Undisclosed Changes in Outcomes in Randomized Controlled Trials: An Observational Study

To the editor:

We applaud the efforts of Ewart et al in bringing attention to undisclosed changes in outcomes between trial registration and the publication of results. However, we disagree with their assessment of our article, Tolterodine and Tamsulosin for Treatment of Men With Lower Urinary Tract Symptoms and Overactive Bladder A Randomized Controlled Trial (JAMA 2006; 296:2319-28).1

Ewart et al are correct that 3 patient-reported outcomes were listed as primary outcomes in the trial registry, and that only data for one of these were reported as a primary endpoint in the primary manuscript. We believed that these outcomes, despite the fact that they were statistically significant in support of our hypothesis, would be of less interest to the JAMA audience than more quantitative diary outcomes. Analyses of the other 2 patient-reported outcomes revealed statistically significant treatment effects that supported our hypothesis. These data were published in a subsequent article;2 they were not identified as secondary outcomes.

Ewart et al also state that many secondary outcomes that were listed in the registry were not reported in the primary manuscript, including “several clinically important secondary measures,” identified as “continence, overactive bladder, and erectile function.” However, changes in the rate of urgency urinary incontinence and every other defining symptom of overactive bladder were clearly reported in the primary manuscript.3 Change in erectile function was an exploratory endpoint in this study.4 It is not an outcome commonly included in studies on OAB or LUTS treatment, nor was it central to the hypothesis of the study. Several other registered outcomes were reported in subsequent papers.2,5 Some secondary outcomes were not published because of conceptual overlap, or because they did not logically “fit” in any of the manuscripts (eg, erectile function). Given journal word and figure/table limits, as well as reader stamina, there are constraints on the number of outcomes that can reasonably be included in a manuscript, particularly with 4 treatment groups and outcomes assessed at multiple time points. We firmly believe that anyone with expertise in the therapeutic area would agree that the most pertinent endpoints were published, regardless of whether there were statistically significant treatment effects.

Lastly, Ewart et al report that 2 secondary outcomes, postvoid residual volume and maximum flow rate, included in the publication were not listed in the registry. At the time this trial was registered in 2005, not all outcomes were required to be listed in the registry.

Again, we agree with Ewart et al that complete transparency in the performance and reporting of clinical trials is vital. While their efforts are laudable, we strongly disagree with their classification of our article as a case where changes were made in outcomes between registration and publication.

Steven A. Kaplan, MD kaplans@med.cornell.edu Weill Cornell Medical College New York, NY
Claus G. Roehrborn, MD University of Texas Southwestern Medical Center Dallas, TX
Eric S. Rovner, MD Medical University of South Carolina Charleston, SC
Zhonghong Guan, MD, PhD Pfizer Inc New York, NY

1. Kaplan SA, Roehrborn CG, Rovner ES, et al. Tolterodine and tamsulosin for treatment of men with lower urinary tract symptoms and overactive bladder: a randomized controlled trial. JAMA. 2006;296:2319-2328.
2. Rovner ES, Kreder K, Sussman DO, et al. Effect of tolterodine extended release with or without tamsulosin on measures of urgency and patient reported outcomes in men with lower urinary tract symptoms. J Urol. 2008;180:1034-1041.
3. Abrams P, Cardozo L, Fall M, et al. The standardisation of terminology of lower urinary tract function: report from the Standardisation Sub-committee of the International Continence Society. Neurourol Urodyn. 2002;21:167-178.
4. Kohler TS, McVary KT. The Relationship between Erectile Dysfunction and Lower Urinary Tract Symptoms and the Role of Phosphodiesterase Type 5 Inhibitors. Eur Urol. 2008.
5. Kaplan SA, Roehrborn CG, Chancellor M, et al. Extended-release tolterodine with or without tamsulosin in men with lower urinary tract symptoms and overactive bladder: effects on urinary symptoms assessed by the International Prostate Symptom Score. BJU Int. 2008;102:1133-1139.

Competing interests:   Steven A. Kaplan has served as a consultant for Pfizer Inc, Astellas, and Allergan.
Claus G. Roehrborn has served as a consultant to Pfizer Inc, AMS, spectrum, Aeterna Zentaris, Lilly, GSK, and Sanofi Aventis, and an investigator for AMS, Aeterna Zentaris, and GSK.
Eric S. Rovner is an investigator, consultant, meeting participant, and advisory board member for Pfizer Inc.
Zhonghong Guan is an employee of Pfizer Inc.