Correspondence

Reporting of Trials of Gabapentin

N Engl J Med 2010; 362:1641-1642April 29, 2010DOI: 10.1056/NEJMc1000964

Article

To the Editor:

In the Special Article on outcome reporting in industry-sponsored trials of gabapentin for off-label use, Vedula et al. (Nov. 12 issue)1 claim to have identified unacceptable reporting practices in certain clinical trials.

Our study on gabapentin for pain due to traumatic nerve injury2 was included as an example of disagreement between the definition of the primary outcome in the protocol and that in the published report. We strongly disagree.

New primary outcomes were not introduced in our published report. The primary-outcome variable (i.e., the mean visual-analogue score of the last week of each of the two treatment periods adjusted for the baseline [a run-in period for period 1 and a washout period for period 2]) was identical in both the protocol and the published report.

Protocol-specified primary outcomes were not relegated to secondary outcomes in our published report. The statement in the Supplementary Appendix that accompanied the article by Vedula and colleagues (available at NEJM.org) that tactile allodynia, cold allodynia, and pinprick hyperalgesia were primary variables but were reported as secondary outcomes in the published report is incorrect. They were not outcome variables but were baseline data only. These variables were measured at the start of the study but were not measured again after intervention; therefore, they could not possibly serve as outcome variables.

Our study, originally proposed by Scandinavian pain researchers to Parke-Davis, showed that, as compared with placebo, gabapentin in doses of up to 2400 mg per day had no effect on neuropathic pain (P=0.20).

Torsten E. Gordh, M.D., Ph.D.
Uppsala University, Uppsala, Sweden
torsten.gordh@akademiska.se

Troels S. Jensen, M.D., Ph.D.
Danish Pain Research Center, Aarhus, Denmark

Eija Kalso, M.D., Ph.D.
University of Helsinki, Helsinki, Finland

No potential conflict of interest relevant to this letter was reported.

2 References

1. 1

   Vedula SS, Bero L, Scherer RW, Dickersin K. Outcome reporting in industry-sponsored trials of gabapentin for off-label use. N Engl J Med 2009;361:1963-1971
   Free Full Text | Web of Science | Medline

2. 2

   Gordh TE, Stubhaug A, Jensen TS, et al. Gabapentin in traumatic nerve injury pain: a randomized, double-blind, placebo-controlled, cross-over, multi-center study. Pain 2008;138:255-266
   CrossRef | Web of Science | Medline

Author/Editor Response

The concerns of Gordh et al. reflect in part the complexity of the data that we presented and analyzed in our article. Although we applied standardized rules for disagreement among the statements of primary outcome in the protocol, research report, and published report,1 it is possible that other researchers might have interpreted the same information differently. Statistical language must be precise, and it is possible that the differences we observed were attributable to imprecision of language. For the study by Gordh et al., the protocol stated that “The primary efficacy variable will be the mean pain intensity score (VAS) during the last week of each treatment period.” The protocol stated elsewhere that the score was “adjusted for the corresponding mean during the last week of baseline before the start of titration.” The description of the primary outcome in the published report did not clearly reflect the timing of the primary-outcome measurement or whether the analysis approach was part of the definition of outcome. In this study, as indicated in our article, the use of different descriptions for the primary outcome did not lead to a change in the conclusion of efficacy, since findings from the analysis of covariance were not statistically significant (P=0.20). The relevant pages from the protocol and the full publication for study 945-271 are in the Supplementary Appendix, available with the full text of this letter at NEJM.org.

Similarly, for the purposes of our study, we decided that ancillary studies would be defined as part of the main trial, and then we applied the rule consistently to all included trials. We agree that the rules we used may have obscured some of the details of the study by Gordh et al.

S. Swaroop Vedula, M.D., M.P.H.
Kay Dickersin, Ph.D.
Johns Hopkins Bloomberg School of Public Health, Baltimore, MD
svedula@jhsph.edu

Since publication of their article, the authors report no further potential conflict of interest.

1 Reference

1. 1

   Gordh TE, Stubhaug A, Jensen TS, et al. Gabapentin in traumatic nerve injury pain: a randomized, double-blind, placebo-controlled, cross-over, multi-center study. Pain 2008;138:255-266
   CrossRef | Web of Science | Medline

Citing Articles (1)

Citing Articles

1. 1

   Kerry Dwan, Douglas G Altman, Lynne Cresswell, Michaela Blundell, Carrol L Gamble, Paula R Williamson, Kerry Dwan. . 2011.