Randomised controlled trials (RCTs) have the potential to change clinical practice. Nevertheless, concerns persist regarding the interpretation of statistical analyses, particularly hypothesis testing. Over-reliance on the ubiquitous 0.05 significance threshold has led researchers to make potentially inappropriate statements about their findings. A critical issue is the erroneous ‘acceptance’ of the null hypothesis instead of correctly stating a ‘failure to reject’ when results are statistically non-significant.1 While statistical reporting practices have been explored in cardiology research,2 there has been limited focus on inappropriate interpretations of hypothesis testing. Therefore, we aimed to examine their occurrence among five major cardiology journals.
We searched Embase.com for primary reports of RCTs in Circulation, European Heart Journal, Heart, JAMA Cardiology and Journal of the American College of Cardiology using their online International Standard Serial Number. We limited the publication years from 2024 to 2025 to capture the most recent reporting practices. Conference abstracts, secondary analyses, post hoc analyses and pooled analyses based on RCTs, as well as other study designs, namely errata, commentaries and editorials, were not eligible for inclusion. The analysis reviewed the abstracts and full texts of the articles, focusing on two categories: (1) ‘null acceptance’ statements, defined as phrases interpreting a non-significant p value from hypothesis testing, defined by the original authors, as indicating ‘no difference,’ ‘no effect’ or ‘similarity’ between intervention and …
