Characterising dysfunctional breathing seen in post-acute sequelae of SARS-CoV-2 using approximate entropy

Area of Science:

• Pulmonary Medicine
• Cardiorespiratory Physiology
• Biomedical Data Analysis

Background:

• Dysfunctional breathing (DB) is a common issue in post-acute sequelae of SARS-CoV-2 (PASC) patients, potentially causing persistent symptoms.
• Current identification of DB during cardiopulmonary exercise testing (CPET) relies on subjective visual assessment of breathing patterns.
• There is a need for objective methods to reliably detect DB in PASC patients.

Purpose of the Study:

• To investigate the utility of approximate entropy (ApEn) in objectively distinguishing DB from normal breathing patterns in PASC patients undergoing CPET.
• To determine if ApEn of minute ventilation (V'E) can serve as a reliable metric for DB detection.

Main Methods:

• Collected breath-by-breath CPET data from 82 PASC subjects and 25 controls.
• Normalized and detrended time-series data for minute ventilation (V'E), tidal volume (VT), and breathing frequency (BF).
• Calculated ApEn for V'E, VT, and BF at varying exercise intensities up to the anaerobic threshold (AT).

Main Results:

• ApEn V'E was significantly higher in PASC subjects with visually identified DB compared to controls and PASC subjects without DB (p<0.05).
• Receiver operating characteristic (ROC) analysis identified an optimal ApEn V'E cut-off of 0.17, yielding 81% sensitivity and 72% specificity for DB detection.
• ApEn VT and ApEn BF showed less distinction among PASC groups but were elevated compared to controls.

Conclusions:

• ApEn V'E provides an objective and reliable method for differentiating dysfunctional breathing from normal breathing patterns during CPET.
• This quantitative metric can complement subjective visual interpretation of CPET data for improved DB diagnosis in PASC patients.
• Objective assessment of breathing patterns using ApEn may aid in understanding and managing persistent symptoms in PASC.