Age-adjusted NT-proBNP thresholds for patients aged over 75 years attending a single-centre rapid access heart failure clinic

Background

Diagnosing heart failure (HF) in the outpatient setting remains challenging. European guidelines recommend an NT-proBNP threshold of 125pg/ml to refer for specialist assessment and echocardiography(1). The UK NICE guidelines recommend a higher threshold of ≥400pg/ml(2). Despite this higher threshold, growing demand outstrips capacity. The ESC HFA consensus statement recently recommended the use of age-adjusted BNP thresholds advising ≥500pg/ml for patients over 75(3). Retrospective UK GP data suggest age-adjusted thresholds reduce sensitivity but improve specificity when compared with the ESC threshold of 125pg/ml(4).

Purpose

This single-centre study evaluated the ESC HFA consensus age-adjusted threshold by examining outcomes of patients over 75 referred with NT-proBNP values of 400-499pg/ml.

Methods

Data were collected for patients over 75 referred by primary care to the Rapid Access Heart Failure Clinic (RAHFC) at our UK tertiary referral centre between Jan 2022 and Dec 2024. NT-proBNP values were taken from referral letters. Referrals were accepted or rejected based on NT-proBNP ≥400pg/ml, clinical signs/symptoms of HF, pulse check for AF. Data were anonymised and analysed using Excel.

Results

Over a three-year period, our RAHFC was referred a total of 980 patients over the age of 75. 56 patients had NT-proBNP values of 400-499pg/ml. Ten were rejected due to inadequate information, previous normal echocardiogram, incorrect referral or an alternative likely diagnosis. Of the remaining 46, five were not investigated (two did not attend, two declined tests and one died of myocardial infarction before review).

41 patients were reviewed. HF was diagnosed in 22% (n=9) on first review: seven with HFpEF and two with HFmrEF. One HFmrEF patient was later diagnosed with cardiac amyloidosis and had one HF hospitalisation; the other HFmrEF patient had two HF hospitalisations.

One additional patient was diagnosed with Stage B HFpEF or ‘pre-HF’. No new HFrEF cases were identified.

Among those not diagnosed with HF, two had severe aortic stenosis (AS). Two others were later diagnosed with HF during follow up, one of whom was hospitalised. Within the eleven patients of 46 (24%) diagnosed with HF, there were four hospitalisations and one death over a mean follow up period of 27.4 months.

Conclusions

In patients over 75 with NT-proBNP values of 400-499 pg/ml, 24% were ultimately diagnosed with HF, mainly HFpEF, with 18% HFmrEF. HF disease severity was generally mild with no HFrEF cases found. The single case of cardiac amyloidosis is unusual given the low NT-proBNP value. Two patients without HF had an alternative significant diagnosis, severe AS (4%).

Raising the threshold to ≥500 pg/ml for this group may reduce unnecessary investigations without significantly affecting HF hospitalisations. However prospective studies are needed to assess long term outcomes of adopting the ≥500 pg/ml threshold.