Assessment of Pressure-Volume Loop Accuracy Under Circuit Leak Conditions

Background: Mechanical ventilator pressure-volume (PV) loops offer a unique opportunity to assess the degree of recruitable lungs, determine optimal PEEP, and administer lung recruitment maneuvers. The presence of inflation and deflation limb hysteresis provides a signal of lung recruitability. The concept of normalized maximum distance (NMD% = inflation-deflation hysteresis/inflation volume) has been proposed to quantify a value that represents recruitment potential. An NMD% of greater than 40%, has shown a high potential for recruitable lung. This bench study aimed to create a lung model that can deliver a reproducible PV loop maneuver and test the PV loop’s accuracy under various percentages of circuit leaks. Methods: A lung model (patent pending) was created to generate reproducible PV loop shapes that demonstrated a consistent lower inflection point (LIP) and a recruitable lung profile while connected to a critical care mechanical ventilator with PV loop software. An initial validation sequence of 20 maneuvers was performed under 0% leak conditions to confirm the reproducibility (LIP and NMD%) of the device. At 0% leak, a 10-s, 40 cm H2O inflation recruitment maneuver was performed to validate reproducibility and establish a recruitable volume baseline. Assessments of the LIP, inflation volume (I Vol), hysteresis (P20 dV), and NMD% were tested with a circuit leak of 0%, 5%, 10%, and 15%. A recruitment maneuver was also performed at the respective leaks to determine the recruited volume. Results: At 0% circuit leak, the mean LIP was 11.17 ± 0.22. The mean LIP during all other testing conditions was 11.15 ± 0.24 (P = .86). At 0% leak, the NMD% was 28.3%. Adding a 5% leak to the circuit increased the mean NMD% to 56.6%. NMD% at 10 and 15 % leak were 63.4 and 72.4% respectively (Table 1). Recruitable volume (Rec Vol) during a 10-s recruitment maneuver at 0% leak showed Rec Vol of 17.6 mL. Rec Vol at 5, 10, and 15% leak were 420.8, 559.1, and 993.2 mL respectively (Table 2). Conclusions: Our group was able to construct a reproducible and reliable PV loop evaluation model. The presence of a leak can drastically impact the PV tool’s assessment of recruitable lung. Adding a 5% leak to the ventilator circuit nearly doubled the NMD% compared to no leak, giving a false impression of high recruited volume. More data are needed to support this evidence in a clinical setting, but consideration of leak when performing a PV loop maneuver is necessary.