Introduction

Acute respiratory distress syndrome (ARDS) is a severe form of respiratory failure in critically ill patients, associated with high mortality rates of 30–40%1,2. Patients with ARDS commonly develop hypoxemia and require invasive mechanical ventilation, which can itself cause ventilator-induced lung injury if not carefully managed3,4. Lung-protective ventilation strategies—including low tidal volume and adequate positive end-expiratory pressure (PEEP), are recommended, particularly for severe ARDS5,6,7. Low tidal volume ventilation has been shown to reduce mortality compared with higher tidal volumes (10–12 mL/kg)8,9. Inappropriately high PEEP may cause alveolar overdistension, barotrauma, or hemodynamic compromise, while insufficient PEEP can result in alveolar collapse and hypoxemia10. However, the optimal strategy for adjusting PEEP remains uncertain11.

Recent international guideline emphasize the importance of applying higher or individualized PEEP strategies in patients with moderate-to-severe ARDS to optimize alveolar recruitment while minimizing overdistension5. However, despite these advances, the optimal approach to PEEP selection remains uncertain because of substantial interindividual variability in lung mechanics and recruitability. To address this challenge, multiple approaches for PEEP titration have been explored, ranging from conventional FiO₂/PEEP tables and compliance-based adjustment to advanced physiologic methods such as stress index monitoring, esophageal pressure guidance, and the recruitment-to-inflation ratio, as well as imaging-based techniques including computed tomography, lung ultrasound, and electrical impedance tomography (EIT)12,13,14,15. Among these approaches, EIT has emerged as a promising, non-invasive bedside tool for optimizing ventilatory management16,17. By allowing clinicians to visualize alveolar recruitment and overdistension simultaneously, EIT offers the potential for individualized PEEP titration18,19. Early studies have suggested that EIT-guided PEEP may improve physiologic parameters20. Recent trials by Jimenez et al. and the RECRUIT study further supported physiologic-guided PEEP titration to optimize lung recruitment and reduce mechanical power21,22. Similarly, Sella et al. and Somhorst et al. reported more homogeneous ventilation and oxygenation with EIT-guided strategies without increasing adverse events23,24. However, these findings are primarily limited to short-term physiologic outcomes in small cohorts, and it remains unclear whether such improvements translate into meaningful clinical benefits, including reduced mortality, fewer ventilator days, or shorter ICU stays. EIT-guided strategies have not been widely validated or systematically compared with conventional ARDSnet-guided PEEP.

Considering these limitations, this randomized controlled study aimed to investigate whether PEEP titration guided by EIT could improve oxygenation, respiratory mechanics, and other clinically relevant outcomes compared with the low PEEP/FiO₂ strategy in patients with moderate-to-severe ARDS.

Methods

This single-center randomized trial was conducted at Bach Mai Hospital, Vietnam, between 2024 and 2025. The protocol was approved by the Bach Mai Hospital Ethics Committee (5532/BM-HDDD), and written informed consent was obtained from all patients. The study adhered to the Declaration of Helsinki and CONSORT guidelines. The trial was registered at ClinicalTrials.gov (NCT06733168; registered on 13/12/2024).

Patient population

Eligible patients were ≥ 18 years old, intubated, and diagnosed with moderate-to-severe ARDS according to the New Global Definition25, with a PaO₂/FiO₂ ratio ≤ 200 mmHg. Exclusion criteria included undrained or newly developed pneumothorax; unstable hemodynamics defined as mean arterial pressure (MAP) < 60 mmHg unresponsive to resuscitation and/or heart rate < 60 bpm; contraindications to EIT (e.g., pacemakers, automatic external defibrillators, chest trauma, or recent chest surgery preventing EIT belt application); pregnancy; and severe neuromuscular disease.

Randomization

Patients were randomized 1:1 to either the EIT group or the low PEEP/FiO₂ group using simple randomization. The allocation sequence was generated by a computer-based random number list.

Intervention

EIT group

Baseline: The EIT device (PulmoVista 500, Dräger Medical, Lübeck, Germany) was continuously monitored via a silicone belt with 16 electrodes positioned around the thorax at the 4th–6th intercostal space; the reference electrode was placed on the abdomen26. Before recruitment maneuvers, patients were ventilated according to the low PEEP/FiO₂ strategy for 10 min, targeting SpO₂ 88–95%, PaO₂ 55–80 mmHg, and MAP ≥ 65 mmHg. Patients were positioned supine and received deep sedation with neuromuscular blockade to maintain a Richmond Agitation-Sedation Scale (RASS) score less than or equal to − 3, ensuring complete suppression of spontaneous respiratory efforts. Mechanical ventilation was then switched to pressure-controlled ventilation (PCV) mode with a driving pressure of 15 cmH₂O, while all other ventilatory parameters were kept unchanged.

