Why randomized controlled trials do not always reflect reality.

Abstract

Randomized controlled trials (RCTs) constitute level I evidence and are used as the backbone of guidelines and recommendations for treatment. Although RCTs are theoretically the studies of choice for the assessment of the effectiveness of health care interventions, these trials (and their interpretation) may sometimes result in erroneous conclusions, erroneous therapeutic decisions, and incorrect recommendations. We aimed to check the applicability of the results of RCTs to everyday practice. We reviewed the literature for studies comparing the results of RCTs with observational or population-based studies in the field of vascular surgery, focusing on two specific topics: the results of carotid artery stenting vs carotid endarterectomy for the management of carotid artery stenosis; and the results of open surgical repair vs endovascular aneurysm repair for the management of ruptured abdominal aortic aneurysms. We found considerable discrepancy in the results of RCTs with real-life registries and observational studies in both topics. In the management of carotid artery stenosis, observational studies reported worse outcomes after carotid artery stenting compared with carotid endarterectomy. Regarding ruptured abdominal aortic aneurysms, population-based studies reported better results for endovascular aneurysm repair compared with open repair. In contrast, RCTs in both topics reported similar results for the two procedures. There is evidence that RCTs sometimes do not reflect clinical reality and are therefore potentially misleading to the reader. Every RCT has to be interpreted and applied carefully using complete available evidence and good clinical judgment.