Gary Wolf

March 25, 2019

Last October, the physician researcher Richard Kravitz published an article in the Journal of the American Medical Association (JAMA) reporting the results of a study investigating whether people being treated for pain would benefit from using N-of-1 experiments to tailor their treatments. The patients were randomly assigned to groups whose treatment was tailored using N-of-1 experiments and those who weren’t. At six months, there was no significant difference between the two groups in terms of how much pain interfered in daily activities.

The commentary introducing the article described N-of-1 methods in clinical care as “another instance of a beautiful idea being vanquished by cruel and ugly evidence.” The comment is all the more powerful in being authored by two physicians who have worked hard to bring these methods into wider use, including Gordon Guyatt, one of the true pioneers of the field.

I met with Dr. Kravitz before the article was published but after the results were known, and we talked a bit about how to interpret it. We both share the conviction that reasoning using evidence is a useful way to make decisions, but in this case a well considered method of experimentation did not yield additional benefit to patients. One possibility that I asked him about: is it possible that the various pain treatments didn’t produce outcomes different enough for the quality of the decision making to matter very much? (If you are choosing between two things that produce substantially the same results, then it doesn’t matter very much how good you are at choosing.) He wasn’t certain about this. He was still thinking about what it all meant, and I’ve been thinking about it, too.

This month, the article produced a number of thoughtful letters by other physicians and researchers who have been working on N-of-1 methods. I thought one of them, by Dr. Alexander Smith, would be of particular interest to readers of this blog, because it contains a first person report of an experiment that worked. Dr. Smith found the trigger of his throat pain by eliminating a likely culprit from his diet, noting that the pain disappeared, and then reintroducing the offending food and noticing that the pain came back. This simple protocol substituted for a much more difficult process that is typically recommended, saving him a lot of time, stress, and money.

It’s very interesting to note how different this kind of N-of-1 experiment is from the kind that Dr. Kravitz was so carefully investigating. The pain-treatment experiments in Dr. Kravitz’ study were meant to explore systematic benefit of using N-of-1 methods in a group of patients. The experiment Dr. Smith described is meant to discover the effect of an intervention on a single person, who can chose to stop the experiment when a desirable level of confidence is reached. The first approach has the power to produce generally applicable knowledge and justify policy decisions; the second is highly useful in everyday life but is merely exploratory and anecdotal by clinical research standards. Is it possible that the phrase “N-of-1” contains two very different kinds of practices?

Dr. Smith gave me permission to publish the entire text of the commentary describing his experiment. I’ve added a couple of definitions of medical terms in square brackets.

Several months ago, I began experiencing dysphagia [difficulty swallowing] and solid food impaction. Findings from an esophagogastroduodenoscopy (EGD) with biopsy confirmed eosinophilic esophagitis [inflammatory disease of the esophagus]. My gastroenterologist recommended the 6-food elimination diet to eliminate milk products, wheat, soy, eggs, seafood, or nuts from my diet for 6 to 8 weeks until my symptoms resolved, then add them back 2 at a time until symptoms recurred.

This diet seemed draconian. Cruel even. And nearly impossible. I quickly reviewed the literature and found the 6-food elimination diet is indeed standard of care.(1) Compliance with the 6-food elimination diet in real-world settings is understandably low. Some recommend repeat EGDs at each step of the elimination diet, for up to 5 EGDs per patient, at an immense burden to patients and expense to the health care system.(2)

My review also revealed that the most common food trigger was dairy.(3) This seemed a perfect situation for an unblinded n-of-1 trial. The condition is chronic. Response to food elimination varies by individual. The symptoms are experienced near daily. And if I was going to eliminate a food from my diet for the rest of my life, I wanted to know for certain that it was the cause of my symptoms.

I eliminated all milk products from my diet for 8 weeks. My symptoms resolved. I then ate bread pudding with ice cream, sweet butter on my corn, and milk in my cereal. My symptoms recurred. I am now off dairy, symptom free, and convinced I know my food trigger. No further EGDs.

At the end of my trial, I read the article by Kravitz and colleagues that demonstrated a lack of effectiveness of patients randomized to n-of-1 trials compared with usual care for treatment of chronic musculoskeletal pain.(4) This was a laudable negative study that advanced the science of n-of-1 trials. Still, my experience suggests there may yet be a role for n-of-1 trials. My hope is that the study by Kravitz and colleagues spurs other researchers and clinicians not to abandon n-of-1 trials, but rather animates us to think creatively about scenarios in which n-of-1 trials might simplify treatment regimens, improve patient compliance, and reduce health care costs.

1. Liacouras CA, Furuta GT, Hirano I, et al. Eosinophilic esophagitis: updated consensus recommendations for children and adults. J Allergy Clin Immunol. 2011;128(1):3-20.e6. doi:10.1016/j.jaci.2011.02.040
2. Health facts for you:the six food elimination diet for eosinophilic esophagitis. University of Wisconsin Health. https://www.uwhealth.org/ healthfacts/nutrition/553.pdf. Published 2017. Accessed September 18, 2018.
3. Teoh T, Mil lC, Chan E, Zimmer P, Avinashi V. Liberalized versus strict cow’s milk elimination for the treatment of children with eosinophilic esophagitis [published online July 4, 2018]. Can J Gastroenterol. doi:10.1093/jcag/gwy030
4. Kravitz RL, Schmid CH, Marois M, et al. Effect of mobile device–supported single-patient multi-crossover trials on treatment of chronic musculoskeletal pain: a randomized clinical trial. JAMA Intern Med. 2018;178(10):1368-1377. doi:10.1001/jamainternmed.2018.3981