A New Dynamic Slow-Flow Esophageal Balloon Pressure-Volume Curve Calibration Method to Assess Transpulmonary Pressure Monitoring

Introduction

Transpulmonary pressure (P_L) monitoring is becoming increasingly popular for estimating mechanical stress applied to injured lungs or tailoring ventilator settings during ARDS.^1–5 The computation of P_L relies on pleural pressure, for which the esophageal pressure (P_es) is an acceptable regional approximation.^2,6

Clinicians typically measure P_es using a balloon catheter filled with air. The conventional methodology for balloon volume calibration involves filling it with a standardized volume and validating its position using the Baydur modified test, also known as the occlusion test (OT).^1,3 However, this methodology has recently been called into question at the bedside. Indeed, Mojoli and colleagues^7,8 proposed an original method based on the assessment of pressure-volume (PV) curve of the esophageal balloon catheter, which yielded promising results. Whereas this calibration procedure is attractive, it remains cumbersome and time consuming at the bedside, hindering widespread adoption for P_L monitoring.^2,7

In this short report, we propose a simplified method based on Mojoli's hypothesis. We developed an experimental protocol to dynamically build and record the PV curve of the esophageal balloon. Our method involves filling the balloon at a slow and continuous rate using an automatic process that does not require any intervention during recording.

Methods

The study was approved by the ethics committee of Rennes University Hospital and the French national authority. It is part of a larger observational study (NCT05697666) investigating the impact of neuromuscular relaxants on respiratory mechanics. Subjects or their family provided written informed consent after receiving a detailed protocol.

Subjects included had moderate to severe ARDS according to the Berlin definition and were eligible for muscular paralysis. Subjects were monitored using the NutriVent (Sidam, Mirandola, Italy) P_es catheter device (to evaluate the partitioning of respiratory mechanics) which was positioned according to recommended guidelines.^1–3 Ventilation was conducted using protective volume controlled–continuous mandatory ventilation mode, with a tidal volume approximately 6 mL/kg of predictive body weight, and the PEEP level was at the discretion of the clinician.

An air-filled syringe was connected to the P_es monitoring circuit using a Y-branch and controlled using a syringe driver. Baseline pressure was zeroed before air injection. P_es and balloon volume were continuously recorded up to 8 mL inflation rate at 100 mL/h (Fig. 1A–B) using a manometric sensor coupled with respiratory variables (airway pressure and flow) via a FluxMed device (MBMed, Buenos Aires, Argentina). The balloon was then emptied and zeroed. A step-by-step OT sequence was then performed by filling the esophageal balloon with fixed volumes from 0.5–8 mL, and an OT was conducted for each static volume level according to Mojoli's description (Fig. 1B and 1E).⁷ OT ratio was defined as the ratio of increment of P_es (ΔP_es) to the increment in airway pressure (ΔP_aw) during expiratory (exp) occlusion and compression of the chest, expressed as ΔP_es/ΔP_aw ratio (with usual validated values ranging from 0.8 to 1.2). The value associated with the ratio closest to 1 was the most suitable volume balloon for assessing P_es according to Mojoli's method^1,2,7 (Fig. 1E–F).

Fig. 1.

Synopsis of experimental protocol. Fig 1A: Set-up of experimental protocol for high-resolution dynamic esophageal balloon pressure-volume (PV) curve. Esophageal balloon was filled with air with a piloted syringe (filling balloon at rate of 100 mL/h, ie, 1.67 mL/min). Fig 1B: Air-filled esophageal balloon PV curve built with a continuous slow-flow inflation sequence (and subsequent occlusion test [OT] maneuver): recording and graphical/computerized analysis. Fig 1C–D: Continuous slow-flow PV curve graphics focus on V_bestXP: recording and graphical/computerized analysis. C: Volume balloon with maximal P_es tidal variation. D: Analytics of dynamic PV curve leading to an estimate of V_bestXP = 1.37 mL. Fig 1E–F: Detailed description focused on OT sequence. E: Computation of OT (as ΔP_es/ΔP_aw ratio) at each volume step (see text for detailed description of OT sequence). F: Maximal ratio defining V_bestOT (top green triangle): here V_bestOT = 1.5 mL, which closely match with V_bestXP (1.37 mL) estimated by our experimental method (panel D). PV = pressure-volume; P_es = esophageal pressure; esp = expiratory; P_aw = airway pressure; OT = occlusion test.

The esophageal balloon signals were processed and cleaned using software programs (Graphysio and MATLAB [MathWorks, Natick, Massachusetts]), and esophageal balloon PV curves were calculated (Fig. 1B). The experimental variables (XP variables) including V_bestXP (mL), E_ew (cm H₂O/mL), and P_es insp/exp_XP (cm H₂O), were compared to OT variables [V_bestOT (mL), P_es insp/exp_OT (cmH₂O)] derived from static PV steps using Mojoli's method as a standard. V_bestXP represents maximal tidal variation of P_es; V_bestOT is the balloon volume closer to 1 for OT ratio; E_ew is esophageal wall elastance, and P_es insp/exp are the values of insp/exp P_es at V_best with OT method and at volume = 4 mL (V4).

Statistical comparison between continuous variables (V_bestXP and V_bestXP) and P_es (P_esOT and P_esV4) were performed using the non-parametric Mann-Whitney test with a significance level of .05. Correlations between volumes (V_bestXP and V_bestOT) and P_es analyzed using Pearson correlation coefficient and linear regression statistics (R²).

