Proportional Assist Ventilation and Respiratory Effort Monitoring During Transition to Spontaneous Breathing

Abstract

Introduction:

Proportional assist ventilation plus (PAV+) uses a work of breathing bar (WOB-Bar) to provide real-time, noninvasive estimates of patient work of breathing (WOB_pt). Although designed to guide titration during the transition to spontaneous ventilation, the clinical accuracy of this tool remains unvalidated against gold-standard measures.

Methods:

This physiological randomized crossover study included 35 subjects weaning from mechanical ventilation. We compared the WOB-Bar against the esophageal pressure-time product (PTP_es-min), the pressure drop in the airway during the first 100 ms of an inspiratory effort against an occluded airway (P_0.1), and the pressure variation during an end-expiratory occlusion (P_occ) across 3 randomized assistance PAV+ levels (20%, 50%, and 70%). Effort was categorized as low, adequate, or high based on suggested physiological thresholds.

Results:

The WOB-Bar correctly classified effort (relative to PTP_es-min) in only 44.5% of cases (95% CI 32.9–56.4%), with a negligible weighted κ of 0.01. P_occ demonstrated the best performance with 66.5% agreement (95% CI 57.9–75%) and a κ of 0.35. A multivariable mixed-effects model (R² = 0.84) showed that WOB_pt (P < .001), P_0.1 (P = .040), and P_occ (P = .02) were all independent predictors of PTP_es-min. However, despite a positive association between WOB_pt and PTP_es-min, the categorical WOB-Bar showed limited accuracy and tended to underestimate respiratory effort, particularly in high-effort states. Inter-individual variability accounted for 77.5% of the total variance in respiratory effort.

Conclusions:

The WOB-Bar frequently underestimated effort and was insufficient as a stand-alone tool for PAV+ titration. Airway occlusion maneuvers (P_occ and P_0.1) were more reliable indicators of PTP_es-min. Since current thresholds are extrapolated from healthy individuals and unvalidated during weaning, these findings are hypothesis generating. Future studies should establish context-appropriate thresholds and evaluate their clinical implications.