The use of Positive End Expiratory Pressure (PEEP) is a corner stone in the management of Acute Respiratory Distress Syndrome (ARDS), and was linked to decreased mortality in patients with ARDS [1]. Optimum PEEP value can be considered as the one which improves lung recruitment and hence oxygenation as well as ventilation, without compromising patient's hemodynamics or inducing alveolar hyperinflation.

Several methods were investigated to determine the optimal PEEP value in ARDS patients, which reflects the heterogeneity in the pathology of ARDS. These methods include, but not limited to, the use of multiple pressure volume curves [2] measurement of lung volumes, [3] ARDS Clinical Network Mechanical ventilation protocol (ARDSNet) FiO₂- PEEP combination, [4] and others. However, there is no gold standard method for determining such optimal value.

Volumetric capnogram can be used to monitor VCO₂ during mechanical ventilation in ARDS patients. It is postulated that increments of PEEP values cause a decrease in VCO₂, which recover quickly to baseline if not associated with a compromise of the pulmonary perfusion (cardiac output). Optimum PEEP may therefore be considered the maximal PEEP value which was not associated with a decrease in cardiac output and hence CO₂ delivery and elimination which can be assessed by VCO₂ value. Thus, when a decrease in VCO₂ with failure to recover to baseline is observed, the preceding PEEP value can be considered as the optimum PEEP. Our study aimed at comparing the use of VCO₂ obtained from volumetric capnogram to detect optimal PEEP vs. traditional ARDSNet FiO₂- PEEP combination, through its effects on oxygenation, alveolar ventilation and static compliance.

This prospective randomized controlled study was conducted in Anesthesia and Surgical Intensive Care Department of a university hospital during the period from March 2016 to March 2017. This is a tertiary hospital in a central catchment area covering an estimated population of 5 million citizens. The study was approved by the institutional review board of faculty of medicine and university hospitals, Tanta University with code: 30452/08/15.

Over the study period, all patients admitted to the surgical intensive care unit (SICU) who were mechanically ventilated via an oro-tracheal tube and were screened. Patients fulfilling Berlin definition [5] of ARDS were registered. Patients with known history of cardiac disease, acute coronary syndromes or low functional capacity; history of hepatic disease, alcohol intake or positive virology; history of chronic kidney diseases or rising creatinine; with unstable hemodynamics in term of mean arterial blood pressure if less than 65 mm Hg or on vasopressors were excluded from the study. If a patient fulfilled the criteria, explanation to the family was done by the SICU doctor then a written informed consent from participants' legally authorized representative, according to national regulations, was obtained.

The patients were randomized into two groups, on a ratio of 1:1, using end-user computer-based randomization software. Samples, data input into ventilator and PEEP titration were done by the principal investigator.

Sixty patients who met the previous criteria were enrolled in the study. Patients were randomly allocated in equal proportions of 30 patients to each of the studied groups:

Group A

In this group, after baseline ventilation, optimum PEEP was determined using FiO₂-PEEP combination. Fraction of inspired oxygen (FiO₂) and PEEP were titrated based on the FiO₂-PEEP combination chart (Table 1) every 20 minutes. A minimum PaO₂ of 55-80 mmHg or SpO₂ 88-95% was targeted as recommended by ARDSNet trials [4]. However, up titration was continued until reaching plateau oxygen saturation. The least combination that produces same oxygen saturation was considered optimal.

Table 1: FiO₂-PEEP combinations [5].

Group B

In this group, after baseline ventilation, the same oxygenation levels were targeted as recommended by ARDSNet trials. PEEP was increased in increments of 2 cmH₂O every 20 minutes and VCO₂ was monitored. Once it failed to recover to baseline, the preceding PEEP value was considered optimum.

Baseline ventilation, monitoring and PEEP titration protocol:

All patients were mechanically ventilated using Engström Carestations (General Electric, New York, USA). Basic monitoring for patients were done using BSM-2301K monitors (Nihon Kohden, Tokyo, Japan). A radial arterial catheter and triple-lumen central venous catheter were inserted for frequent sampling of arterial blood gas and central venous blood gas analysis, respectively using the AVL-988 multi-gas analyzer (Roche, Basel, Switzerland).

Adequate sedation of all patients (Richmond agitation sedation scale score -5) [6] was achieved with continuous infusions of midazolam 0.1 mg/kg/h. Bolus muscle relaxant injection of 3 mg cisatracurium were injected as required during PEEP titration. Maintenance intravenous fluids were infused and fluid responsiveness was monitored regularly with passive leg raising test and bolus fluid infused as required. All patients were kept in supine position.

