Getting Fitness Data into Clinical Records
For several decades, conventional wisdom about the amount of exercise the typical person needs to reach optimal health has stayed relatively constant at 30 minutes a day, five days a week. But a recent study in the Journal of the American College of Cardiology, covering more than 55,000 people over a mean 15-year period, discovered those who ran or jogged as little as five to 10 minutes a day also achieved substantial benefits--including 30 percent less risk of all-cause mortality and 45 percent less risk of cardiovascular-cause mortality.
The study's authors also concluded that running was as vital a prognostic indicator as smoking, obesity, and hypertension. The findings could revolutionize exercise science. Health Data Management spoke with one of the study's co-authors, Carl "Chip" Lavie, M.D., medical director of cardiac rehabilitation and prevention at the New Orleans-based Ochsner Health System, about the importance of the new data and how best to include activity data in clinical records.
HDM: You've been conducting studies on the effects of physical activity and fitness on overall health for many years. What is your impression of how far efforts to bring that data into charts formally have come?
CL: From the standpoint of the EHR, certainly what is lacking in probably almost all of them is information about physical activity. Every time I'm seeing a patient I have something in the beginning of my notes about physical activity, but because I put it there. It was not automatically in the record. When my patient gets put in the room, they have their height, weight, blood pressure, and BMI printed out. Their medications come out. But there is nothing about physical activity. I work with exercise as medicine. I don't know exactly all the behind the scenes work, but I know people are working on trying to get the big EHRs to include physical activity. It could be simple, it could be two questions. It could be a questionnaire, then you could code the questionnaire to have the kind of data you need for publications: "Are you physically active and if you are, how many times per week?" and that would be flagged. Then you could do all kinds of stuff. For example, if they say they're not active, a thing could pop up when you're doing your note and ask "Have you discussed physical activity recommendations?" Things like that could be done.
HDM: The benchmark recommendation for physical activity for years was really quite stable, more or less distilled to 150 minutes of moderate to vigorous activity a week. Your latest study showed that people who ran five to 10 minutes a day realized immense benefits. How revolutionary is this latest study?
CL: I think it's really great news for many people. You have to take a study like this for what it is. It's going to be a big study, it's gotten a lot of publicity, it'll get lots of citations. It'll be talked about for many years. It was a large study--55,000 people, 13,000 of them runners, with a 15-year average follow-up. The biggest finding, the strongest statistical point, is that the runners had a 30 percent lower all-cause lower mortality rate, a 45 percent cardiovascular-cause lower mortality rate, and a three-year average life extension. The biggest finding was when we divided the runners into quintiles, about 2,600 per quintile, those in the lowest quintile, who ran less than six miles per week, less than 52 minutes per week, and just one or two times per week, did just as well as those in quintiles 2-4, and the trend looked actually better than quintile 5. So it's suggesting you don't have to do a lot of running to get the maximum benefit. It can be misinterpreted--each one of those was done separately as a quintile. It wasn't like someone ran one time a week for five minutes--the bottom quintile which got the maximum benefit wasn't running a half mile per week, they were running six miles per week. But the point is that you didn't have to do three miles five times a week--if you ran two miles two or three times a week, you would still be getting maximum benefit. Doing more might burn more calories and you might enjoy it, but from an overall mortality and cardiovascular mortality standpoint, which are pretty important endpoints, it looks like you get the maximum benefit at quite low doses.
HDM: You champion the idea of including activity data in EHRs, but many, if not most, clinicians say they don't entirely trust patient-reported data--is that the biggest obstacle to including it?
CL: No. Why would that be different than getting people to answer whether they smoke and how much, or if they drink, how many drinks they have? We know we don't always get the accuracy. We know sometimes smokers tell us they're not smoking, and they're embarrassed to say they continue to smoke. But I believe the majority of patients tell the clinicians the truth. That could change if it was used against you. For example, in your workplace, if they ask you if you smoke, and if you would be charged a higher health premium if you smoke, you may not tell them the truth if you do. Or, if you were going to get a reduction in premium for exercising, you might tell them, "Sure, I follow the guidelines, 30 minutes five times a week." Now, for both of those, you can do a test. There's a urine test that can measure if somebody's smoking or not. And with exercise, there are a lot of ways to quantify that--the simplest way is with a pedometer. You can't make somebody do x number of steps, but you can give them incentives, the same kind of thing you give them for getting their annual physical.
Now, people cheat on all those things, but I do think from a health record standpoint it's usually pretty accurate information. Somebody who says they're smoking five cigarettes a day is maybe smoking eight to 10, and somebody who says they drink five drinks a day is probably drinking seven or eight. Now, a person who says they have a drink every three or four days is probably pretty accurate. It's the same for exercise. Usually the person, if they're not doing any exercise, will say "No, but I walk a lot," or something like that. And when you ask them for a little more detail you find they walk the typical amount everybody walks just to get through the day--everybody has to walk to their car.
