Chatellier, ‘Feasibility Study of N-of-1 Trials With Blood Pressure
Hypertension, 25 (2): 294 – Hypertension
I measure blood pressure at home. Unfortunately, it is easy to become bored with this procedure, and neglect it. In fact, it is more fun to wonder why measuring blood pressure is so boring than actually measuring blood pressure, so of course that’s what I’ve been spending some of my time on lately. My guess is that part of the problem is that home blood pressure measurements vary a lot. I’ve had single sessions in which my systolic ranged 11 points and my diastolic 16 points. This measurement range is larger than the likely effect of any intervention I’m going to be making. Therefore, a single measurement session doesn’t give me the feeling that I’m adding any information. It’s frustrating and stupid. Damn measurements.
Of course a good way to track measurements with a lot of random error is to use a moving average. So here’s the question: how many blood pressure measurements does it take to get results that accurate enough to discern the effects of treatment? Here is a graph from a paper published in Hypertension that suggests an answer. I won’t break down the method here. There is a link to the paper at the bottom of the graph and you can explore it for yourself. The quick version is that researchers compared the difference between two series of measurements taken at home, varying the number of measurements in the series, and watched the difference decrease as the number of measurements went up. The graph shows a nice, smooth decrease in variation. You achieve 80% of the total drop in variation after 15 measurements.
In other words, if you take three measurements per day, you can get a decent baseline for blood pressure experiments in five days. This seems like good news.