A noninvasive bedside imaging technique can individually calibrate positive end-expiratory pressure settings in patients on extracorporeal membrane oxygenation (ECMO) for severe acute respiratory distress syndrome (ARDS), a study showed.
The step-down PEEP (positive end-expiratory pressure) trial could not identify a single PEEP setting that optimally balanced lung overdistension and lung collapse for all 15 patients. But, electrical impedance tomography (EIT) allowed investigators to individually titrate PEEP settings for each patient, Guillaume Franchineau, MD, wrote (Am J Respir Crit Care Med. 2017;196[4]:447-57. doi: 10.1164/rccm.201605-1055OC).
Samir/Wikimedia Commons/CC ASA-3.0
The 4-month study involved 15 patients (aged, 18-79 years) who were in acute respiratory distress syndrome for a variety of reasons, including influenza (7 patients), pneumonia (3), leukemia (2), and 1 case each of Pneumocystis, antisynthetase syndrome, and trauma. All patients were receiving ECMO with a constant driving pressure of 14 cm H2O. After verifying that the inspiratory flow was 0 at the end of inspiration, PEEP was increased to 20 cm H2O (PEEP 20) with a peak inspiratory pressure of 34 cm H2O. PEEP 20 was held for 20 minutes and then lowered by 5-cm H2O decrements with the potential of reaching PEEP 0.
The EIT device, consisting of a silicone belt with 16 surface electrodes, was placed around the thorax aligning with the sixth intercostal parasternal space and connected to a monitor. By measuring conductivity and impeditivity in the underlying tissues, the device generates a low-resolution, two-dimensional image. The image was sufficient to show lung distension and collapse as the PEEP settings changed. Investigators looked for the best compromise between overdistension and collapsed zones, which they defined as the lowest pressure able to limit EIT-assessed collapse to no more than 15% with the least overdistension.
There was no one-size-fits-all PEEP setting, the authors found. The setting that minimized both overdistension and collapse was PEEP 15 in seven patients, PEEP 10 in six patients, and PEEP 5 in two patients.
At each patient’s optimal PEEP setting, the median tidal volume was similar: 3.8 mL/kg ideal body weight for PEEP 15, 3.9 mL/kg ideal body weight for PEEP 10, and 4.3 mL/kg ideal body weight for PEEP 5.
Respiratory system compliance was also similar among the groups, at 20 mL/cm H2O, 18 mL/cm H2O, and 21 mL/cm H2O, respectively. However, arterial partial pressure of oxygen decreased as the PEEP setting decreased, dropping from 148 mm Hg to 128 mm Hg to 100 mm Hg, respectively. Conversely, arterial partial pressure of CO2 increased (32-41 mm Hg).
EIT also allowed clinicians to pinpoint areas of distension or collapse. As PEEP decreased, there was steady ventilation loss in the medial-dorsal and dorsal regions, which shifted to the medial-ventral and ventral regions.
“Most end-expiratory lung impedances were located in medial-dorsal and medial-ventral regions, whereas the dorsal region constantly contributed less than 10% of total end-expiratory lung impedance,” the authors noted.
“The broad variability of EIT-based best compromise PEEPs in these patients with severe ARDS reinforces the need to provide ventilation settings individually tailored to the regional ARDS-lesion distribution,” they concluded. “To achieve that goal, EIT seems to be an interesting bedside noninvasive tool to provide real-time monitoring of the PEEP effect and ventilation distribution on ECMO.”
Positive PEEP trial, but questions remain
This first study to examine EIT in patients under extracorporeal membrane oxygenation shows important clinical potential, but also raises important questions, Claude Guerin, MD, wrote in an accompanying editorial. (Am J Respir Crit Care Med. doi: 10.1164/rccm.201701-0167ed).
The ability to titrate PEEP settings to a patient’s individual needs could substantially reduce the risk of lung derecruitment or damage by overdistension.
The current study, however, has limitations that must be addressed in the next phase of research, before this technique can be adopted into clinical practice, noted Dr. Guerin, a pulmonologist at the Hospital de la Croix Rousse, Lyon, France. The 5-cm H20 PEEP steps may be too large to detect relevant changes, he said.
In several other studies, PEEP was reduced more gradually in 2- to 3-cm H2O increments. “Surprisingly, PEEP was reduced to 0 cm H2O in this study, with this step maintained for 20 minutes, raising the risk of derecruitment and further stretching once higher PEEP levels were resumed.”
The investigators did not perform any recruitment maneuvers before proceeding with PEEP adjustment. This is contrary to what has been done in prior animal and human studies.
The computation of driving pressure was done without taking total PEEP into account. “As total PEEP is frequently greater than PEEP in patients with [acute respiratory distress syndrome], driving pressure can be overestimated with the common computation.”
The optimal PEEP that the investigators aimed for was determined retrospectively from an offline analysis of the data; this technique would not be suitable for bedside management. “When ‘optimal’ PEEP was defined from [EIT criteria], from a higher PaO2 [arterial partial pressure of oxygen] or from a higher compliance of the respiratory system during the decremental PEEP trial, these three criteria were observed together in only four patients with [acute respiratory distress syndrome].”
The study was done only once and cannot comply with the need for regular PEEP-level assessments over time, as could be done with some other strategies.
“Further studies should also consider taking into account the role of chest wall mechanics,” Dr. Guerin said.
Nevertheless, he concluded, EIT-based PEEP titration for each individual patient represents a prospective tool for assisting with the treatment of acute respiratory distress syndrome, and should be fully investigated in a large, prospective trial.
Dr. Franchineau reported receiving speakers fees from Mapquet. Dr. Guerin had no relevant financial disclosures.