Step 1: PEEP was increased in 5 cmH₂O increments every minute from 10 to 15, 20, and 25 cmH₂O, with a maximum pressure limit of 40 cmH₂O.

Step 2: After the recruitment maneuver at the final PEEP level, recommended PEEP was determined. PEEP was first set to 20 cmH₂O and then decreased by 2 cmH₂O every 30 s to 6 cmH₂O. Recommended PEEP was defined as the intersection of the overdistension and collapse curves as measured by EIT (Fig. 1).

Fig. 1
Fig. 1
Full size image

Determination of recommended PEEP using EIT-based analysis. The recommended PEEP level is identified at the intersection between the cumulative collapse (white) and overdistension (yellow) curves, representing the balance point that minimizes both atelectasis and overinflation. In this example, the recommended PEEP was determined to be 16 cmH₂O.

Step 3: Once recommended PEEP was identified, a second recruitment maneuver was performed at 25 cmH₂O for 1 min, followed by setting PEEP at the recommended level.

Discontinuation criteria: Recruitment maneuvers were stopped if persistent hypotension (MAP decrease > 15 mmHg) or persistent hypoxemia (SpO₂ < 85% for ≥ 1 min) occurred. Ventilator settings were then returned to pre-maneuver values. Recruitment was repeated during the acute phase if SpO₂ dropped < 90% and did not improve with PEEP/FiO₂ adjustments, or after airway pressure loss (e.g., suctioning, nebulization, bronchoscopy, or ventilator disconnection).

PEEP was maintained at a constant level for 24 h after titration unless emergency adjustments were required (SpO₂ < 90% or hemodynamic deterioration). Thereafter, PEEP was gradually reduced (by 2 cmH₂O every 6–8 h) when the PaO₂/FiO₂ ratio exceeded 200 with FiO₂ < 0.5.

Control group

Routine recruitment maneuvers were not performed. PEEP was set according to the same low FiO₂–PEEP table of the ARDSnet protocol to achieve SpO₂ 88–95% and PaO₂ 55–80 mmHg.

Data collection

Primary outcomes were oxygenation (PaO₂/FiO₂) and respiratory mechanics (static compliance). Secondary outcomes included in-hospital and 28-day mortality, ventilator-free days to discharge, ICU length of stay, pneumothorax, use of rescue therapies (prone positioning, ECMO, tracheostomy, corticosteroids), and sequential organ failure assessment (SOFA) scores.

Statistical analysis

Data were analyzed using SPSS version 22.0 (IBM Corp., Armonk, NY, USA). Categorical variables were expressed as frequency and percentage, while continuous variables were reported as mean ± standard deviation if normally distributed, or as median (interquartile range) if not. Between-group comparisons were performed using the independent Student’s t-test for normally distributed variables and appropriate non-parametric tests for non-normally distributed variables (Mann–Whitney U test). Within-group comparisons were assessed using paired t-tests.

Because oxygenation and respiratory mechanics parameters were evaluated before and after intervention (PEEP titration) in both groups, two-way repeated-measures ANOVA was used to assess the effects of group (EIT vs. control) and time (before vs. after intervention), followed by Bonferroni’s post hoc test for multiple comparisons. Kaplan–Meier survival analysis was applied to compare 28-day survival between groups, and differences were assessed using the log-rank test. Subgroup analyses were performed to evaluate changes in oxygenation (ΔPaO₂/FiO₂) and static compliance of the respiratory system (ΔCstat) stratified by ARDS severity (moderate vs. severe) and by BMI categories, using the independent Student’s t-test. A two-sided p-value < 0.05 was considered statistically significant.

Results

Of 150 patients screened, 116 were randomized (58 to EIT and 58 to control). Two patients in the EIT group and four in the control group were withdrawn (< 24 h recommended PEEP), and two were excluded due to missing outcome data. Thus, 108 patients (56 in the EIT group and 52 in the control group) were included in the final analysis (Fig. 2). No significant baseline differences were observed between groups in patient characteristics, SOFA scores, comorbidities, ARDS severity, or etiology (Table 1).

Fig. 2
Fig. 2
Full size image

Flow chart of the study.

