Your Patient Is at 21% Risk for Something -- Now What?

Henry R. Black, MD; Andrew J. Vickers


June 03, 2011

This feature requires the newest version of Flash. You can download it here.

Henry A. Black, MD: Hi. I'm Dr. Henry Black, Clinical Professor of Internal Medicine at the New York University School of Medicine and immediate past President of the American Society of Hypertension. I'm here with my friend and colleague Dr. Andrew Vickers from Sloan-Kettering talking about biostatistics in the 21st century. We're way beyond chi squares now. To understand what to do, we need help from our biostatistical colleagues.

How do we risk stratify most effectively? If we knew all the things that increased risk and all the things that were protective, we would throw the book at the people with risk and not expose people at low risk to any of the risks of the treatment.

Andrew J. Vickers, DPhil: Right.

Dr. Black: We are not very good at that, however, although we are getting better as we go along.

Dr. Vickers: We are not very good at it because we just ignore it, right? At the moment, we tend to treat. The analogy I always use is if we read the news like we practiced medicine, we'd say things like, "The Yankees played the Red Sox and both teams scored more than 4 runs." We wouldn't actually give the result. We tend to say, "Well, you have hypertension. And so do you."

Dr. Black: So this is noninferiority.

Dr. Vickers: Right -- you give the numbers. Cardiovascular disease was really the pioneer in this because of the Framingham model. It said, "Well, we'll take all these things like your lipids and your age and whether you smoke and so on, and we'll calculate your risk. If you're at high risk, we'll treat you. If you're low risk, we won't."

I wish people would do more of that because what you often find is people go in to the doctor. They are totally healthy 45-year-old guys. They work out. They don't smoke. They're thin and they're in great shape, but their cholesterol is raised. So we say, "We'd better put you on a statin to get your cholesterol down." That's not treating risk. That's treating a risk factor.

Dr. Black: Risk factors, for sure.

Dr. Vickers: The reason we do that is that 20, or 30, or 40 years ago we had to. We didn't have desktop risk prediction models. We had to put people in binary categories. That way of practicing medicine has sort of carried forward into the technological era.

Dr. Black: I have an issue with Framingham, not necessarily with this idea.

Dr. Vickers: Sure.

Dr. Black: This could get dangerous if you have a 20% 10-year risk based on the Framingham score, which doesn't include a lot of obesity issues or diabetes or inflammation, but just what they measured, and risk of 19.9% means you don't treat. Or 10.9% means you don't treat. Or what could be even worse is that 19.9% means that you are not going to get reimbursed for treating; whereas, 20.1% means you are reimbursed, and you do treat. So you're taking a continuous variable and making it into a categorical variable.

Dr. Vickers: Right. The nice thing about risk, though, is that you can have a discussion about it. You can alter it with respect to patient benefit. So, for example, you have a patient who is at 21% risk and starts taking the drug. The patient says, "This is making me feel terrible. I'm sick all the time. Do I really have to take it?" You might say, "Well, maybe not."

Another patient comes in who is at 50% risk. You say, "Look, you're going to take this drug because you're at just such high risk." Now, a decision analyst can tell you that 20% risk means that the disease is 4 times worse than the treatment. At a 10% risk threshold, you're saying the disease is 9 times worse than the treatment.

So you can actually have rational discussions about the appropriate cut-point and how that should vary according to patient preference. Now if we are talking about something like hypertension, you would just use a cut point of 140, right? You're taking a continuous measure like blood pressure and cutting it. You could have exactly the same payment concerns. "We won't reimburse you if it's 139. We will reimburse if it's 141."

But you can't have a rational debate about it. "Why 140?" "Well, I think 138 and you think 140." Or a patient is intolerant to medication and you think, well, how high of a blood pressure is high enough? It's much less easy to have rational conversations when you turn that into risk.

Dr. Black: Dr. Vickers, thank you very much for your input. It's so valuable right now, when we're trying to know what to do with at patient and what to tell a patient and what a patient is supposed to interpret from what they read, that we really understand how this happens.

Dr. Vickers: Well, thank you.

Dr. Black: We really appreciate your time.

Dr. Vickers: It was fun to be here.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: