Investigation of the Relationship between Clinical Trial Design of Randomized Controlled Trials and Scientific Influence of the Trial Outcome in Non-small Cell Lung Cancer

Abstract

The results of clinical trials are essential for the advancement of evidence-based medicine (EBM). Our previous research using the Relative Citation Ratio (RCR), article-based citation metric developed by the National Institutes of Health (NIH), identified that the outcomes of clinical trials designed to have a high evidence level, such as large sized randomized trials, and those with novel interventions have higher scientific influence. In the present study, aiming to further investigate the factors that improve the scientific influence of clinical study results, we focused on randomized controlled trial (RCT) and evaluated the relationship between study design and RCR of drug intervention RCTs in patients with non-small cell lung cancer (NSCLC). Publications of drug intervention RCTs for NSCLC patients published between 2007 and 2016 were included. Clinical trial design factors were compared among three RCR categories with 50 trials in each: lowest RCR (LOW50), median RCR (50 NIH percentile [50NIH％ile]), and highest RCR (TOP50) group. The numbers of Phase 3, multi-country, for-profit company sponsored/supported, and statistically powered RCTs, which confirmed pre-defined differences between/among interventions, increased by more than 30％ from 50NIH％ile to TOP50. The number of novel interventions such as EGFR-TKI and Immune checkpoint inhibitors were also increased as RCR increased. Our results indicate quantitatively that the clinical trial design is an important factor that affects scientific influence of the RCT outcome. RCTs should be sufficiently large to generate statistically confirmatory outcomes, be conducted in multiple countries, and use newer drugs in order to better contribute to the progress of EBM.