OPTIMUM LEVEL OF POSITIVE END-EXPIRATORY PRESSURE IN ACUTE RESPIRATORY DISTRESS SYNDROME CAUSED BY INFLUENZA A(H1NI)PDM09: BALANCE BETWEEN MAXIMAL END-EXPIRATORY VOLUME AND MINIMAL ALVEOLAR OVERDISTENSION.

The Aim: to determine optimum level ofpositive end-expiratory pressure (PEEP) according to balance between maxi- mal end-expiratory lung volume (EEL V)(more than predicted) and minimal decrease in exhaled carbon dioxide volume (VCO) and then to develop the algorithm of gas exchange correction based on prognostic values of EEL K; alveolar recruitability, PA/FiO2, static compliance (C,,,) and VCO2.

Materials And Methods: 27 mechanically ventilatedpatients with acute respiratory distress syndrome (ARDS) caused by influenza A (HINJ)pdm09 in Moscow Municipal Clinics ICU's from January to March 2016 were included in the trial. At the beginning of the study patients had the following characteristic: duration offlu symptoms 5 (3-10) days, p.0/FiO2 120 (70-50) mmHg. SOFA 7 (5-9), body mass index 30.1 (26.4-33.8) kg/m², static compliance of respiratory system 35 (30-40) ml/mbar: Under sedation and paralysis we measured EELV, C VCO and end-tidal carbon dioxide concentration (EtCO) (for CO₂ measurements we fixed short-term values after 2 min after PEEP level change) at PEEP 8, 11,13,15,18, 20 mbar consequently, and incase of good recruitability, at 22 and 24 mbar. After analyses of obtained data we determined PEEP value in which increase in EELV was maximal (more than predicted) and depression of VCO₂ was less than 20%, change in mean blood pressure and heart rate were both less than 20% (measured at PEEP 8 mbar). After that we set thus determined level of PEEP and didn't change it for 5 days.

Results: Comparision of predicted and measured EELV revealed two typical points of alveloar recruiment: the first at PEEP 11-15 mbar, the second at PEEP 20-22 mbar. EELV measured at PEEP 18 mbar appeared to be higher than predicted at PEEP 8 mbar by 400 ml (approx.), which was the sign of alveolar recruitment-1536 (1020-1845) ml vs 1955 (1360-2320) ml, p=0,001, Friedman test). we didn't found significant changes of VCO₂ when increased PEEP in the range from 8 to 15 mbar (p>0.05, Friedman test). PEEP increase from 15 to 18 mbar and more lead to decrease in VCO₂ (from 212 (171-256) ml/min to 200 (153-227) ml/min, p<0,0001, Friedman test, which was the sign of overdistension. Next decrease of VCO₂ was observed at PEEP increase from 22 to 24 mbar (from 203 (174-251 ml/min) to 185 (182-257) ml/min, p=0.0025, Friedman test). Adjusted PEEP value according to balance between recruitment and overdistension was higher than the one initially set (16(15-18) mbar vs 12(7-15) mbar, p <0.0001). We observed increase of SpO₂ from 93 (87-96) to 97(95-100)% (p<0.0001 followed by decrease in inspiratory oxygen fraction from 60(40-80) to 50(40-60)%(p<0.0001). Low EELV VCO₂ and VCO₂/EtCO₂ at PEEP 8 mbar has low predictive value for death (AUROC 0,547, 0706 and 0.596, respectively).Absolute EELV value at PEEP 18 and 20 mbar were poor predictors of mortality (AUROC 0.61 and 0.65 respectively) Alveolar recruit ability was measured by subtraction of EELV at PEEP 20 and at PEEP II mbar - value below 575 ml was a good predictor of death (sensitivity 75%, specificity 88%, AUROC 0.81). Lowering of VCO₂ at PEEP 20 mbar to less than 207 ml/min was a marker of alveolar overdistension and associated with poor prognosis (sensitivity 83%, specificity 88%, AUROC 0,89). C has poor predictive value at PEEP 8 and 20 mbar (AUROC 0,58 and 0,74 respectively.

Conclusion: PEEP adjustment in ARDS due to influenza A (H1N1) pdm09 in accordance with balance between recruitment and overdistension (based on EELV and VCO measurements) can improve gas exchange, probably, not leading to right ventricular failure. This value of "balanced" PEEP is in the range between 15 and 18 mbar: Low lung recruitabiilty is associated with poor prognosis. Measurements of EELV and VCO₂ at PEEP 8 and 20 mbar can be used to make a decision on whether to keep "high" PEEP level or switch to extracorporeal membrane oxygenation in patient with ARDS due to influenza A (N1H1).