Chronic pain treatment preferences change following participation in N-of-1 trials, but not always in the expected direction

Richard L Kravitz,Maria Marois,Ida Sim,Deborah Ward,Samika S Kanekar,Allison Yu,Peach Dounias,Jiabei Yang,Youdan Wang,Christopher H Schmid

doi:10.1016/j.jclinepi.2021.08.007

Abstract

ObjectiveTo examine pain treatment preferences before and after participation in an N-of-1 trial. Study Design and SettingIn this observational study nested within a randomized trial, we examined chronic pain patients’ preferences before and after treatment in relation to N-of-1 trial results; assessed the influence of different schemes for defining comparative “superiority” on potential conclusions; and generated classification trees illustrating the relationship between pre-treatment preferences, N-of-1 trial results, and post-treatment preferences. ResultsTreatment preferences differed pre- and post-trial for 40% of participants. The proportion of patients whose N-of-1 trials demonstrated “superiority” of one treatment regimen over the other varied depending on how superiority was defined and ranged from 24% (using criteria that required statistically significant differences between regimens) to 62% (when relying only on differences in point estimates). Regardless of criteria for declaring treatment superiority, nearly three-fourths of patients with equivocal N-of-1 trial results nevertheless expressed definite preferences post-trial. ConclusionA large segment of patients undergoing N-of-1 trials for chronic pain altered their treatment preferences. However, the direction of preference change did not necessarily correspond to the N-of-1 results. More research is needed to understand how patients use N-of-1 trial results, why preferences are “sticky” even in the face of personalized data, and how patients and clinicians might be educated to use N-of-1 trial results more informatively.

Highlights

Shared decision-making (SDM) about treatment rests on the assumption that clinicians have medical expertise and patients have first-hand knowledge of their own preferences and values
We considered patients to have derived positive information value [17] from their N-of-1 trial if their initial treatment preference was retained after trial confirmation, reversed after trial refutation, or changed from uncertain to a preference consistent with the results of the trial
The proportion of participants who had definite post-trial treatment preferences despite equivocal N-of-1 trial results was consistently ≥74% (Table 4). (Detailed tabulations of the relationship between N-of-1 trial results and final treatment preferences are provided in the e-Appendix B, Tables B1B4). In this secondary analysis of a randomized controlled trial, we studied 87 chronic pain patients who completed N-of-1 trials comparing two treatment regimens

Summary

Introduction

Shared decision-making (SDM) about treatment rests on the assumption that clinicians have medical expertise and patients have first-hand knowledge of their own preferences and values. Clinical evidence may be thin, and patients’ views are shaped by multiple influences, including past experience, advice of previous providers, opinions of family and friends, mass media, social media, and direct-toconsumer advertising [2]. The foundations of SDM are shakier than proponents might hope. This could explain, in part, why high-fidelity implementation of SDM is observed infrequently in practice [3]. Clinical practice guidelines derived from such evidence may not adequately account for variations in patients’ social circumstances, comorbidities, or values [7]. Evidence from personal experience can be collected and analyzed along a spectrum of rigor, ranging from informal trial and error, to careful logging of clinical data, to performance of single-patient (“N-of-1”) trials

Methods

Results

Conclusion