Ventilation settings were adjusted according to the ARDSNet protocol [5]. FiO₂ was initially set at 0.4 and PEEP at 5cmH₂O then recruitment maneuver in the form of a sustained application of PEEP at 40cmH₂O for 40 seconds was performed. PEEP was then titrated in steps of 2 cmH₂O according to method used in each group.

During the procedure, baseline hemoglobin level was obtained as well as arterial blood gas and central venous blood gas samples at each step as input data for volumetric capnogram module to calculate VA. In addition, patients were monitored for signs of hemodynamic instability in the form of hypotension with a mean blood pressure of less than 65 mmHg and/or pneumothorax.

The primary outcome was the PaO₂/FiO₂ while secondary outcome was: CO₂ production VCO₂, alveolar ventilation (VA), static compliance, MAP and complications if occurred with modulation of initial settings.

The sample size was calculated using Epi-Info software statistical package created by World Health organization and center for Disease Control and Prevention (Atlanta, Georgia, USA) version 2002. The sample size was calculated at N=24 and approximated to 30

The criteria used for sample size calculation were as follows:

95% confidence limit

80% power

The ratio between experimental and control groups is 1:1

Expected outcome in treatment group is double times better than control groups. (40-80% of optimal required) Statistical analysis was done by an independent statistician, who was blinded to allocation and methodology used in each group. The collected data were organized, tabulated and statistically analyzed, using SPSS (Statistical Package for Social Studies) version 19, created by IBM (Illinois, Chicago, USA).

For numerical values the range, mean, and standard deviations were calculated. The differences between mean values of the two studied groups were tested using student's t test. Differences of mean values at baseline and end of intervention were tested using paired t test. For categorical variables, the number and percentage were calculated and analyzed using chi square test. The level of significance was adopted at p <0.05.

The enrollment and allocation of patients are shown in Figure 1.

Figure 1: Study flow chart - patient enrollment and allocation.

Sixty patients who fulfilled the Berlin definition of ARDS [5] were enrolled. Demographic data and ARDS grade for each registered patient was collected (Table 1).

Significantly higher values of PEEP were applied in group B (p <0.001), while significantly higher values of FiO₂ were applied in group A (p <0.001). Both groups showed a significant increase in SpO₂ and PaO₂ from baseline (p <0.001 in both groups), with no significant difference between the two groups. However, PaO₂/FiO₂ values were significantly higher in group B compared with group A (p <0.001). The static compliance in both groups showed a significant increase from baseline values (p <0.001 in both groups) and was significantly higher in group B compared to group A (p <0.004) (Table 2).

Table 2: Comparison between demographic data and ARDS grade between the studied groups.

The baseline values of VCO₂, VA and MAP were comparable in both groups. VA values were significantly improved in both groups (p <0.001 in both groups), with no significant difference between the two groups. Neither VCO₂ nor MAP significantly changed from baseline in both groups, with no significant difference between the two groups (Table 3).

Table 3: Comparison between measured data between the studied groups at baseline and end of intervention. Values are expressed as mean (± SD).

The use of fixed PEEP levels based on FiO₂ values according to ARDSNet protocol does not consider the heterogeneity of ARDS pathology among different patients [7]. Despite its criticism, it has been used successfully in all ARDSNet trials [8] and hence it was used in the current study. However, the current study targeted higher SPO₂ than recommended by ARDSNet trials. This modification was based on Villar et al. [9] who reviewed publications from the last 2 years and the current guidelines dealing with ARDS. They recommended a target SPO₂ of 90–97%. This higher target was also suggested by Bein et al. [10].

Application of PEEP in ARDS patients to open potentially recruitable alveoli does not only improve oxygenation, but also augments CO₂ elimination. The approach of the current study to detect optimal PEEP depended on monitoring effect of PEEP on CO₂ elimination rather than oxygenation which better reflects PEEP related hemodynamic compromise. Measuring VCO₂ is a simple bedside tool, with superiority over end-tidal CO₂ in evaluation of CO₂ elimination [11,12].

Our findings showed that determination of optimum PEEP by monitoring VCO₂ is associated with improvement of oxygenation as well as lung compliance, without hypotension or pneumothorax, compared with use of FiO₂-PEEP combination in ARDS patients.

