Document Type


Degree Name



Department of Public Health and Preventive Medicine


Oregon Health & Science University


Background: Early trials of tricyclic antidepressants showed efficacy against neuropathic pain compared with placebo. Later trials of anticonvulsants demonstrated less efficacy than antidepressants, but a greater placebo effect. Head-to-head trials comparing pharmacotherapies for neuropathic pain are limited. Indirect comparisons from placebo-controlled trials could provide information about comparative effectiveness. However, differences in placebo response rates could affect the reliability of indirect comparisons. Objectives: To improve the interpretation of trial evidence through better understanding of the placebo response in neuropathic pain trials by determining study-specific predictors of the placebo response and study-specific predictors of pain reduction attributable to treatment after controlling for the placebo response. Methods: Meta-regression techniques were used to identify study-specific predictors of the placebo response as well as predictors of pain relief while controlling for the placebo response in pharmacotherapy trials. Data from an existing systematic review on treatments for neuropathic pain were used to identify qualifying randomized, placebo-controlled trials, with updated searches to identify published and unpublished studies current through March 2014. The primary outcome variable was whether ≥50 percent reduction in pain from baseline to study endpoint had occurred. Predictor variables included study design characteristics (e.g., parallel trial, size of trial, length of trial), drug characteristics (e.g., drug category, flexibility of dosing), population characteristics (e.g., diagnosis, baseline pain level, gender distribution), and other time and space variables (e.g., when the study was published or completed, geographic location of study).

Sensitivity analyses involving only parallel group studies, studies of pregabalin, and studies of only painful diabetic peripheral neuropathy were conducted. In addition, the predictors for proportion of patients with ≥30 percent pain reduction were compared with the predictors for ≥50 percent pain reduction. Results: Thirty-nine placebo-controlled trials, analyzed as 40 studies, reported the primary outcome (one study randomized patients to two included medications in addition to placebo and was analyzed as two studies). Dosing flexibility, baseline pain levels, gender distribution, and whether or not patients had painful diabetic neuropathy were included in the final model predicting the proportion of patients reporting at least a 50 percent pain reduction in the placebo group. After controlling for the placebo response, length of treatment on the study drug was inversely correlated with treatment effect. In sensitivity analyses, length of treatment accounted for 100 percent of the between-study variance in published and unpublished studies, parallel group trials, studies enrolling only patients with painful diabetic neuropathy, trials of pregabalin, and studies in which the outcome variable was ≥30 percent pain reduction. Conclusion: The placebo response varies in trials of neuropathic pain. Standardization of trial design by incorporating flexibility of drug dosing, duration of treatment at least 12 weeks, including a single type of neuropathic pain, and stratifying results by gender may facilitate interpretation and generalizability of trial results. Systematic evidence reviews and meta-analyses should also incorporate consideration of flexibility of dosing, length of treatment, type of neuropathic pain, gender distribution within trials, and baseline levels of pain when pooling studies or comparing drugs across trials with the placebo group as the common comparator.





To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.