The fate of prospective spine studies registered on www.ClinicalTrials.gov

Background context: There has been concern expressed about research ethics with respect to not fully reporting data collected during clinical studies. One site available for all clinical trials is ClinicalTrials.gov. The original purpose of this site was to facilitate patients seeking a trial for the treatment of their particular condition. The internationally available site offers general information about the study, sponsor name, principal investigator, patient selection criteria, enrollment goal, study design, outcome measures, participating centers, initiation date, date posted, date completed, and other pertinent data. The site can be used to identify studies conducted for a particular condition or intervention.

Purpose: The purpose of this study was to investigate the fate of spine-related studies registered on www.ClinicalTrials.gov, with particular focus on the publication rate of completed trials.

Study design/setting: Analysis and classification of clinical studies posted on an international research registry Web page and literature search for related publications.

Patient sample: Not applicable.

Outcome measures: The primary outcome measure was publication of the study registered on ClinicalTrials.gov.

Methods: Multiple searches were conducted on ClinicalTrials.gov Web site to identify studies related to commonly treated spinal conditions, including herniated disc, degenerative disc disease, stenosis, and spondylolisthesis. Studies related to tumors, fractures, or that included nonspine conditions were not included. For studies classified as completed more than 18 months before this review, literature searches were conducted to determine if the results of the study had been published and factors related to publication. The author has no financial conflict related to this work.

Results: There were 263 spine-related studies identified from searches on the ClinicalTrials.gov site. Data on the site had the studies classified as follows: 72 completed, 70 active, not recruiting (generally indicates collecting follow-up data), 74 recruiting, 11 recruiting by invitation, 13 not yet recruiting, 18 terminated, 4 withdrawn, and 1 suspended. Among the 72 studies indicated to be completed, 28 (38.9%) have been published. The mean time to publish was 27.9 months from the date of completion. Among unpublished studies, the mean length of time from study completion to the preparation of this article was 62.0 months. There was no difference in the likelihood of publication based on the geographic region of study origin or whether the study was registered before or after initiation. There were statistically significant relationships between the publication rate and the funding type as well as the research type (p<.05) with industrial-funded studies and those evaluating devices having a lower publication rate and those that were funded by a federal agency and comparing surgery to nonoperative care had the highest publication rates.

Conclusions: Although the 38.9% publication rate for spine-related studies found in this study appears low, it is in line with other studies reporting a 22.8% publication rate for arthroplasty trials and 43.2% for orthopedic trauma trials. In addition to ClinicalTrials.gov Web site fulfilling its original goal of providing patients information about clinical studies, it can also provide a means of tracking publication of prospective studies, changes to protocols, matching publication content to posted study design, and others and raise queries concerning the reasons for not publishing what appear to be well-designed studies. The posting of spine studies before initiation can increase transparency and ability to evaluate clinical trials in spine.