The high rate of re-hospitalization for heart failure might be reduced by improving noninvasive techniques for identifying elevated left ventricular (LV) filling pressure. We previously showed that changes in a finger photoplethysmography (PPG) waveform during the Valsalva maneuver (VM) reflect invasively measured LV end-diastolic pressure (LVEDP). We have since developed a hand-held device that analyzes PPG while guiding the expiratory effort of a VM. Here we assessed the sensitivity and specificity of this device for identifying elevated LVEDP in patients. We tested 82 participants (28 women), aged 40 to 85 years, before a clinically indicated left heart catheterization. Each performed a VM between 18 and 25 mm Hg for 10 seconds into a pressure transducer. PPG was recorded continuously before and during the VM. LVEDP was measured during the catheterization. An equation for calculating LVEDP was derived using (1) ratio of signal amplitudes: minimum during VM to average at baseline, (2) ratio of peak-to-peak time intervals: minimum during VM to average at baseline, and (3) mean blood pressure. Calculated and measured LVEDP were compared. The range of measured LVEDP was 4 to 35 mm Hg. Calculated LVEDP correlated with measured LVEDP (p <0.0001, r = 0.56). A calculated LVEDP >20 mm Hg had a 70% sensitivity and 86% specificity for identifying measured LVEDP >20 mm Hg (area under receiver-operating characteristic curve 0.83). In conclusion, a hand-held device for assessing LV filling pressure had high specificity and good sensitivity for identifying LVEDP >20 mm Hg, a clinically meaningful threshold in heart failure.
Copyright © 2017 Elsevier Inc. All rights reserved.