for systolic BP (DOI: https://doi.org/10.1016/s0140-6736(21)00590-0) and Andrew for diastolic BP. There is a new MA, with three additional studies, showing that there is no interaction by baseline diastolic BP for the incidence of the composite CV outcomes and all-cause mortality (DOI: https://doi.org/10.1681/asn.0000000539). 100 to live 100.
Great discussion. Thanks for all your time and expertise.
Thanks. Sorry I missed it. I love your work.
How do modern surgical methods (Da Vinci) or center experience in prostate resection influence the downsides/side effects of prostatectomy? I would - probably naively - expect that side effects go way down with above mentioned points. Any insights in this?
Of course the side effects go down with increased surgical expertise. And surgical expertise for radical prostatectomies basically comes by doing nothing but and many of them. The results of Da Vinci versus open surgery are not different providing the level of expertise of the 2 surgeons is the same. In some ways open surgery is better because on can actually feel things. But of course, the pain at the incision site for 2 days post-op is a lot worse for open surgery than laparoscopic/robotic surgery.
could you add a transcription of a video each time it is included. When I watch a video, I have no time to reflect and nothing to look back and review. Just saying. Thank you.
Transcript button available just below the picture on the website.
Very interesting as always, but with respect it is not doctors business to make preference judgments, as you imply when discussing what are ‘acceptable’ NNs (for whatever) in the patient , not public health ,context..The individual patient is not giving their informed consent to PSA screening when they are not told (at the very least) the NNBenefit in terms of avoiding (i) dying from PC AND (ii) dying from Any Cause, from being diagnosed and treated for PC; as well as being told the NNHarm from (at least) (i) incontinence and (ii) erectile dysfunction. A person’s choice, if informed by these numbers, may be to be screened or not, but neither can then represent either 'over' or 'under' diagnosis. These are purely population-level phenomena. I personally believe ODOT in cancer (and many other conditions) at the population level is mainly produced by condition-specific pressure groups (and healthcare systems!) that, for various reasons, do not wish to have people required to give truly informed consent to screening. Since you all agree it’s the patient’s decision, to be based on their informed preferences, isn’t it your ethical duty, not just legal obligation, to always reveal the 4 basic (‘secret’) NNs? And to push back against the huge pressures mentioned by Andrew .