Results

Complete acquisition and recording of esophageal balloon PV curve were performed in 6 subjects, resulting in a total of 10 PV curves. Demographic and clinical characteristics were as follows: age 44 ± 11 y; moderate to severe ARDS with a P_aO2/F_IO2 116 ± 35; plateau pressure 26 ± 4 cm H₂O; PEEP level 12 ± 3 cm H₂O; and etiology of ARDS including pneumonia (n = 2), pancreatitis (n = 3), and aspiration (n = 1). A typical pattern of dynamic low-flow inflation esophageal balloon PV curve is illustrated in Figure 1B and 1C. Of 6 subjects, 4 were included in the comparison of continuous esophageal balloon PV curve to complete OT. One subject was excluded due to high OT ratio (> 1.2 for every volume tested) and another due to lack of consent. One subject was included 5 times during ARDS. Eight recordings were analyzed.

Parameters including V_bestXP, E_ew, V_bestOT, and P_es (insp/exp) were analyzed. V_bestXP and E_ew were identified in all 8 recordings. V_bestOT and V_bestXP were comparable (1.6 ± 0.6 mL and 1.9 ± 0.8 mL, respectively, P = .21) (Fig. 2A) with a high correlation (R² = 0.84). P_esexp at V_bestXP and V_bestOT were closely related (R² = 0.78; Fig. 2B). V_bestXP and V_bestOT varied between subjects and were different from the standardized fixed volume of 4 mL (V4) (P < .01). The OT ratio at V4 was 0.9 ± 0.1 (range 0.8–1.2). Furthermore, P_es (exp/insp) computed at V_bestOT and V4 were different, with P_es at V4 being > P_es at V_bestOT, with P_es,expOT at 10.4 ± 3 and P_es,expV4 at 15.5 ± 3.9 cm H₂O (P = .02) and P_es,inspOT at 16.3 ± 3.9 and P_es,inspV4 at 19 ± 3.9 cm H₂O (P = .01). Figure 2C illustrated the discrepancy between P_es,expOT and P_es,expV4, (R² = 0.34). Intra-individual variability in volume calibration was observed in subject 4, with V_bestXP ranging from 1.5–2.3 mL over 5 days of ARDS evolution.

Fig. 2.

Reliability of experimental data versus reference. A: Scatterplot of the V_bestOT versus V_bestXP relationship (dots) superimposed with its associated linear fit (dashed line). B: Scatterplot of the P_es,expOT versus P_es,expXP relationship (dots) superimposed with its associated linear fit (dashed line). C: Scatterplot of the P_es,expOT versus P_es,expV4 relationship (dots) superimposed with its associated linear fit (dashed line). Linear coefficients, intercepts, and R² statistics are displayed for each panel. V = volume; OT = occlusion test; XP = experimental slow-flow method variables; P_es = esophageal pressure; exp = expiratory; V4 = 4 mL volume.

We found that it took 8 ± 1 min (4.8 min for the syringe filling process alone) to set up the experimental part and record data for the PV method. Furthermore, the average time to perform OT at each step (n = 10) and compute the OT ratio closest to 1 was 12 ± 3 min (for trained practitioners). The time required to perform the PV procedure was significantly shorter than that for the OT (P = .01).

Discussion

In this report, we propose a new process to evaluate the optimal esophageal balloon among 4 subjects compared to the standard methodology of Mojoli et al. V_bestXP and V_bestOT were similar in our cohort (Fig. 2A), and dynamic esophageal balloon PV curve reproduced graphically static esophageal balloon PV curve framework described by Mojoli et al, with a high-resolution added value⁷ (Fig. 1B–C). We found that our process was less time consuming than the complete Mojoli et al procedure, which required setting various volumes and performing Baydur tests at each level.

Our approach advocates for individual volume assessment instead of standardized volume (4 mL), which may lead to an overestimation of P_es (Fig. 2C).^7,8 It should be noted that our population showed lower calibration volumes compared to those found in literature,^7–9 as well as more variable esophageal wall elastance (range 0.6–4 cm H₂O/mL). OT ratio at V4 was in acceptable range in our cohort (0.8–1.2), but our data of P_es values captured at different volumes (V_best and V4) were significantly different.^1–3,7 Beyond inter-individual variability, we can also speculate on the possibility of an intra-patient variability as suggested by the several measurements performed in the same subject during the course of ARDS (optimal volumes ranging from 1.4–2.3 mL).

Our study had a descriptive and exploratory nature, aimed at assessing the feasibility of the PV protocol. Due to the small sample size and the fact that it was conducted in a single-center, our experimental protocol needs to be validated in a larger cohort. We can note that previous studies did not perform direct pleural manometry when using static or dynamic balloon volume PV curve to guide volume calibration. Therefore, although the use of V_best for calibration seems physiologically plausible, its improved reliability for P_es monitoring remains hypothetical.^10–12 Additionally, the computation and analysis of esophageal balloon PV curve were performed off line, and informative results for volume titration were not readily available. To address this issue, future research could focus on developing dedicated software tools and considering Baydur modified OT at V_bestXP as a confirmatory rule. Our study suggests that a dynamic esophageal balloon inflation procedure may help monitor P_L with greater accuracy and simplicity in clinical practice. Personalized and repeated measurements are important due to observed inter- and intra-individual variability of optimal balloon volume. Our preliminary results are exploratory but can contribute to refining P_es monitoring, leading to potential automation.