Precision, Health and the Real World
I attended the American College of Sports Medicine’s annual conference this summer. It was a great experience overall and I learned a lot. I enjoy seeing the state of new research tremendously but one of the things that I have found most useful from attending conferences over the years is the process of evaluation I do with my beliefs and practices to see how they match with what is around me. For decades, I’ve debated with myself and others the appropriate levels of precision for health and physiological information when it is to be used in the real world. Clearly there is a time and place for precision but is it universally needed?
This question became something I wanted to write about after attending a particular panel at this conference. The session was focused on the effect of medication timing and exercise prescription. In particular, the research focused on how time of day and medication status influenced exercise zones for patients undergoing cardiac rehabilitation. There a variety of commonly used cardiac medications that have an influence on heart rate, the most prevalent being a class of medications called beta blockers. These are some of the most commonly prescribed medications in the United States overall and are a near default for people who have had a heart attack or undergone bypass surgery. Beta blockers typically lower both resting and exercising heart rates.
Many people find that beta blockers significantly change how exercise feels and a lot of that is in the reduction of heart rate, particularly as exercise intensity increases. Many people who take beta blockers note that as effort increases during exercise, their heart rate does not keep up, and eventually just won’t allow for further raising of exercise intensity. Given this reality, the overall point of this session that we should analyze and better understand patient dosing of medications to influence their response to exercise made intellectual sense to me. If people took their medications close to when they exercised, the influence of these medications would potentially be stronger than those who took their medications many hours earlier. These differences might influence how people experienced their exercise and what their bodies did during that exercise. So far, so good.
When the analysis part of the session got going, I was continually surprised by the level of precision that the presenters brought to the topic. I fully get that academic research can’t be done without some precision. That said, the data was presented to the tenth of a decimal place for heart rates, i.e. comparing 123.2 with 126.6 for example. I found this frustrating and my annoyance was further exasperated when they compared the patients’ levels of perceived exertion at that level of precision as well. I would not have been so put out if they were using a realistic measure of perceived exertion. They were not. As is common in some academic physiology research, the presenters used the classic Borg scale of perceived exertion. This scale is a numeric, single digit scale, from 6 to 20 with 6 being at rest and 20 being maximal exertion. This scale is hard for users, I have been around the field for decades and could not readily distinguish many of these values while exercising even though I know the definitions. This scale has been validated, I full well know but I really just don’t think it’s a usable tool for the real world. In my mind, it’s artificially specific. The intellectual foundation of the scale is quite limited to boot. In short, this is a scale used by tradition or for a level of academic validity that has limited real world value at the level of the whole number, let alone at the level of decimal places.
I want to make clear, I did not write this post to complain about this session or the attempt to measure and discuss things with specificity and accuracy. My complaint and issue comes when we discuss aspects of human health or physical functioning with a high level of precision that has no perceptible difference in the real world. We owe It to ourselves and our clients (patients) to be realistic about how much precision we present. Details matter and there are things when small differences are critical. There is also a lot of times when we present and discuss variations that mean nothing in the real world. Let’s be precise when we need to and simultaneously acknowledge the limits of our analysis and our understanding. It’s ok for a mean value of 11.28 to be rounded to “about 10” in many circumstances. The real world is a messy place and by using our data analysis and recommendations with that in mind, everyone is the better for it. Imprecision is not always a bad thing, sometimes it’s the perfect Goldilocks response – just right.