Table 1 Baseline demographic and clinical characteristics of patients.
Full size table

Oxygenation results are shown in (Fig. 3). Both PaO₂ and the PaO₂/FiO₂ ratio improved after PEEP optimization in both groups. On day 1, the EIT group had a significantly higher PaO₂/FiO₂ (adjusted p = 0.04), with a significant group–time interaction [F(1, 106) = 6.99, p = 0.01; Fig. 3A]. Trends toward higher PaO₂ on days 1 and 2—and higher PaO₂/FiO₂ on day 2—were observed in the EIT group compared with the control group, although these differences were not statistically significant (Fig. 3A, B).

Fig. 3
Fig. 3
Full size image

Oxygenation. Ratio of arterial partial pressure of oxygen to fraction of inspired oxygen (PaO₂/FiO₂; A) and arterial partial pressure of oxygen (PaO₂; B). Data are presented as mean ± SD. Differences were analyzed using two-way repeated-measures ANOVA with Bonferroni’s post hoc test. *p < 0.05 (between groups).

Respiratory mechanics parameters are shown in (Fig. 4). After titration, the EIT group demonstrated a wider range and a higher median PEEP level compared with the control group (Fig. 4A; p = 0.03, Mann–Whitney U test). Plateau pressure did not differ significantly between groups (Fig. 4B). Static compliance was significantly higher in the EIT group on day 1 (adjusted p = 0.01; group–time interaction [F(1, 106) = 7.08, p = 0.009]) and day 2 (adjusted p = 0.05; group–time interaction [F(1, 106) = 4.23, p = 0.04]) (Fig. 4C). Driving pressure was significantly lower in the EIT group on day 1 (adjusted p < 0.001; group–time interaction [F(1, 106) = 9.99, p = 0.002]) and day 2 (adjusted p = 0.03; group–time interaction [F(1, 106) = 4.42, p = 0.04]) (Fig. 4D).

Fig. 4
Fig. 4
Full size image

Respiratory mechanics. Positive end-expiratory pressure (PEEP; A), plateau pressure (Pplat; B), static compliance of the respiratory system (Cstat; C), and driving pressure (Pdriv; D). Data are presented as median [range] or mean ± SD. Differences were analyzed using the Mann–Whitney U test (A) and two-way repeated-measures ANOVA with Bonferroni’s post hoc test (B–D). *p < 0.05, **p < 0.001 (between groups).

Clinical outcomes are shown in (Table 2 and Fig. 5). ΔSOFA at days 1 and 2 decreased in both groups, with a significantly greater reduction in the EIT group. Twenty-eight-day mortality was lower in the EIT group than in the control group (28.6 vs. 44.2%), although the difference was not statistically significant (p = 0.09; Fig. 5). Rates of successful weaning, ICU length of stay, duration of mechanical ventilation, and the incidence of barotrauma, ECMO use, tracheotomy, prone positioning, corticosteroid administration, and neuromuscular blocker use were similar between groups.

Table 2 Other outcomes and ventilation strategies in the two groups.
Full size table
Fig. 5
Fig. 5
Full size image

Kaplan–Meier 28-day survival curves for the EIT and control groups.

Subgroup analyses are shown in (Fig. 6). When stratified by ARDS severity, the benefit of EIT in improving PaO₂/FiO₂ (ΔPaO₂/FiO₂ = PaO₂/FiO₂ at follow-up – PaO₂/FiO₂ at admission) was evident on both days 1 and 2 in patients with severe ARDS. On day 1, ΔPaO₂/FiO₂ in the EIT group was 97.5 ± 61.4 mmHg, significantly higher than in the control group (55.8 ± 40.6 mmHg, p = 0.02; Fig. 6A). On day 2, ΔPaO₂/FiO₂ remained higher in the EIT group (112.9 ± 71.5 vs. 70.2 ± 61.2 mmHg, p = 0.03; Fig. 6B). Notably, a significant negative correlation was observed between PaO₂/FiO₂ at admission and improvement in PaO₂/FiO₂ in the EIT group (r = –0.43, p = 0.001). Similarly, static respiratory compliance (ΔCstat) improved significantly in the EIT group compared with the control group (Fig. 6C–D). On day 1, ΔCstat was 4.80 ± 0.60 mL/cmH₂O vs. 2.85 ± 0.41 mL/cmH₂O (p = 0.009), and this difference persisted on day 2 (5.29 ± 0.56 vs. 3.75 ± 0.53 mL/cmH₂O, p = 0.036). Subgroup analysis stratified by BMI showed no significant differences in ΔPaO₂/FiO₂ improvement between the two groups.

