Deciphering a decision support tool (CoDE-HF) to improve the diagnosis of acute heart failure

Comment on ‘Development and validation of a decision support tool for the diagnosis of acute heart failure: systematic review, meta-analysis, and modelling study’, which was published in the BMJ, https://doi.org/10.1136/bmj-2021-068424. Acute HF is one of the most frequent causes of unplanned hospitalization, producing challenges in diagnosis, treatment, and health service organization, beside the heavy economic burden.2 Delivering an appropriate and timely care to patients is challenged by diagnostic uncertainty, variable individual response to therapy, and limited integration of data at the point of care.3 Although current guidelines recommend immediate NT-proBNP testing to aid the diagnosis of acute HF,4 its utility is limited by the high variability in diagnostic performance among clinically relevant subgroups, such as the elderly and patients with renal disease or obesity.5 Statistical approaches incorporating patient characteristics with clinical and laboratory data may enable a more individualized assessment and enhance diagnostic performance across patient subgroups.6