This Viewpoint is about the upper boundary of a 95% confidence interval, specifically the one around the reduction in total mortality observed in meta-analyses of randomized trials of ultrasound screening for abdominal aortic aneurysm (AAA) in men older than 65 years. Four such trials involving more than 100 000 men have been conducted and have together demonstrated a 50% reduction in AAA–related mortality,1 resulting in a recommendation to screen by the US Preventive Services Task Force (USPSTF). Abdominal aortic aneurysm–related deaths made up fewer than 3% of all deaths in these trials, so the 50% reduction in AAA-related deaths should correspond to a 1.5% reduction in total deaths, and this is almost exactly what was observed. My topic is whether this reduction can be considered statistically significant. Although this may sound like hairsplitting, it is how the question posed in the title is already being addressed.
Lederle FA. Does Abdominal Aortic Aneurysm Screening Save Lives?. JAMA Surg. 2016;151(8):697-698. doi:10.1001/jamasurg.2016.0044