RÉSUMÉ
OBJECTIVE: To assess the proportion, the reasons, and the factors associated with the discontinuation or nonpublication of randomized controlled trials (RCTs) on connective tissue diseases (CTDs). METHODS: We searched all interventional RCTs on CTDs registered in ClinicalTrials.gov since 2000. Two reviewers selected studies according to prespecified eligibility criteria. Completion status, publication status, and reported reasons for discontinuation or nonpublication were retrieved on ClinicalTrials.gov, through literature search, and by contacting investigators. Multivariable logistic regression was used to identify factors associated with study noncompletion and nonpublication. RESULTS: We included 175 studies, mostly phase III, placebo-controlled trials on pharmacologic treatments recruiting patients with systemic lupus erythematosus (51%), systemic sclerosis (20%), Sjögren's syndrome (12%), or other CTDs. Fifty-eight (33%) had been discontinued, mainly for insufficient patient accrual, with no differences in discontinuation rates across the CTDs (P > 0.5). Forty-six (35%) of 130 studies having included at least 1 patient were unpublished, and 86 (65%) were published in a peer-reviewed journal after a median of 24 months (interquartile range 15-41) from completion, with a significantly higher publication rate in completed versus discontinued studies (81% versus 22%; P < 0.001). We were able to obtain reasons for nonpublication in one-third of cases. Small sample size (<100 participants) was the only factor associated with study noncompletion and nonpublication. CONCLUSION: One of 3 registered RCTs on CTDs fails to be completed or published. This represents a waste of resources and raises ethical concerns regarding hidden clinical data and unfruitful participation by patients.