1

Reply

We thank Dr. Jolobe for the comments on our paper regarding NT-proBNP to rule-out heart failure among patients with atrial fibrillation [ , ]. Dr. Jolobe raises two concerns: the potential misclassification of patients due to the definition of heart failure used in our study and the presence of patients in our cohort with comorbidities associated with high NT-proBNP levels.

We acknowledge that the definition of heart failure used in our study as an ejection fraction below 40% is dichotomous and may be considered as too simple. Dr. Jolobe refer to the Framingham and the Boston criteria for diagnosing congestive heart failure. Although these clinical scales may identify patients with reduced ejection fraction these scales have varying sensitivity and specificity and may also not identify patients with low ejection fraction [ ]. Importantly, there is a broad contemporary consensus that heart failure with reduced ejection fraction is defined as an ejection fraction below 40% similar to the definition used in our study [ , ].

The purpose of our study was to investigate whether a significant reduced ejection fraction can be ruled out in an emergency department setting using biochemical testing, as this may help guide the initial treatment strategy of atrial fibrillation in case ultrasound evaluation is not readily available. Accordingly, we aimed to investigate whether the NT-proBNP may be used as an adjunct for the clinician to decide whether or not to treat a patient with atrial fibrillation in an emergency department setting using beta blockers or calcium antagonists.

We agree with Dr. Jolobe that NT-proBNP levels can be elevated due to multiple causes other than reduced ejection fraction. It is well known that a wide range of conditions in addition to atrial fibrillation are associated with elevated NT-proBNP levels such as diastolic dysfunction, COPD, kidney failure, aortic stenosis, etc. [ ]. Therefore, we included an unselected population of patients with atrial fibrillation presenting to the emergency department of which many had comorbidities that could lead to higher NT-proBNP levels. Thus, our study shows that heart failure with ejection fraction below 40% can only be ruled out at NT-proBNP levels below 739 ng/L as slightly elevated levels above this cut-off may be due to other comorbidities or atrial fibrillation itself. Interestingly, the performance of NT-proBNP to rule out heart failure with reduced ejection fraction appeared much better in patients <75 years of age as compared to older patients, likely due to competing causes of higher NT-proBNP levels i.e. comorbidities. In our study, we studied an unselected cohort to achieve a high negative predictive value to rule-out of heart failure. We believe this is of importance if NT-proBNP should help guide clinicians in a real-life setting reflecting patients who are admitted with atrial fibrillation to the emergency department. Based on our study we conclude NT-proBNP can be used to rule out heart failure in atrial fibrillation patients with a high negative predictive value, but low specificity.

CRediT authorship contribution statement

Cecilie Budolfsen: Conceptualization, Writing – original draft, Writing – review & editing. Camilla Bang Hoeks: Conceptualization, Writing – review & editing. Kasper G. Lauridsen: Conceptualization, Writing – review & editing. Bo Løfgren: Conceptualization, Writing – review & editing.