Predicting CVD in People at the Center of Risk
Hi. I am Dr. Henry Black, Clinical Professor of Internal Medicine at the New York University School of Medicine and former President of the American Society of Hypertension. As we develop newer and newer tests to define risk, it is becoming very important to determine which tests are worth it, which are not, and what we learn by adding new tests to what we think we already know. We have had this discussion with respect to imaging studies and regarding biomarkers, such as high-sensitivity C-reactive protein (hs-CRP).
Two studies recently looked at this issue. One is the MESA study in the United States with about 6800 individuals from across the country, including people of different races and ethnicities, who had a baseline evaluation done around the year 2000. We have a 7.5-year follow-up on strokes, heart attacks, and other events and which tests predicted those.
The other study comes primarily from Europe, with about 250,000 people included from a variety of observational studies and many of the same factors measured. The European study focused on hs-CRP and fibrinogen, but also looked at white blood cell count and albumin.
This is important because right now, we are pretty sure what to do with low-risk individuals: people whose Framingham Heart Score is less than 1% per year. We are also pretty sure what to do with high-risk individuals, whose risk is defined as more than 2% per year. What about the people with intermediate risk? Many of us, especially as we age, will fall into that intermediate risk category.
The MESA study looked at carotid intima-media thickness (IMT); coronary artery calcium; and a whole variety of other things, including family history. (Family history is not really part of the original Framingham Heart Score but was used in the Reynolds Heart Score, which has pretty much similar findings.) The findings are extremely interesting. The investigators had to do some adjustment because they had a 7.5-year follow-up rather than the 10-year follow up we had with Framingham. The MESA study included 1330 individuals who fell into the intermediate-risk category defined by Framingham.
Particularly interesting is that all the factors they looked at -- and there were 6 of them -- had increased sensitivity and picked out people who had events. The best predictor was coronary artery calcium. Flow-mediated vasodilation did not add much information, nor did other imaging tests and biomarkers.
The investigators found that coronary artery calcium up-titrated people about 25% of the time, meaning that the patients went from intermediate to low risk for down-titration or from intermediate to high risk for up-titration. This has very important implications because if someone is high risk, we tend to feel comfortable giving them statins, aspirin, and blood pressure medication, and if someone is low risk, we think we can get by with lifestyle modifications.
Coronary artery calcium was very valuable relative to everything else the investigators looked at. Family history was next. hs-CRP was not very helpful. They concluded that if we wanted to assess predictive factors, coronary artery calcium would be extremely useful.
There is a downside, as usual. In all, 23 million Americans are thought to be intermediate risk, and if we were to go and look at coronary artery calcium, we would probably expose those people to radiation risks, and we might produce additional carcinomas that would not have developed otherwise. Thus, we have to balance risks vs benefits -- not to mention cost, equipment, and salaries -- relative to simply drawing blood for a test, which is pretty safe and pretty useful.
In the European study, hs-CRP was a better predictor than fibrinogen or albumin, but it still did not really help much. Of course, that is a very inexpensive test with virtually no risk. In the European study, the number needed to screen using hs-CRP was about 440 individuals. We would perform 439 unnecessary hs-CRP tests; it was 490 for fibrinogen. Clearly, hs-CRP is better as far as that goes.
What does the clinician do when faced with an individual who wants to know whether or not he or she is at risk? What tests should you do? Who will pay for them? How many things are you going to find when you screen that are false-positives or that are not helpful? How many false-negatives will result?
This is an evolving field. I want to congratulate the American College of Radiology (ACR), which looked at the number of imaging tests we do that are not useful. It turns out to be quite significant. As we get to a cost-effectiveness era, we will have to take this into account. The ACR estimated that 5% of our national healthcare budget involves imaging tests that are not useful. That is a very large number.
Stay tuned. This is going to be a very interesting development as we go forward. Thank you.
Hi. I am Dr. Henry Black, Clinical Professor of Internal Medicine at the New York University School of Medicine and former President of the American Society of Hypertension. As we develop newer and newer tests to define risk, it is becoming very important to determine which tests are worth it, which are not, and what we learn by adding new tests to what we think we already know. We have had this discussion with respect to imaging studies and regarding biomarkers, such as high-sensitivity C-reactive protein (hs-CRP).
Two studies recently looked at this issue. One is the MESA study in the United States with about 6800 individuals from across the country, including people of different races and ethnicities, who had a baseline evaluation done around the year 2000. We have a 7.5-year follow-up on strokes, heart attacks, and other events and which tests predicted those.
The other study comes primarily from Europe, with about 250,000 people included from a variety of observational studies and many of the same factors measured. The European study focused on hs-CRP and fibrinogen, but also looked at white blood cell count and albumin.
This is important because right now, we are pretty sure what to do with low-risk individuals: people whose Framingham Heart Score is less than 1% per year. We are also pretty sure what to do with high-risk individuals, whose risk is defined as more than 2% per year. What about the people with intermediate risk? Many of us, especially as we age, will fall into that intermediate risk category.
The MESA study looked at carotid intima-media thickness (IMT); coronary artery calcium; and a whole variety of other things, including family history. (Family history is not really part of the original Framingham Heart Score but was used in the Reynolds Heart Score, which has pretty much similar findings.) The findings are extremely interesting. The investigators had to do some adjustment because they had a 7.5-year follow-up rather than the 10-year follow up we had with Framingham. The MESA study included 1330 individuals who fell into the intermediate-risk category defined by Framingham.
Particularly interesting is that all the factors they looked at -- and there were 6 of them -- had increased sensitivity and picked out people who had events. The best predictor was coronary artery calcium. Flow-mediated vasodilation did not add much information, nor did other imaging tests and biomarkers.
The investigators found that coronary artery calcium up-titrated people about 25% of the time, meaning that the patients went from intermediate to low risk for down-titration or from intermediate to high risk for up-titration. This has very important implications because if someone is high risk, we tend to feel comfortable giving them statins, aspirin, and blood pressure medication, and if someone is low risk, we think we can get by with lifestyle modifications.
Coronary artery calcium was very valuable relative to everything else the investigators looked at. Family history was next. hs-CRP was not very helpful. They concluded that if we wanted to assess predictive factors, coronary artery calcium would be extremely useful.
There is a downside, as usual. In all, 23 million Americans are thought to be intermediate risk, and if we were to go and look at coronary artery calcium, we would probably expose those people to radiation risks, and we might produce additional carcinomas that would not have developed otherwise. Thus, we have to balance risks vs benefits -- not to mention cost, equipment, and salaries -- relative to simply drawing blood for a test, which is pretty safe and pretty useful.
In the European study, hs-CRP was a better predictor than fibrinogen or albumin, but it still did not really help much. Of course, that is a very inexpensive test with virtually no risk. In the European study, the number needed to screen using hs-CRP was about 440 individuals. We would perform 439 unnecessary hs-CRP tests; it was 490 for fibrinogen. Clearly, hs-CRP is better as far as that goes.
What does the clinician do when faced with an individual who wants to know whether or not he or she is at risk? What tests should you do? Who will pay for them? How many things are you going to find when you screen that are false-positives or that are not helpful? How many false-negatives will result?
This is an evolving field. I want to congratulate the American College of Radiology (ACR), which looked at the number of imaging tests we do that are not useful. It turns out to be quite significant. As we get to a cost-effectiveness era, we will have to take this into account. The ACR estimated that 5% of our national healthcare budget involves imaging tests that are not useful. That is a very large number.
Stay tuned. This is going to be a very interesting development as we go forward. Thank you.