Back

Trial-level factors affecting accrual rate of systemic sclerosis randomized clinical trials over 20 years.

RUSSO, B.; CHIRICA, I.-S.; COURVOISIER, D. S.; IUDICI, M.

2024-09-28 rheumatology
10.1101/2024.09.26.24314451 medRxiv
Show abstract

ObjectivesTo estimate the average time to complete patient enrollment and identify factors associated with accrual rates in systemic sclerosis (SSc) randomized controlled trials (RCTs). MethodsWe searched published SSc-RCTs indexed in PubMed from 2000 to 2024, selecting those with recruitment completed before the COVID-19 pandemic. We recorded key trial features (country, phase, randomization ratio, intervention, blinding, funding source, outcome type) and enrollment year(s). We measured enrollment duration and accrual rate (participants per month). A multivariable negative binomial generalized linear model was used to identify factors associated with accrual rate. ResultsWe included 80 studies, mostly single-country (75.0%) and industry-funded (57.5%), mainly recruiting in Europe (36.2%) and North America (22.5%). In 65% of studies, both limited and diffuse SSc patients were enrolled. The median sample size was 40.5 patients, with 20% of RCTs enrolling [≥]100 patients. The median recruitment time was 15 months (IQR 9.9 - 30.0), with a median accrual rate of 3.1 (IQR 1.6 - 5.5) participants per month. Recruitment rates varied over time, with faster accrual early in the 2000s and after 2012, and a slower period in between. Multivariable analysis showed that accrual rate was positively associated with skewed randomization, non-industry funding, international recruitment, and inclusion of both SSc subsets, especially compared to studies involving only dcSSc patients. ConclusionsRecruiting SSc patients for RCTs has been challenging, with generally slow accrual over the past 20 years and no significant improvement over time.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.