The size of circles is proportional to the number of studies. Arrows indicate collaborations between countries or regions.
Dechartres A, Riveros C, Harroch M, Faber T, Ravaud P. Characteristics and Public Availability of Results of Clinical Trials on Rare Diseases Registered at Clinicaltrials.gov. JAMA Intern Med. 2016;176(4):556-558. doi:10.1001/jamainternmed.2016.0137
As an obligation to participants and to ensure evidence-based decision making, results of trials must be reported. Rare diseases are defined by a prevalence of fewer than 200 000 people affected in the United States,1,2 but overall they may affect up to 25 million people in North America.2 The limited number of patients for each rare disease makes research in this field challenging2 and availability of results even more crucial.3 We aimed to describe the characteristics of completed phase 3 or 4 trials of rare diseases registered at clinicaltrials.gov and to assess whether their results were publicly available.
On March 27, 2015, we searched clinicaltrials.gov for the topic “rare diseases,” which is based on the list from the Office of Rare Diseases of the US National Institutes of Health; we downloaded records for all studies registered for each rare disease listed. We selected all completed phase 3 and 4 trials assessing drugs with a primary completion date between January 2009 and December 2013. For each trial, we systematically assessed whether results were (1) posted at clinicaltrials.gov, (2) published in journals by searching MEDLINE via PubMed and Google Scholar, and (3) publicly available (ie, posted at clinicaltrials.gov or published in journals). The Kaplan-Meier method was used to assess cumulative incidence over time from the primary completion date up to July 16, 2015. We also collected whether the trial was likely subject to the US Food and Drug Administration Amendments Act (FDAAA) requiring posting of results within 1 year of trial completion based on an algorithm developed by the US National Library of Medicine.
We found 1531 rare diseases with at least 1 study registered at clinicaltrials.gov and identified 802 completed phase 3 or 4 trials assessing drugs. Their characteristics are described in the Table. Briefly, 251 trials (31%) were single-group interventional trials, 552 (69%) were reported to be randomized, and 426 (53%) were open-label trials. In all, 500 trials (62%) were conducted in a single country and 312 (39%) were single-center trials. There were 270 trials (34%) conducted in the United States and 284 (35%) in Europe (Figure).
At 36 months, results were posted at clinicaltrials.gov for 35% of trials (95% CI, 32%-38%), published in journals for 47% (95% CI, 43%-51%), and publicly available for 63% (95% CI, 59%-66%).
Among the 325 trials likely subject to the FDAAA, 220 (68%) had results posted at clinicaltrials.gov: 4% (95% CI, 2%-6%) had results posted at 12 months and 57% (95% CI, 51%-63%) at 36 months. Overall, 98 of 466 trials (21%) not likely subject to the FDAAA had results posted at clinicaltrials.gov.
We found studies registered at clinicaltrials.gov for 1531 rare diseases, which is a small number compared with the 7000 rare diseases that have been described.1 Many studies were conducted in a single center, whereas the limited number of patients makes collaboration between centers necessary. Many also studied a single group or were not randomized, as previously reported,4 which can compromise the validity of results.
Although we found a greater proportion of trials with results posted than did previous studies,5,6 many trials are still not compliant with the FDAAA. Potential explanations may include lack of knowledge of the legislation, uncertainty regarding the applicable trials, and lack of deadline reminders. In contrast, 21% of studies not likely subject to the FDAAA had results posted, which suggests that some researchers are prone to share results for transparency.
Our study has limitations. It is only representative of trials on rare diseases registered at clinicaltrials.gov. The algorithm used to identify studies subject to FDAAA regulation may have misclassified some studies. Finally, trial characteristics are submitted by sponsors or principal investigators and thus may be subject to error.
Corresponding Author: Agnes Dechartres, MD, PhD, Centre de Recherche Epidémiologie et Statistique, INSERM U1153, 1 place du Parvis Notre Dame, 75004 Paris, France (email@example.com).
Published Online: March 7, 2016. doi:10.1001/jamainternmed.2016.0137.
Author Contributions: Dr Dechartres had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Dechartres, Ravaud.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Dechartres.
Critical revision of the manuscript for important intellectual content: Riveros, Harroch, Faber, Ravaud.
Statistical analysis: Dechartres.
Study supervision: Dechartres, Ravaud.
Conflict of Interest Disclosures: None reported.
Additional Contributions: Elise Diard, Centre de Recherche Epidémiologie et Statistique, INSERM U1153, French Cochrane Center, provided help with figures. Mickael Randrianandrasana, BSc, Centre d’Epidemiologie Clinique, Hôpital Hôtel-Dieu, provided help with data management. Jacques Li, MD, Centre d’Epidemiologie Clinique, Centre de Recherche Epidémiologie et Statistique, INSERM U1153, developed a method to conduct repeated actions on the Internet. Ignacio Atal, MSc, French Cochrane Center, Centre de Recherche Epidémiologie et Statistique, INSERM U1153, provided supplementary data such as study location. None of these additional contributors received any compensation for their roles in the study.