Layperson here. I found this a helpful sentence in articulating why all cause mortality is a valid endpoint in this discussion. Thank you.
“The attributable fraction of death from breast cancer is very low compared to cardiovascular disease and thus, using all-cause death as the bar for determining efficacy for a “comprehensive, advanced cardiovascular screening program” is appropriate.”
Very nice review.. Like most studies with multiple endpoints, DANCAVAS analyzed the endpoints separately. This fails to answer a global question of "did screened persons fare better?". One should recognize that some endpoints are worse than others, and consider re-analyzing the data to answer the question "was the worst event that happened to a screened person less bad than the worst event happening to a non-screened person?". This can be done with an ordinal outcome analysis. Secondly, the analysis by age group is not valid unless one thinks that magic happens on a specific birthday. Age should only be thought of as a continuous variable, and its effect assessed with a smooth interaction analysis. Finally, continuously scoring the severity of the results of the screening tests, rather than considering the screening result as all-or-nothing would have given some interesting results.
Here we go again with statins and saturated fat... ;) If you're going to screen people, it would work a little better if you have an effective intervention to use.
For example, if they had put the patients with high blood pressure or high cholesterol on a whole food diet / avoid processed food (or at least done some intensive nutritional counseling to that effect), there probably would have been significantly more net benefit versus statins and limiting saturated fat.
Perhaps this is an unduly broad assumption, but having been familiar with the culture of deliberate and widespread corruption of the Penn State School of Medicine, I cannot consider an “authority” from Cardiology to merit serious attention. The gravity of dishonesty which the institution engaged in is loathsome. I am sorry to consider Dr. Foy, whom I do not know, to be tarred with the same brush. He should take his academic associations wit less naïveté.
"There is evidence of treatment effect heterogeneity. Individuals <70 years experienced a 1.1% reduction in death compared to 0.2% increase for those ≥70, respectively."
I'm pretty sure the p on that difference between the two age subgroup HRs is not significant...
Maybe I’m missing it but how do they know that the control group didn’t receive these screening tests or therapeutic interventions as part of their routine care?
“Although further research is needed in certain areas, current research to date suggests that hawthorn may potentially represent a safe, effective, nontoxic agent in the treatment of CVD and ischemic heart disease (IHD).”
Poorly designed study with wrong screening tests and end points that don't correlate with the pathology being assessed.
1. PCI for stable CAD has not been proven to lower mortality (PCI for ACS or FFR driven PCI in stable CAD has mortality benefit). So, CAC by itself seems highly unlikely to correlate to mortality benefit.
2. Static EKG to identify AFib is nonsensical. As a screening tool (study volunteers are presumably asymptomatic), Static EKG would only identify chronic AFib. For paroxysmal occult AFib it can take 80+ days to identify occult Afib even with implantable cardiac monitoring. Also, the main benefit of identifying Afib is to prevent stroke (morbidity), not mortality (though some recent trials indicate mortality benefit with rhythm control also). Even with these limitations the study did identify statistical reduction in ischemic stroke.
3. All comer screening for attic aneurysm would not seem to have as much benefit as limiting screening to those with smoking history or family history of aortic aneurysm.
4) BP: Identifying hypertension is not the same as achieving BP control.
5. PAD: The main benefit of identifying lower extremity PAD is improvement in functional / exercise capacity and the correlation that 50% of PAD patients have CAD. Unless those with abnormal ABI also had aggressive CAD evaluation, it's unlikely to have any mortality benefit.
6) As with BP, identifying HgbA1c>6.5 is not the same as achieving control
7) Total Cool ≥309: Curious why they used such a high cutoff. One could argue that having such a high T. Cholesterol level at such an advanced age (and still no symptoms presumably) yields a subset of those who do not seem to have clinically significant atherosclerotic heart disease despite elevated Cholesterol levels. Again, identifying high cholesterol is not the same as achieving control.
8. Age group: 65-74 yo are those who have already outlived their life expectancy for those born at that time. Perhaps screening in 50-64 yo might have yielded different results.
Overall: the study uses screening tests already known to not correlate to mortality outcomes, does not emphasize achieving control of the target parameter assessed, and includes an older population for whom interventions may not change outcomes anyway (the underlying disease process may already be irreversible or too advanced).
