Structured Abstract

Objectives

To assess the impact of including studies identified from ClinicalTrials.gov on the conclusions and strength of evidence (SOE) grading from an ongoing systematic review of treatments for diabetic peripheral neuropathy symptoms.

Data sources

We searched ClinicalTrials.gov through March 2016 to identify trial records. Peer-reviewed publications were identified from an ongoing systematic review of treatments for diabetic peripheral neuropathy symptoms.

Review methods

Two independent reviewers screened ClinicalTrials.gov records for randomized controlled trials evaluating treatments for diabetic peripheral neuropathy symptoms. We matched ClinicalTrials.gov records to publications. Two reviewers extracted data from ClinicalTrials.gov records. We compared conclusions and SOE grade with and without ClinicalTrials.gov records for pain and quality of life, conducting sensitivity analyses where possible.

Results

We identified 53 studies from ClinicalTrials.gov (46 completed, 3 recruiting, 2 withdrawn, 2 with unknown status). 37% of the completed trials posted results. We compared 25 ClinicalTrials.gov records with 25 matched publications. These differed in the number enrolled (8 studies, 32%), the primary outcome (14 studies, 56%), and adverse event reporting (2 of 10 studies with posted results, 20%). Pooled results of published trials showed greater effectiveness of pregabalin than placebo at reducing pain, but pooled results of unpublished trials were not statistically significant. Otherwise, ClinicalTrials.gov was mostly useful in confirming suspected reporting biases and did not meaningfully change either the overall conclusions or the SOE grading.

Conclusions

Researchers conducting systematic reviews should account for reporting bias in their analyses. But, until outcomes data are more consistently reported, the usefulness of searching ClinicalTrials.gov is limited.

Citation

Wilson LM, Sharma R, Dy SM, Robinson KA. Supplemental Project to Assess the Transparency of Reporting Requirements for Studies Evaluating the Effectiveness of Treatment Options for Symptoms of Diabetic Peripheral Neuropathy. Methods Research Report. (Prepared by the Johns Hopkins University Evidence-based Practice Center under Contract No.290-2015-00006-I). AHRQ Publication No. 17-EHC006-EF. Rockville, MD: Agency for Healthcare Research and Quality. July 2017. doi: https://doi.org/10.23970/AHRQEPCMETH1.

Journal Publications

Wilson LM, Sharma R, Dy SM, et al. Searching ClinicalTrials.gov did not change the conclusions of a systematic review. J Clin Epidemiol. 2017 Jul 28. Epub ahead of print. DOI: https://dx.doi.org/10.1016/j.jclinepi.2017.07.009.