Levels of Evidence Supporting American College of Cardiology/American Heart Association and European Society of Cardiology Guidelines, 2008-2018

Abstract

IMPORTANCE Clinical decisions are ideally based on evidence generated from multiple randomized controlled trials (RCTs) evaluating clinical outcomes, but historically, few clinical guideline recommendations have been based entirely on this type of evidence. OBJECTIVE To determine the class and level of evidence (LOE) supporting current major cardiovascular society guideline recommendations, and changes in LOE over time.

Filling the glass of evidence‐based medicine

“…11 In this system, virtually all recommendations associated with low or very low levels of evidence are graded weak (or conditional), unless there are clear concerns about major harm. Second, guideline developers should strive to minimise the high numbers of consensus recommendations, which, according to a recent review of cardiovascular guidelines published over a 10-year period, persistently account for 40% or more of all recommendations, 12 of which a significant minority are proven incorrect when finally evaluated in clinical trials. Comparing consecutive versions of 11 US cardiology guidelines published 6 years apart, among 448 index 'should do' recommendations with known level of evidence, 90% (95/105) of those supported by level 1 evidence were retained in later versions versus 74% (109/148) of those supported by expert opinion 13 a near threefold increase in the likelihood of being downgraded, reversed or omitted if based on little evidence.…”

Research in Vascular Medicine: Where We Are and Where We Are Going

“…One disadvantage of this rapid technological progress is the difficulty of achieving a high level of evidence, as currently defined in the medical community (14). Multilevel disruptive innovations typically imply frequent changes and rapid improvements of technologies, which limit their evaluation in multiple randomized controlled trials.…”

Response to Commentary by Giuseppe Mancia

“…

We appreciate the commentary by Prof. Mancia on our manuscript [1].He argues that randomized trials are not feasible to certain areas, particularly epidemiology, diagnosis, and patients' follow-up [2]. We agree that in the current climate, randomized trials of such topics are not feasible.However, all observational studies are subject to confounding; there are numerous examples in the literature of observational studies or expert opinion being contradicted by the results of well-conducted randomized clinical trials [3,4].

…”