Assessing Evidence in Pain Medicine: Clarifications and Misconceptions

Abstract

Dear Editor: We appreciate the interest shown by Boswell et al. in my articles on how to use evidence in pain practice [1,2] and the opportunity to clarify how some of the systematic reviews I discussed were assessed. As described in the articles, a review of facet joint interventions by Boswell et al. applied its exclusion criteria inconsistently [3]. The methods state, “To be included in this review, ideally, all studies should have documented the existence of spinal pain of facet origin using controlled diagnostic facet joint or nerve blocks. Due to the scarcity of such studies, some studies with single blocks were also considered”[4]. Boswell et al. excluded a negative randomized trial of radiofrequency by Leclaire et al. [5] because it used a single uncontrolled diagnostic block (based on a positive response in the following week), yet included a positive trial by van Kleef et al. [6] that also used an uncontrolled block (based on a positive response 30 minutes after the injection). The application of post hoc exclusion criteria related to the timing of the positive response to an uncontrolled diagnostic block was not specified in the methods section, appears arbitrary, and is not appropriate in a systematic review. Either …