Very well done. Who is sitting on the FDA advisory panel? How could they give this test a 6-4 vote (I'm presuming in favor of) for efficacy and 7-2 (who didn't vote?) for benefits>harm? I looked them up, and from what has been released publicly (which may be incorrect) there are pathologists, oncologists, biostatisticians, surgeons, internists... but no epidemiologists! In my experience pathologists and oncologists frequently oversell benefits of screening. Epidemiologists do better detecting lead-time bias, length bias, and overdiagnosis etc.
Not assessing harms of subsequent testing seems unfathomable to me. If they measure and report OS over 5-10 years I suppose that could be a decent proxy but the whole concept of medicalizing otherwise healthy individuals without measuring harms is wild. Anyone run the NNT / NNH on this?
I'm disappointed to find out GRAIL didn't turn out better but appreciate the analysis. As someone more in favor of proactive screening than the Sensible Medicine authors, I had been hoping GRAIL would have a favorable testing profile for younger individuals.
I ran some rough numbers based on the specificity and a ~80% false positive in 40 year olds and 65% false positive rate in 50 year olds is not it.
Very well done. Who is sitting on the FDA advisory panel? How could they give this test a 6-4 vote (I'm presuming in favor of) for efficacy and 7-2 (who didn't vote?) for benefits>harm? I looked them up, and from what has been released publicly (which may be incorrect) there are pathologists, oncologists, biostatisticians, surgeons, internists... but no epidemiologists! In my experience pathologists and oncologists frequently oversell benefits of screening. Epidemiologists do better detecting lead-time bias, length bias, and overdiagnosis etc.
Fantastic review. Great job on the epidemiological analysis. Very important work!
Excellent summary, thank you!
Not assessing harms of subsequent testing seems unfathomable to me. If they measure and report OS over 5-10 years I suppose that could be a decent proxy but the whole concept of medicalizing otherwise healthy individuals without measuring harms is wild. Anyone run the NNT / NNH on this?
I'm disappointed to find out GRAIL didn't turn out better but appreciate the analysis. As someone more in favor of proactive screening than the Sensible Medicine authors, I had been hoping GRAIL would have a favorable testing profile for younger individuals.
I ran some rough numbers based on the specificity and a ~80% false positive in 40 year olds and 65% false positive rate in 50 year olds is not it.
From the perspective of a Medical Director for an insurance plan - Not Medically Necessary (Not a Covered Benefit)…!
And don’t get me started on the FDA nonsense - ugh.
Can someone explain the math; what does the rate of detection (30%) divided by the probability of a false positive result (0.5%) mean?
Wasn’t the result more positive for pancreatic cancer, which is hard to detect?
Fascinating (in like a not good way…) - as all your articles, thank you