There is a ton of interesting acupuncture research being conducted these days, and there is no doubt that we all will benefit from this work, practitioners and patients alike. Our aims are the same – bringing about healing. As I’ve been thinking about the actual research itself, though, I’ve become aware of a kind of disconnect between medical researchers and the patients they recruit for their work. These researchers are most interested in “Does treatment/drug X have a measurable effect on symptom Y?” And as a matter of good research protocol, they have to remain as neutral as possible about the outcome. That is to say, they can’t bias their work by hoping for one result or another.
Double blind research protocol
More importantly, some sort of randomized trial has to be set up and a placebo chosen. In the case of acupuncture studies, some type of “sham acupuncture” is employed (either inserting needles at random points disregarding established meridians, or only pretending to insert needles). Controls such as these are needed to ascertain whether the main treatment variable is the one responsible for any positive effects that might ensue. And then, in an attempt to remove bias from the gathering of data, a double-blind situation is set up whereby neither the doctor nor the patient is aware of which group he/she is in. (A real hurdle, for obvious reasons – the acupuncturist knows what group the patient is in, but at least the recording of results can be performed by someone who is unaware of treatment group assignments.)
A conundrum – two different views of sample size
The medical researcher is interested in the statistical results of the test. Does the treatment group differ significantly from the placebo group(s)? Statistics are based on samples of patients, and the larger the sample, the better. Sample sizes are referred top as “N,” so a sample size of 100 patients would be N = 100. The patients, on the other hand, are interested in their own personal health. Each is a sample of one. Hence the N = 1 in the title of this blog. How does the perspective of N = 1 relate to the perspective of the N = 100? Here’s where it gets interesting.
Lots of studies have failed to establish a statistically significant difference between treatment and control groups. So the researchers must conclude that the treatment is not effective. But there may be many individual patients who experienced significant individual benefits! It’s just that the significant individual benefits in the treatment group did not exceed the benefits of the control group.
The researcher, in a sense, has to dismiss the treatment’s effectiveness as being no better than the placebo’s. But each patient that responded positively has to be delighted, right? Nobody ever claims that their symptoms were less severe or that their recovery was less real. From their perspective the treatment (even if it was the placebo, and I’ll have more to say about placebos in a later blog) was successful.
What this means to me as a practitioner
As I treat patients, I have to be aware of what research has revealed about the kind of medicine I practice, and it certainly informs my treatment options, but every patient I see is a sample size of one. And what that means is that I look for unique, individual avenues to explore, and these avenues are not constrained by statistical evidence. I find this liberating in a way because it allows me to be creative in my practice; but also because it reminds me that my goals and my patients’ are the same – healing. And it really doesn’t matter if the exact same treatment given to another patient would not have the same result.
Significant individual benefits are not dependent on statistical significance in a clinical setting. I’ll have to talk to my family doctor the next chance I get and see if she has come to this same conclusion.