Detection of strong inspiratory efforts from the analysis of central venous pressure swings: a preliminary clinical study.

Abstract

Swings of central venous pressure (ΔCVP) may reflect those of pleural and esophageal (ΔPES) pressure and, therefore, the strength of inspiration. Strong inspiratory efforts can produce some harm. Herein we preliminarily assessed the diagnostic accuracy of ΔCVP for strong inspiratory efforts in critically-ill subjects breathing spontaneously. We measured ΔCVP and ΔPES in 48 critically-ill subjects breathing spontaneously with zero end-expiratory pressure (ZEEP) or 10 cmH<inf>2</inf>O of continuous positive airway pressure (CPAP). The overall diagnostic accuracy of ΔCVP for strong inspiratory efforts (arbitrarily defined as ΔPES >8 mmHg) was described as the area under the receiver operating characteristic (ROC) curve, with 0.50 indicating random guess. The agreement between ΔCVP and ΔPES was assessed with the Bland-Altman analysis. ΔCVP recognized strong inspiratory efforts with an area under the ROC curve of 0.95 (95% confidence intervals, 0.85-0.99) with ZEEP and 0.89 (0.76-0.96) with CPAP, both significantly larger than 0.50 (P<0.001). With the best cut-off value around 8 mmHg, the diagnostic accuracy of ΔCVP was 0.92 (0.80-0.98) with ZEEP and 0.94 (0.83-0.99) with CPAP. With ZEEP, the median difference between ΔCVP and ΔPES (bias) was -0.2 mmHg, and the 95% limits of agreement (LoA) were -3.9 and +5.5 mmHg. With CPAP, bias was -0.1 mmHg, and 95%-LoA were -5.8 and +4.5 mmHg. In both cases, ΔCVP correlated with ΔPES (r<inf>s</inf> 0.81 and 0.67; P<0.001 for both). In critically-ill subjects breathing spontaneously, ΔCVP recognized strong inspiratory efforts with acceptable accuracy. Even so, it sometimes largely differed from ∆PES.