Abstract
There has been significant progress over the last two decades in translational research and with resultant strides in the management of axial spondyloarthritis; 1 undoubtedly, challenges remain with the unmet needs for future endeavours. 2 Over the same period, however, there has been in parallel an increasing amount of literature highlighting the significance of discontinuation and nonpublication of randomised controlled trials (RCTs) and its inherent bias.3–5 Subpar quality of conducting and nonreporting of trials both contribute to wastage of health care resources and to a distortion of the evidenced-based landscape, ultimately jeopardising patient safety. 6
Ankylosing spondylitis (AS), a specific and severe subset of axial spondyloarthritis, was found to be one of the top three topics of spinal trials registered at ClinicalTrials.gov. To investigate whether there is resource wastage in AS research, we systematically analysed the discontinuation and nonpublication rates of relevant trials.
All completed RCTs on the topic of AS, registered on ClinicalTrials.gov between 1 January 2013 and 31 December 2020, were ascertained. Data as per progress status, topic of interest, source of funding, start and completion date, intervention type, number of participating centres, and publication status were extracted.
From a total of 21 relevant randomised controlled trials, five were discontinued; the discontinuation rate was 24%. The leading causes of discontinuation were slow recruitment, sponsor’s decision, drug failure to achieve key endpoints, and termination of the development programme.
The nonpublication rate was 48% (10 of 21). The majority of these were completed but unpublished (80% of the subgroup and 38% of total group). Our search only prevails a cross-sectional perspective of the publication outcome. However, the most recently completed trial included in our data was in August 2020, which is beyond 12 months from the point we ran the search on the trial registry. Twelve months is an important time window for publication as it is the FDA’s recommend period for publication outcome. Unfortunately, only 2 of 11 publications were published within this recommended time frame. Industry funding status did not reach statistical significance, with 71% (15 of 21) studies being industry funded; three of those (3 of 15, P = 0.52) were discontinued and six (6/15, P = 0.27; 40% of the subgroup & 29% of the total group) remained unpublished. The reason for the high nonpublication rate remains unreported.
The discontinuation (24%) and nonpublication (48%) rates of RCTs in AS research remain worrisome both from an ethical and financial perspective. Both underreporting and nonpublication adversely affect efforts in evidence synthesis and can compromise clinical guideline development. We support a call for greater transparency behind the nonpublication of completed trials; for improved adherence to reporting guidelines; and further identification of strategies to increase patient recruitment, in order to protect research resource capacity and to prevent financial wastage and bias in clinical research.
