Being actively involved in your own care leads to better outcomes. No test is a panacea but information contributes to agency. There is a lot of undiagnosed disease that benefits from early detection - I believe in giving people more ways to access the care system when it can do the most good.
This is a terrific piece, thank you! II'm currently helping a family member deal with the ramifications of a Prenuvo scan. She's a healthy young adult who paid for the scan looking for reassurance that all is well. Unfortunately, it found a few incidental findings. The most significant of these is a liver lesion that has now been evaluated with labs, a CT, a dedicated abdominal MRI, and most recently a liver biopsy, results still pending. The amount of anxiety this has caused is extraordinary. If somehow we learn that this is a primary liver cancer or metastatic disease from an as yet unidentified primary tumor, there's no evidence to suggest that knowing this now will result in better outcomes than if we had waited for the disease to present itself. Right now the anxiety and life disruption is the most salient issue for my family member; I can't help but also notice how much cost this Prenuvo scan finding has driven to her health insurance. All of the above is to say that I find myself squarely in the "NO" camp!
Great piece! One thing to note is that even though individuals may pay for the tests, they are socializing the costs of chasing down the incidentalomas, all of which likely is billed to insurance. So we all have a vested interest in keeping this whole process sensible and efficient.
Gil Welch and others have nicely and serially described the appeal and hazard regarding overzealous use of screening tools. Read Overdiagnosed- making people sick in pursuit of health . Some day, probably emerging from x-US, with shared access to full population level health related data, we will be able to combine the following - clinical (EHR) data combined with proteomic, metabalomic, genomic, epigenomic, radiographic and full stack digital medical diagnostic data (including serially updated microbiome data)- Once we have sufficient multimodal access AND longitudinal tracking (if we are willing), then we will have achieved a health system which learns. Mt Sinai and Sema4 were attempting this years ago- The effort is worthy, expensive and highly complex.
While we work on the obvious logical though expensive multimodal solution, it is certain that approaches such as screening whole body MRI will randomly harm some people and randomly help others.
responsible entrepreneurs would maintain a registry or will donate their data to a secure, deidentified registry so that we can learn as we go.
There will be a huge number needed to scan to pick up something actionable. And there will be a smaller number needed to harm. But we don't really know because these companies never do clinical trials.
The use of screening tests in asymptomatic people needs to be carefully researched to determine if it causes more harm than benefit--overall. No doubt incidental findings of cancer or aneurysms will be found and prevent untimely deaths. But this should be balanced against the knowledge that some incidental findings are harmless--but biopsies have their own risks. Seems we went through this decades ago with whole body CT scanners. Of course, MRI doesn't cause potential radiation injury as CT scans do, but it was the same issue then as it is now. Benefit has not been established.
As with anything (and everything), before diving into the minutiae of a test, the question each person will want to ask is "What is it for?" If you go searching in a forest for broken branches, you'll find them. It doesn't mean the forest isn't healthy, vibrant, and functioning.
If you're searching for something specific, or you have a list of considerations you want to talk about, like "My menstrual period is so painful, and I have a history of breast cancer, and I've had back pain for a long time," then this can be an (expensive) way to get more information.
But the information has to be triangulated with the presence of symptoms, the goal(s) at the forefront, and the wisdom of both the medical professionals as well as the wisdom of the body & person getting the scan done.
If the test is done to assuage anxiety or confirm that nothing is wrong, those are such broad goals and more likely to result in a long winding rabbit hole and (potentially) solving for problems that aren't problems.
Wider use of DEXA scans for people with metabolic syndrome would be a better use of point-in-time diagnostic imaging as an entre into treatment for obesity/diabetes/CVD, and as a test in the future to assess / reinforce progress and change. We will gain more from accumulating longitudinal data from cheaper non-invasive tests associated with chronic disease risk and feeding these results into systems to improve sensitivity and specificity of more focused predictive assays. Of course, that's not something health insurers, as the system is now configured, values -- predictive medicine, that is. But it's possible to imagine other financial frameworks, including life insurance, that will.
(Side note, some insurers' beginning to encourage liquid biopsy tests, though in a somewhat unfocused way, is progress in terms of insurer interest, but less promising in terms of the likelihood of unnecessary invasive biopsies. It would be better to underwrite the tests as a broader, longitudinal surveillance method to inform improvements in assay accuracy, especially for earlier detection.)
Being actively involved in your own care leads to better outcomes. No test is a panacea but information contributes to agency. There is a lot of undiagnosed disease that benefits from early detection - I believe in giving people more ways to access the care system when it can do the most good.
