|| Checking for direct PDF access through Ovid
Introduction: There was a 20 years span between the landmark NINDS r-tPA trials and the recent spate of positive endovascular trials of acute ischemic stroke. The evolution of clinical trial design, characteristics of subjects, and nature of background standard care have not been properly studied. Such information could have consequences for future clinical trial designs.Objective: To systematically search the literature for randomized controlled trials of acute ischemic stroke and identify key patterns.Methods: We interrogated Pubmed from 1995 to 2015 with MeSH terms (Fig 1A). Inclusion criteria are: 1) acute intervention delivered within 24 hours of onset; 2) randomized trial with a control (either placebo or standard care) group with ≥20 subjects in each arm; 3) mRS used as an outcome at 3 months; 4) published in the English language. Data were extracted and effect sizes were calculated for both mortality and favorable outcome (mRS: 0-1). We also extracted data for baseline characteristics and analyzed trends over time.Results: Forty-six clinical trials were included with 19951 subjects and 38.5% of them were females. There were 8 phase II and 11 phase III trials, nine out of 46 are positive trials. Average number of subjects per trial was 443 and number of patient enrollments per year ranged from 23-270. Study-wise odd’s ratio ranged 0.15 - 3.97 for mortality and 0.30- 3.06 for favorable outcomes. Rate of mortality was 13.6% (0-44%) and favorable outcomes ranged 6-37% in control and 7.5-74.8% in intervention arm. Median onset of treatment ranged from 90 - 972 minutes. There are trends of increased NIHSS score at baseline for studies over the years. The type of intervention and baseline NIHSS are major determinants of mortality rate.Conclusions: Patterns existed in the acute ischemic stroke treatment trials. Understanding these patterns may inform the stroke community to better design trials in the future.