Device Measured Rapid Shallow Breathing Index Reflects Changing Respiratory Patterns but Minute Ventilation Reflects Changing Activity During Worsening Heart Failure in Ambulatory Patients

Background Respiratory distress is common in heart failure (HF) and a primary driver for HF hospitalizations. Minute Ventilation (MV), a product of respiratory rate and tidal volume, is known to be elevated in HF patients due to ventilation/perfusion (V/Q) mismatch. However, it is not known if changes in MV accurately reflect emergence of rapid shallow breathing patterns in ambulatory patients preceding a HF event. Methods The MultiSENSE trial enrolled 900 patients implanted with a COGNIS CRT-D and followed them up to 1 year. Device software was modified to permit collection of chronic diagnostic sensor data including impedance based respiration rate (RR) and tidal volume (TV), which was used to compute MV (= RR*TV) and Rapid Shallow Breathing Index (RSBI = RR/TV), and activity (XL). Daily averages were separately computed over entire 24 hours as well as during resting epochs. HF events (HFEs) were independently adjudicated and defined as HF admissions or unscheduled visits with intravenous HF treatment. Relative changes preceding HFEs were computed between a baseline 30–60 days prior to HFEs (BL) and 3-day pre-HFE (ST) as (ST-BL)/BLx100% and reported as mean +/- SEM. Significance was tested using Wilcoxon signed-rank test. Results 900 patients followed for a year experienced 192 HFEs. Using 24-hour averages, significant changes were observed in RR, TV and RSBI indicating the emergence of rapid shallow breathing pattern leading up to HFE. MV average over 24 hours showed nonsignificant decrease coincident with decreased patient activity but showed no change when daily averaging was limited to resting epochs. In contrast, RR, TV and RSBI were significantly changed even at rest in directions consistent with the emergence of rapid shallow breathing pattern. Conclusion Device measured rapid shallow breathing is significantly elevated in the three day epoch preceding HFEs, whereas minute ventilation is not, in both 24-hour as well as resting period daily averages. Automatic ambulatory longitudinal monitoring of changes in rapid shallow breathing patterns may enable better monitoring for emerging respiratory distress in HF patients.