I think Vinay Prasad's response to this was pretty compelling: https://www.drvinayprasad.com/p/why-you-should-not-get-a-whole-body
This is a terrific piece, thank you! II'm currently helping a family member deal with the ramifications of a Prenuvo scan. She's a healthy young adult who paid for the scan looking for reassurance that all is well. Unfortunately, it found a few incidental findings. The most significant of these is a liver lesion that has now been evaluated with labs, a CT, a dedicated abdominal MRI, and most recently a liver biopsy, results still pending. The amount of anxiety this has caused is extraordinary. If somehow we learn that this is a primary liver cancer or metastatic disease from an as yet unidentified primary tumor, there's no evidence to suggest that knowing this now will result in better outcomes than if we had waited for the disease to present itself. Right now the anxiety and life disruption is the most salient issue for my family member; I can't help but also notice how much cost this Prenuvo scan finding has driven to her health insurance. All of the above is to say that I find myself squarely in the "NO" camp!
"Do we really want AI or an untrained eye to be interpreting this?" Yes to AI; no to "untrained eye" - which no-one is suggesting...
Great piece! One thing to note is that even though individuals may pay for the tests, they are socializing the costs of chasing down the incidentalomas, all of which likely is billed to insurance. So we all have a vested interest in keeping this whole process sensible and efficient.
Gil Welch and others have nicely and serially described the appeal and hazard regarding overzealous use of screening tools. Read Overdiagnosed- making people sick in pursuit of health . Some day, probably emerging from x-US, with shared access to full population level health related data, we will be able to combine the following - clinical (EHR) data combined with proteomic, metabalomic, genomic, epigenomic, radiographic and full stack digital medical diagnostic data (including serially updated microbiome data)- Once we have sufficient multimodal access AND longitudinal tracking (if we are willing), then we will have achieved a health system which learns. Mt Sinai and Sema4 were attempting this years ago- The effort is worthy, expensive and highly complex.
While we work on the obvious logical though expensive multimodal solution, it is certain that approaches such as screening whole body MRI will randomly harm some people and randomly help others.
responsible entrepreneurs would maintain a registry or will donate their data to a secure, deidentified registry so that we can learn as we go.
RG Hoffman, MD
There will be a huge number needed to scan to pick up something actionable. And there will be a smaller number needed to harm. But we don't really know because these companies never do clinical trials.
The use of screening tests in asymptomatic people needs to be carefully researched to determine if it causes more harm than benefit--overall. No doubt incidental findings of cancer or aneurysms will be found and prevent untimely deaths. But this should be balanced against the knowledge that some incidental findings are harmless--but biopsies have their own risks. Seems we went through this decades ago with whole body CT scanners. Of course, MRI doesn't cause potential radiation injury as CT scans do, but it was the same issue then as it is now. Benefit has not been established.
As with anything (and everything), before diving into the minutiae of a test, the question each person will want to ask is "What is it for?" If you go searching in a forest for broken branches, you'll find them. It doesn't mean the forest isn't healthy, vibrant, and functioning.
If you're searching for something specific, or you have a list of considerations you want to talk about, like "My menstrual period is so painful, and I have a history of breast cancer, and I've had back pain for a long time," then this can be an (expensive) way to get more information.
But the information has to be triangulated with the presence of symptoms, the goal(s) at the forefront, and the wisdom of both the medical professionals as well as the wisdom of the body & person getting the scan done.
If the test is done to assuage anxiety or confirm that nothing is wrong, those are such broad goals and more likely to result in a long winding rabbit hole and (potentially) solving for problems that aren't problems.
Wider use of DEXA scans for people with metabolic syndrome would be a better use of point-in-time diagnostic imaging as an entre into treatment for obesity/diabetes/CVD, and as a test in the future to assess / reinforce progress and change. We will gain more from accumulating longitudinal data from cheaper non-invasive tests associated with chronic disease risk and feeding these results into systems to improve sensitivity and specificity of more focused predictive assays. Of course, that's not something health insurers, as the system is now configured, values -- predictive medicine, that is. But it's possible to imagine other financial frameworks, including life insurance, that will.
(Side note, some insurers' beginning to encourage liquid biopsy tests, though in a somewhat unfocused way, is progress in terms of insurer interest, but less promising in terms of the likelihood of unnecessary invasive biopsies. It would be better to underwrite the tests as a broader, longitudinal surveillance method to inform improvements in assay accuracy, especially for earlier detection.)