PEEP application is known to improve oxygenation in ARDS patients [13]. The higher values of PEEP used in group B in combination with lower values of FiO₂, resulted in higher values of PaO₂/FiO₂, compared with group A.

In agreement with current study results, a meta-analysis by Breil et al. [14] involving 2299 patients reported that the use of PEEP values higher than suggested by ARDSNet trials was associated with improvement in lung aeration as well as clinical improvement, compared with lower values. Although statistically significant (p < 0.001), some may argue that PEEP difference between two groups may not be considered clinically significant (10.87 (± 2.35) in group B vs. 9.20 (1.13) in group A), yet the combination allowed the use of significantly lower FiO₂ (57.00 (± 8.37) in group A vs. 40 (± 0.0) in group B; p < 0.001). A meta-analysis by Briots et al. [15] on 2312 patients previously enrolled in ARDSNet trials, recommended against the use of high FiO₂ values in ARDS patients due to its association with high mortality rates.

There is no significant difference between both groups, as regard effect on VCO₂ (p= 0.552), which is in agreement with Johnson et al. [12] who found no significant effect on VCO₂ with application of PEEP value of 10 cmH2O in anesthetized patients.

Contrary to our findings, Tusman et al. [16] reported an increase of VCO₂ with incremental increase of PEEP (2 cmH2O/min) as a part of recruitment maneuver in morbidly obese patients. This apparent disagreement may be explained by an increase of VCO₂ in the early non-steady state and accelerated response of VCO₂ by recruitment maneuver applied. In addition, the increase in VA from baseline value after increasing PEEP, noted in our study in both groups, could be reflected on VCO₂ if the period was longer than 20 minutes.

The increase of VA in response to an increase of PEEP values, as noticed in both groups in our study, is in agreement with Tusman et al. [16] However, the opposite was reported by them in their earlier animal study, [17] occurring after 10 minutes. This difference in results may be due to their earlier measurement of VA which possibly didn't allow enough time for VA recovery and increase from initial level.

The lack of significant difference in VA between the two groups (p=) in the present study may be explained by the fact that Group A may also have benefited from recruitment maneuvers at the beginning of the study and increasing PEEP levels which were significantly higher than baseline (p ≤ 0.001). It may, also, be due to inability of the current study to include severe ARDS patients, who could demonstrate maximum response to any interventions, which is considered one of the current study limitations.

The improvement of the static compliance with use of PEEP value determined by VCO₂ in our study was also observed by Maisch et al. [18] who reported that use of static compliance for optimal PEEP determination was superior to the use of PaO₂ in patients undergoing cardiac surgery. This can be explained by remodeling of worsened sigmoid volume-pressure relationship in ARDS by the effect of PEEP, which maintains alveoli open, and prevent flattening of volumepressure curve, as reported by Coruh et al. [19].

No hypotension or pneumothorax was observed in the current study, even with the higher values of PEEP which were used in group B. The lack of detrimental hemodynamic effect may be explained by the work of Gattinoni et al. [20] who reported a decrease in pulmonary artery pressure with no significant adverse effect on the cardiac output, secondary to improvement of lung recruitment. Lung recruitment is accompanied by pulmonary vessels recruitment, which improves right ventricular function, and left ventricular function consequently.

This could suggest that VCO₂ monitoring was useful to avoid the side effects of increasing PEEP on cardiac output. Changes in VCO₂ were observed earlier than any change in MAP.

In contrast, Chikhan et al. [21] reported that high PEEP values up to 20 cm H2O improve arterial oxygenation, but can compromise oxygen delivery to tissues due to decrease in cardiac output. This may be explained by the effect of high rather than optimum PEEP value.

Adding to the previously mentioned limitation regarding lack of recruitment of severe ARDS patients, the different protocol of PEEP titration used in both groups made blinding not feasible. While that may reflect a potential bias, data processing and calculation were done electronically using the ventilator and the statistical analysis was blinded. Another limitation is the inability to divide patients into subgroups according to their ARDS grade due to the small sample size which could otherwise deprive any results of statistical significance.

Based on our results, we conclude that optimum PEEP determined by volumetric capnogram (VCO₂ monitoring) is associated with improvement of oxygenation as well as lung compliance, compared with FiO₂-PEEP combination in ARDS patients without resulting in increased complications.

I would like to thank Prof. Ibrahim Elkabbash, Professor of public health, Faculty of Medicine, Tanta University for his participation and guidance in statistical analysis of data collected in study.