Assessing agreement among non-invasive indicators for inspiratory effort during pressure support ventilation

During pressure support ventilation (PSV), the accuracy of non-invasive indicators in diagnosing high or low inspiratory effort has been validated. However, the correlation and agreement of these indicators remain unclear. This study aims to investigate the correlation and agreement among non-invasive inspiratory effort indicators, and to compare characteristics of inspiratory effort in neurocritical and non-neurocritical patients.

This was a single-centre prospective observational study. We collected three non-invasive inspiratory effort indicators, pressure muscular index (PMI), the maximal negative swing of airway pressure during expiratory occlusion (ΔPocc), and the airway occlusion pressure during the first 100ms (P0.1). Cutoff values for these indicators derived from esophageal pressure-time product (PTPmus) were chosen for this study. The correlation and agreement of these indicators were analyzed using Spearman's rank correlation test and linear weighted Kappa analysis. Characteristics of PSV settings and inspiratory effort in neurocritical and non-neurocritical patients were compared.

Ninety-seven patients were enrolled in this study. Correlation analysis showed a moderate correlation between PMI and ΔPocc (rho = -0.524, p < 0.001), ΔPocc and P0.1 (rho = 0.588, p < 0.001), while no correlation between PMI and P0.1 (rho = -0.140, p = 0.172). There was a moderate agreement between ΔPocc and P0.1 (k = 0.459, p < 0.001), a fair agreement between PMI and ΔPocc (k = 0.362, p < 0.001), but no agreement between PMI and P0.1 (k = 0.134, p = 0.072). The correlation of these indicators was similar in neurocritical patients compared with non-neurocritical patients, but agreement was poor.

The study showed that PMI and ΔPocc had moderate correlation and fair agreement, ΔPocc and P0.1 had moderate correlation and agreement, while PMI and P0.1 had no correlation and agreement.