The study should have assessed history of significant child abuse trauma! The statistic "population Attributable Risk" helps assess the potential impact of interventions on population health. The
PAF does not describe the proportion of patients exposed to the cited risk factor(s), the proportion of cases having any risk factor(s), the probability of causation for a specific disease, nor does its estimation enable epidemiologists to discriminate between those cases caused by, and those not caused by, the risk factor(s) under consideration. Nonetheless, the PAF statistic highlights the impact of a condition on a population and helps to concretize its importance. The CDC using PAFs adjusted for age, race/ethnicity, found the estimated overall percentage reduction associated with preventing all adverse childhood experiences would be, for example, 14.6% for stroke, 15.7% for kidney disease, and 5.7% for diabetes. PAF analysis revealed that CAN contributes to as much as 12.6% (or two million cases annually in the US) of all coronary artery disease.
In so far as the PAF is an epidemiologic abstraction and does not aid in diagnosis or treatment, its significance is clarified by looking at the PAFs for coronary artery disease due to the contributions of the several known treatable conditions evaluated in this study: for treatment adjusted blood pressure equal or greater than 130 mmHg, 28%, for treatment adjusted non-HDL cholesterol equal or greater than 130 mg/dl, 17%, for smoking, 9.8%, and for diabetes 9.6%.
Thus, in fact, the presence or maldistribution, of the confounding factor of child abuse trauma may obscure the impact (or lack thereof) of the interventions in the Danish study.
Thank you Dr. Foy. This study should get more press. It shows that aggressive diagnosis and treatment makes no difference. Of course it won’t change anyone’s use of aggressive diagnosis or treatment and they will continue loving statins and blood thinners.
Perhaps they are looking and evaluating this study all wrong. Maybe the screening is good but the treatment options just don’t work. Perhaps better treatment options other than the standard treatment of dyslipidemia would work better.
Maybe next time use an interaction to lower metabolic syndrome abnormalities as well as lower overall excess dysfunctional inflammation and see if there is a mortality benefit.
Layperson here. I found this a helpful sentence in articulating why all cause mortality is a valid endpoint in this discussion. Thank you.
“The attributable fraction of death from breast cancer is very low compared to cardiovascular disease and thus, using all-cause death as the bar for determining efficacy for a “comprehensive, advanced cardiovascular screening program” is appropriate.”
Very nice review.. Like most studies with multiple endpoints, DANCAVAS analyzed the endpoints separately. This fails to answer a global question of "did screened persons fare better?". One should recognize that some endpoints are worse than others, and consider re-analyzing the data to answer the question "was the worst event that happened to a screened person less bad than the worst event happening to a non-screened person?". This can be done with an ordinal outcome analysis. Secondly, the analysis by age group is not valid unless one thinks that magic happens on a specific birthday. Age should only be thought of as a continuous variable, and its effect assessed with a smooth interaction analysis. Finally, continuously scoring the severity of the results of the screening tests, rather than considering the screening result as all-or-nothing would have given some interesting results.
Here we go again with statins and saturated fat... ;) If you're going to screen people, it would work a little better if you have an effective intervention to use.
For example, if they had put the patients with high blood pressure or high cholesterol on a whole food diet / avoid processed food (or at least done some intensive nutritional counseling to that effect), there probably would have been significantly more net benefit versus statins and limiting saturated fat.
Perhaps this is an unduly broad assumption, but having been familiar with the culture of deliberate and widespread corruption of the Penn State School of Medicine, I cannot consider an “authority” from Cardiology to merit serious attention. The gravity of dishonesty which the institution engaged in is loathsome. I am sorry to consider Dr. Foy, whom I do not know, to be tarred with the same brush. He should take his academic associations wit less naïveté.
"There is evidence of treatment effect heterogeneity. Individuals <70 years experienced a 1.1% reduction in death compared to 0.2% increase for those ≥70, respectively."
I'm pretty sure the p on that difference between the two age subgroup HRs is not significant...
Maybe I’m missing it but how do they know that the control group didn’t receive these screening tests or therapeutic interventions as part of their routine care?
HAWTHORN!
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3249900/
“Although further research is needed in certain areas, current research to date suggests that hawthorn may potentially represent a safe, effective, nontoxic agent in the treatment of CVD and ischemic heart disease (IHD).”
