161 posters,  34 sessions,  7 topics,  1004 authors 

ePostersLive® by SciGen® Technologies S.A. All rights reserved.

4047
The Big Bamboozle: Bending the Truth in Interventional Pain Trials
Session: MP-02a
Thurs, Nov. 16, 10 am-12 pm
Hampton Room

Please note, medically challenging cases are removed three months after the meeting and scientific abstracts after three years.

Poster Presenter
Authors
Affiliations

Rate

No votes yet

1. INTRODUCTION

•Reporting bias (aka “spin”) occurs when unfavorable information is presented in a positive light.
•These distortions in reporting may misrepresent findings, especially for non-significant outcomes.
•Two common pain interventions, epidural steroid injections (ESI) and facet joint interventions (FJI), were examined
 

2. METHODS

•MEDLINE, EMBASE, and CENTRAL databases for RCTs on ESI and FJI published between 2006-2017 were systematically reviewed
•Two authors independently performed study selection and data extraction
•1788 articles identified, 112 articles retrieved for full text review, 30 articles analyzed.
•ESI trials were also rated using a technical quality rating scale (AQUARIUS)
 

3. RESULTS

•More than half of the trials recommended treatment for clinical use, failed to mention a non-significant primary outcome, framed results in unequivocal terms, and failed to recommend confirmatory trials
•No association of spin with AQUARIUS high risk of bias
•Discrepancies in published trial protocols were identified, often containing inaccurate or misleading information
 

SPIN TYPE

1. Focusing on within group rather than between group changes

2. Emphasis on statistically significant secondary analyses

3. Emphasizing beneficial treatment effect of a non-significant outcome

4. Interpreting non-significant primary results as showing treatment equivalence

5. Concluding safety or lack of harm based on non-significant differences in adverse events between groups

 

4. DISCUSSION

•We identified spin as the misrepresentation of the primary outcome in most trials in this systematic review of RCTs for two common pain interventions
•Bias in interventional pain studies may be more prevalent than other fields due to high number of non-significant primary outcomes, which may encourage spin as a way to promote findings from a non-significant trial
•Limitations: only analyzed studies published in English, only including trials with placebo, possibility of analytical bias, and analysis of only a certain subset of biases
Enter Poster ID (e.gGoNextPreviousCurrent