HDM: Will the profusion of wearable devices with interconnect capabilities help the accuracy of this data, and will it help the clinical community accept this data more easily?
C: Let's put it this way. The reason for some of these things is there's a market for it. There are people who just love the gadgets. Some people are really extremely scientific about everything. From a simplicity standpoint, if somebody is trying to get 10,000 or 15,000 steps a day, an app could give them an alert at some pre-determined point of the day that maybe they were way behind pace on reaching their goal. I think reinforcement can certainly help people. Some people would be aided - it's been proven that pedometers increase physical activity - people kind of enjoy it, they compete with each other. A lot of people will walk a little more at work to meet their goal. Obviously, the person who is already doing a fair amount of exercise doesn't need it, but for such a high percentage of the population who are under-exercised, I think it's a very helpful thing.
HDM: In the abstract of the JACC study, you say reduction in mortality is related to continued running activity over time, and running is as important as such other prognostic variables like smoking, obesity, or hypertension. These other factors are in the record. How can we quantify running's mitigating effect on them if any of them are present?
CL: That's actually a pretty important point and goes into health maintenance. I have a website called obesityparadox.com, and my whole book is about this. It's a very important point so much is made of weight as a vital sign, and it's actually very misleading. For example, a 27.5 BMI is considered overweight. In January 2013, Katherine Flegal of the CDC published a meta-analysis of 97 studies that included 2.9 million people, and the BMI that had the best survival was in that range, 6 percent lower mortality than what is considered normal BMI. And, if somebody at that BMI put on 20 pounds that might move them into mild obesity, but that group had a 5 percent lower mortality than the normal BMI, though it wasn't statistically significant. Now, all obese people had a higher mortality, but that's because of higher mortality in Class 2 obesity, with a BMI above 35, and Class 3, above 40. But you can't say the overweight and mild obese BMIs are killing people if anything, they look like they have a better mortality rate. Steve Blair at the University of South Carolina has been publishing this kind of thing for 20 years, that based on the current data you can't say thinner is healthier, because the person who is thin but not physically active or fit has a tremendously high mortality, much higher than the heavier person. The heavier person who is fit is way better off than the thin person who is not fit. I very frequently say fat has been demonized by our society - fat is not the devil. Fitness is way more important than leanness in regard to long-term health.
HDM: How can we measure that baseline fitness for the clinical record? Periodic stress tests?
CL: I do think to demonstrate physical activity in ways like number of steps taken per day, workplaces will be adopting that type of thing. Medical people will adopt to having physical activity or exercise as a vital sign in their parameters. But stress tests will not be adopted--they can't be done cheaply enough. They provide very helpful prognostic information. The question is whether it's worth the several thousand dollars somebody paid for it. If a stress test could be done for $50, nobody would argue with it, and theoretically, a trainer could do a stress test for $50, but in our country you're not going to be able to get trainers to do medical level stress tests--every once in a while somebody's going to arrest and we have lawyers on every corner. That's not going to work here. There are too many false negatives and positives and the cost is too high.
We've actually recently published in JACC a paper on estimating fitness in simple ways by a simple physical activity questionnaire, your age, weight, BMI, and some simple things that can help in getting an estimate of fitness without doing a fitness test. You can have holes in that but it does correlate pretty well and is a good predictor of prognosis. Actually measuring fitness would be the best, and from a research standpoint, any study that has a measure of fitness is extremely helpful. But the problem of getting that in every individual is not possible from a cost standpoint.
HDM: How can your work feed into the creation of healthy communities? Are people asking you about this?
CL: Yes, they are. Corporate America is going to big health systems and asking them, "Can you make our workers more healthy?," not only to improve the productivity of their work force but also to reduce their health costs. Right now, all medical places are in a funny situation, in that most payment is still fee for service, and everybody's been saying the future is going to be a managed model, where a big system will manage the health of a large population. Right now they're using wellness as a marketing model, but they make all their money by taking care of sick people. When and if the model changes, obviously you realign your incentives. The problem is the whole system needs to get aligned in that direction. If you're actually getting paid to keep people healthy as opposed to being rewarded for doing bypass surgeries, it's a different motivation for the whole system.
HDM: Right now, a lot of hospitals and health systems sponsor road races organized by organizations like the YMCA or local running clubs. Could they go further and begin forming their own activities and clubs, keeping the data of things like who finished a road race, and offer incentives, financial or otherwise, to participants?
CL: Sure. And you could tune the incentives to all degrees and reward those who are more fit more generously. I am being a little facetious, but if you run a 5K under 35 minutes, you're demonstrating a pretty good level of fitness. If you're a runner, you know that from a competitive standpoint, a 34:30 5K isn't all that great, but that person is not unfit.