Monitoring diaphragmatic effort during diaphragm neurostimulation-assisted ventilation.
Andrea Castellvi-Font, Idunn S Morris, Francisco José Parrilla-Gómez, Matías Madormo, Catherine A Bellissimo, Viral Thakkar, Nawzer Mehta, Thiago Bassi, Laurent J Brochard, Niall D Ferguson, Ewan C Goligher
Abstract
Open AccessRATIONALE: Diaphragm neurostimulation-assisted ventilation (DNAV) can improve cardiopulmonary function during passive mechanical ventilation. However, this technique requires a reliable method to monitor and titrate diaphragmatic loading to avoid both insufficient and excessive diaphragmatic stimulation. OBJECTIVE: To establish whether the reduction in airway pressure-time product (ΔPTPaw) obtained during diaphragm neurostimulation in assist control volume-controlled mode accurately quantifies the magnitude of respiratory muscle effort elicited by neurostimulation. METHODS: This was a secondary analysis of the STIMULUS trial. Diaphragm neurostimulation was titrated across four levels targeting progressive occlusion pressures of 0, - 5, - 10, and - 15 cm H₂O at two PEEP levels. At each level, airway, esophageal, and gastric pressures were recorded to compute transdiaphragmatic pressure-time product (PTPdi), respiratory muscles pressure-time product (PTPmus), and ΔPTPaw, defined as the difference in airway pressure-time product between non-stimulated and stimulated breaths. Linear mixed-effects models, Bland-Altman analyses, and receiver operating characteristic (ROC) curves were used to assess agreement and discriminative ability. MEASUREMENTS AND MAIN RESULTS: Twelve patients contributed 494 high-quality respiratory cycles (63% of recorded cycles). Valid Pdi data were available in nine patients. Increasing neurostimulation was associated with higher PTPdi and PTPmus and a corresponding reduction in PTPaw. ΔPTPaw was correlated with both PTPdi (R² = 0.82) and PTPmus (R² = 0.92), with good agreement observed (limits: - 4 to 44 cm H₂O·s/min for PTPdi, and - 5 to 39 cm H₂O·s/min for PTPmus). ΔPTPaw demonstrated excellent discrimination for inadequate (area under receiver operating characteristic curve, AUROC ≥ 0.94) and excessive (AUROC ≥ 0.86) diaphragmatic effort. CONCLUSIONS: ΔPTPaw is a reliable, non-invasive surrogate for monitoring diaphragm loading during DNAV under assist-controlled volume-controlled mode and may guide neurostimulation titration in mechanically ventilated patients.