Positive end-expiratory pressure selection based on best respiratory system compliance or collapse/hyperdistension curves in patients with acute respiratory distress syndrome: lack of correlation with alveolar recruitment

Dear Editor,

Electrical impedance tomography (EIT) and the nitrogen washin/washout method (NWI-WO) for measuring recruitable lung volume (Vrec) provide new information for positive end-expiratory pressure (PEEP) selection in patients with acute respiratory distress syndrome (ARDS) [1, 2]. With EIT, the selected PEEP (PEEP_EIT) is the intersection between the collapse and hyperdistension (CH) curves [1], and a higher Vrec may deserve a higher PEEP [2]. We hypothesized that measurement of Vrec may help in our decision of PEEP and thus conducted a study comparing the Vrec with PEEP_EIT or PEEP selection based on best respiratory system compliance (PEEP_{best Crs}) during decremental PEEP titration in ARDS patients. Our institutional review board (B-ER-103-317) approved this study and all patients provided informed consent. The preliminary result has been presented in abstract form at an international meeting [3].

Extended sigh was used for alveolar recruitment [4]. PEEP was sequentially increased from baseline to 15, 20, and 25 cm H₂O and then decreased from 25 cm H₂O to 20 and 15 cm H₂O twice. Each pressure was maintained for 30 s. Tidal volume (Vt) was decreased by 25% from baseline Vt during the incremental phase and increased by 25% during the decremental phase. Airway plateau pressure (P_pl) was determined at PEEP 25 cm H₂O (PEEP₂₅), 20 cm H₂O (PEEP₂₀) during the second recruitment maneuver and following end-expiratory lung volume (EELV) determination by NWI-WO at PEEP_H 15 cm H₂O, PEEP_I 11 cm H₂O, and PEEP_L 7 cm H₂O (Fig. 1a). Cases with a difference between NWI-WO measurements of > 20 % were excluded [2]. Vrec was calculated as the difference between EELV at PEEP_H and PEEP_L, after subtracting the minimal predicted increase in lung volume due to the difference in PEEP [2]. The method of Costa et al. [5] was used to determine the degree of CH during decremental PEEP titration. The intersection between CH curves and PEEP_EIT were determined (Fig. 1b). Respiratory system compliance (Crs) was calculated as Vt/(P_pl–PEEP).

Data are presented as mean ± SD. Spearman’s rank correlation test and Kruskal–Wallis test was used for statistical analysis.

Twenty-three patients completed the study and five cases were excluded for difference between NWI-WO > 20 %. Eighteen patients left for analysis. Baseline Vt was 7.7 ± 0.7 mL/Kg PBW and baseline PEEP level was 12.3 ± 2.2 cm H₂O. PEEP levels were 24.8 ± 1.6 (PEEP₂₅), 20.3 ± 1.5 (PEEP₂₀), 15.7 ± 1.7 (PEEP_H), 11.8 ± 1.9 (PEEP_I), and 8.0 ± 2.2 (PEEP_L) cm H₂O. Corresponding P_pl were 33.7 ± 2.7, 31.6 ± 3.4, 27.9 ± 3.0, 22.4 ± 2.1, and 18.7 ± 2.2 cm H₂O, and corresponding Vt were 210.7 ± 45.7, 342.0 ± 45.7, 456.7 ± 51.1, 454.5 ± 52.9, and 455.6 ± 53.1 mL. EELV was 1756 ± 390, 1494 ± 359.4, and 1201.0 ± 313.6 ml at PEEP_H, PEEP_I, and PEEP_L, respectively. Vrec between PEEP_H and PEEP_L was 218.6 ± 261.4 ml. CH at the intersection point averaged 6.9 ± 2.8 %. PEEP_{best Crs} and PEEP_EIT correlated well (Fig. 1c). Vrec did not correlate with PEEP_{best Crs} or PEEP_EIT (Fig. 1d).

Our findings suggest that PEEP selection based on PEEP_EIT is justified. However, measured Vrec alone did not provide sufficient information for PEEP selection. However, the recruitment maneuver used in current study may not fully recruit the lung. This result may not be applied to a different recruitment protocol.