Misattribution of a Retrospective Case-Control Statistic as Prospective Epidemiologic Evidence in HFpEF: A Correction and a Caution for the Era of AI-Assisted Scientific Writing

A recent editorial published in the European Journal of Heart Failure stated that prospective studies report that up to 76% of patients with preserved left ventricular ejection fraction and unexplained breathlessness are ultimately diagnosed with HFpEF after invasive evaluation, citing Reddy et al. (JAMA Cardiology, 2022) as the supporting reference. This characterisation is factually incorrect in two respects: the cited study was a retrospective case-control study, not a prospective cohort study, and the 76% figure represents the case-control ratio of a deliberately constructed dataset of patients already referred for invasive haemodynamic exercise testing at tertiary centres, not the diagnostic yield of HFpEF in an unselected dyspnoeic population. Applying this figure to general ambulatory practice substantially overestimates HFpEF prevalence and misrepresents the diagnostic challenge in primary care. We document the inaccuracy, provide the correct interpretation of the cited study, and argue that timely correction is especially important in the current era of large language model-based AI writing assistants, which retrieve and reproduce published claims at scale without the contextual scrutiny a careful reader would apply. Uncorrected inaccuracies in authoritative publications risk propagating into the broader literature as established fact.