Non-invasive blood pressure monitoring underestimates hypertensive response to exercise in suspected heart failure with preserved ejection fraction

Exercise intolerance is the commonest morbidity experienced by patients with heart failure with preserved ejection fraction (HFpEF).1 Invasive haemodynamic stress testing has identified ventricular–arterial coupling to be an important contributor to exercise intolerance.2 Its presence, among other parameters, is suggested by exaggerated blood pressure responses to exercise, which are driven by arterial stiffening and impaired vasodilator reserve.3,–5 Although dynamic invasive haemodynamic testing is the gold standard technique for assessing patients with suspected HFpEF, it is not widely available due to its cost and invasive nature. Research groups are currently evaluating non-invasive stress tests to diagnose HFpEF in individuals with exercise intolerance. The accuracy of non-invasive blood pressure (NIBP) measurement using standard oscillometric cuffs in exercising patients with suspected HFpEF is not well reported. We sought to establish the reliability of NIBP monitoring in identifying exaggerated blood pressure responses in patients with suspected HFpEF who underwent exercise right heart catheterisation, versus the gold standard of invasive blood pressure (IBP) monitoring. We retrospectively searched the case notes for all patients who underwent exercise right heart catheterisation for the assessment of suspected HFpEF who had simultaneous invasive and non-invasive arterial blood pressure measurements at a quaternary referral centre between December 2011 and October 2018. Patients had been recruited for a pilot study investigating the correlation between serum brain natriuretic peptide (BNP) and haemodynamics in suspected HFpEF. Exercise right heart catheterisation was performed using supine bicycle ergometry. A weight-corrected exercise protocol was applied, with an initial workload of 0.3 W/kg, rising to 0.6 W/kg after 3 min and to 1 W/kg after a further 3 min. The duration of exercise was limited by patient symptoms (fatigue/dyspnoea). An arterial line was inserted into the right or left radial artery, and an automated oscillometric blood pressure device (system: Phillips Xper Flex Cardio, Koninklijke Philips NV, The Netherlands; cuff: Critikon Dura-Cuf, GE Healthcare, USA) was applied to the opposite arm recording measurements at 3 min intervals. Simultaneous blood pressure readings were documented at rest, 3 min and peak exercise. A hypertensive response to exercise was defined as peak in systolic blood pressure (SBP) of 210 mmHg or greater in men (≥190 mmHg in women) or in diastolic blood pressure (DBP) of 110 mmHg or greater.6,7 This study was granted institutional ethical approval (number 236/19). IBP and NIBP measurements at rest and peak exercise were assessed for agreement using the Bland and Altman method.8 Data are presented as mean (± standard deviation; SD). Correlation was assessed using the Pearson correlation coefficient (r). Finally, the sensitivity and specificity of non-invasive monitoring in diagnosing a hypertensive response to exercise versus invasive measurements was calculated. Nineteen patients were identified, and seven were excluded due to incomplete blood pressure recordings at peak exercise. Twelve patients were studied; mean age was 67 years (SD 10), eight patients (66.6%) were women, mean body mass index (BMI) was 35.3 kg/m2 (SD 8.4). Six patients (50%) had New York Heart Association grade 3 symptoms. The comorbidities of the cohort included hypertension (n = 7, 58%), atrial fibrillation (n = 5, 42%), diabetes mellitus (n = 2, 17%), ischaemic heart disease (n = 5, 42%) and chronic obstructive pulmonary disease (n = 1; 8%). Nine of the patients (75%) were subsequently diagnosed with HFpEF. (a) Bland–Altman plots demonstrating the change in agreement between invasive and non-invasive blood pressure measurements at rest and at exercise. (b) Scatter plots with linear regression lines demonstrating correlation at rest and at peak exercise. (a) Bland–Altman plots demonstrating the change in agreement between invasive and non-invasive blood pressure measurements at rest and at exercise. (b) Scatter plots with linear regression lines demonstrating correlation at rest and at peak exercise. This study is the first to our knowledge to report the accuracy of NIBP monitoring in patients with suspected HFpEF. We found that non-invasive measurements underestimated blood pressure at rest and this effect is exaggerated with exercise; demonstrated by the deterioration in agreement and correlation with invasive measurements at peak exercise. Consequently, only 40% of patients with IBP-confirmed hypertensive response to exercise were detected using non-invasive monitoring. These findings may reflect the inaccuracy in performing non-invasive measurements on an obese and co-morbid cohort. There are several limitations to our study, which are largely due to its retrospective design. We compared radial invasive to brachial NIBP; and radial blood pressure is on average 5 mmHg higher than brachial according to one study.9 In addition, measurements were compared from opposite arms and with no prior screening for a blood pressure differential between arms. Atrial fibrillation can impair the accuracy of NIBP measurement, and the incidence was 42% in our cohort. Finally, the oscillometric system used may not be the optimal method of non-invasively monitoring blood pressure during exercise; more accurate results may be expected with a manual technique or R-wave gated automated system.10 Our main findings are that NIBP monitoring underestimates blood pressure during exercise and lacks sensitivity in diagnosing an exaggerated hypertensive response. An invasive approach may be necessary for the accurate assessment of exercise blood in patients undergoing assessment for HFpEF, and exercise right heart catheterisation represents a good opportunity for this. A conceptual abstract of part of this work has been accepted for poster presentation at the 67th Annual Scientific Meeting of the Cardiac Society of Australia and New Zealand (CSANZ), Adelaide, Australia, 8–11 August 2019.