Poorly designed study with wrong screening tests and end points that don't correlate with the pathology being assessed.
1. PCI for stable CAD has not been proven to lower mortality (PCI for ACS or FFR driven PCI in stable CAD has mortality benefit). So, CAC by itself seems highly unlikely to correlate to mortality benefit.
2. Static EKG to identify AFib is nonsensical. As a screening tool (study volunteers are presumably asymptomatic), Static EKG would only identify chronic AFib. For paroxysmal occult AFib it can take 80+ days to identify occult Afib even with implantable cardiac monitoring. Also, the main benefit of identifying Afib is to prevent stroke (morbidity), not mortality (though some recent trials indicate mortality benefit with rhythm control also). Even with these limitations the study did identify statistical reduction in ischemic stroke.
3. All comer screening for attic aneurysm would not seem to have as much benefit as limiting screening to those with smoking history or family history of aortic aneurysm.
4) BP: Identifying hypertension is not the same as achieving BP control.
5. PAD: The main benefit of identifying lower extremity PAD is improvement in functional / exercise capacity and the correlation that 50% of PAD patients have CAD. Unless those with abnormal ABI also had aggressive CAD evaluation, it's unlikely to have any mortality benefit.
6) As with BP, identifying HgbA1c>6.5 is not the same as achieving control
7) Total Cool ≥309: Curious why they used such a high cutoff. One could argue that having such a high T. Cholesterol level at such an advanced age (and still no symptoms presumably) yields a subset of those who do not seem to have clinically significant atherosclerotic heart disease despite elevated Cholesterol levels. Again, identifying high cholesterol is not the same as achieving control.
8. Age group: 65-74 yo are those who have already outlived their life expectancy for those born at that time. Perhaps screening in 50-64 yo might have yielded different results.
Overall: the study uses screening tests already known to not correlate to mortality outcomes, does not emphasize achieving control of the target parameter assessed, and includes an older population for whom interventions may not change outcomes anyway (the underlying disease process may already be irreversible or too advanced).
What is your TAU screening? Do you recommend CAC in healthy asymptomatic 69 yo male? Why won’t Medicare cover CAC?
The study should have assessed history of significant child abuse trauma! The statistic "population Attributable Risk" helps assess the potential impact of interventions on population health. The
PAF does not describe the proportion of patients exposed to the cited risk factor(s), the proportion of cases having any risk factor(s), the probability of causation for a specific disease, nor does its estimation enable epidemiologists to discriminate between those cases caused by, and those not caused by, the risk factor(s) under consideration. Nonetheless, the PAF statistic highlights the impact of a condition on a population and helps to concretize its importance. The CDC using PAFs adjusted for age, race/ethnicity, found the estimated overall percentage reduction associated with preventing all adverse childhood experiences would be, for example, 14.6% for stroke, 15.7% for kidney disease, and 5.7% for diabetes. PAF analysis revealed that CAN contributes to as much as 12.6% (or two million cases annually in the US) of all coronary artery disease.
In so far as the PAF is an epidemiologic abstraction and does not aid in diagnosis or treatment, its significance is clarified by looking at the PAFs for coronary artery disease due to the contributions of the several known treatable conditions evaluated in this study: for treatment adjusted blood pressure equal or greater than 130 mmHg, 28%, for treatment adjusted non-HDL cholesterol equal or greater than 130 mg/dl, 17%, for smoking, 9.8%, and for diabetes 9.6%.
Thus, in fact, the presence or maldistribution, of the confounding factor of child abuse trauma may obscure the impact (or lack thereof) of the interventions in the Danish study.
Thank you Dr. Foy. This study should get more press. It shows that aggressive diagnosis and treatment makes no difference. Of course it won’t change anyone’s use of aggressive diagnosis or treatment and they will continue loving statins and blood thinners.
I would be interested by the follow up in 10 years. CT scans are likely to cause cancers and according to Gofman, coronary disease.
https://ratical.org/radiation/CNR/RMP/
Perhaps they are looking and evaluating this study all wrong. Maybe the screening is good but the treatment options just don’t work. Perhaps better treatment options other than the standard treatment of dyslipidemia would work better.
Maybe next time use an interaction to lower metabolic syndrome abnormalities as well as lower overall excess dysfunctional inflammation and see if there is a mortality benefit.
Statins and lowering saturated fat?