Fig. 6
Fig. 6
Full size image

Changes in the ratio of arterial partial pressure of oxygen to fraction of inspired oxygen (ΔPaO₂/FiO₂) and static compliance of the respiratory system (ΔCstat), subgrouped by ARDS severity on day 1 (A,C) and day 2 (B,D). Data are presented as mean ± SD. Differences were analyzed using Student’s t-test. Numbers inside the bars indicate the number of participants.

Discussion

In this randomized controlled trial, we found that an EIT-guided strategy combining recruitment maneuvers with individualized PEEP titration significantly improved respiratory mechanics, including reduced driving pressure and increased compliance, and was associated with more favorable short-term SOFA scores compared with the low PEEP–FiO₂ strategy. Importantly, beyond these physiological benefits, we also observed a trend toward improved clinical outcomes, including lower 28-day mortality and higher weaning success, although these differences did not reach statistical significance. The novelty of our findings lies in demonstrating that EIT not only optimizes respiratory mechanics but may also exert clinically relevant effects on organ function and patient-centered outcomes.

Patients managed with EIT-guided titration in our study exhibited higher PaO₂/FiO₂ within the first 24 h. This contrasts with the findings of He et al., who did not observe significant improvement in oxygenation, whereas Zhao et al. reported enhanced PaO₂/FiO₂ with EIT-guided PEEP selection in ARDS patients20,27. These discrepancies may be explained by differences in baseline severity of hypoxemia: He et al. included patients with PaO₂/FiO₂ ≤ 300 mmHg (mild-to-moderate ARDS), while Zhao et al. studied patients with more severe hypoxemia (PaO₂/FiO₂ ≤ 100 mmHg). Our cohort, with PaO₂/FiO₂ ≤ 200 mmHg, falls between these ranges, suggesting that patients may still benefit from individualized PEEP titration in terms of early oxygenation. Moreover, the observed difference in oxygenation in our study may also be partly attributable to the application of a recruitment maneuver prior to PEEP titration in the EIT-guided group. This step, which was not performed in the control group or in the study by He et al.27, may have contributed to more effective alveolar recruitment and improved baseline aeration before optimization of PEEP levels. These observations imply that the magnitude of oxygenation improvement with EIT-guided PEEP may be influenced by baseline PaO₂/FiO₂, underscoring the importance of considering ARDS severity when selecting patients for individualized ventilatory strategies. Indeed, our subgroup analysis demonstrated that patients with severe ARDS achieved significantly greater improvements in PaO₂/FiO₂ than those with moderate ARDS on both day 1 and day 2. This finding reinforces the notion that the beneficial effects of EIT-guided titration are most pronounced in patients with severe hypoxemia, who may derive greater physiological benefit from optimized alveolar recruitment and prevention of overdistension.

The magnitude of improvement has varied across studies, but the overall trend supports the concept that individualized PEEP titration can optimize alveolar recruitment. Differences in patient characteristics may partially explain this variability. In particular, our Vietnamese cohort had a lower average BMI compared with the predominantly overweight populations (BMI ≥ 25 kg/m2) studied by He and Zhao20,27. which may have influenced lung recruitability and the observed response to EIT. It is also important to recognize that PaO₂/FiO₂ is a dynamic variable influenced by multiple factors, including FiO₂ adjustments, hemodynamics, and patient positioning. Nevertheless, the consistent short-term improvement observed in our study suggests that EIT-guided strategies facilitate more effective recruitment in the acute phase of ARDS management.

One of the most notable findings of our study was the significant improvement in static respiratory compliance in the EIT group compared with the control group during the first two days after intervention. This improvement was consistently observed in subgroup analyses, particularly among patients with severe ARDS. These results emphasize the physiological benefits of EIT-guided PEEP titration, which enables clinicians to identify the optimal PEEP level that balances alveolar recruitment and prevents overdistension. By enhancing compliance and maintaining more homogeneous lung ventilation, EIT-guided titration may contribute to mitigating ventilator-induced lung injury. Our findings are consistent with previous studies reporting similar improvements in lung mechanics when using EIT to optimize PEEP settings28,29,30,31. Our data provide additional support for the integration of EIT into routine clinical decision-making, particularly for patients with heterogeneous lung pathology.

Although baseline SOFA scores did not differ between groups, patients managed with EIT-guided strategies exhibited a more favorable trajectory, with greater reductions in SOFA from baseline to day 1 and day 2. As the SOFA score is a well-validated predictor of mortality in ARDS32, this improvement suggests that enhanced ventilatory mechanics and oxygenation may translate into better short-term multi-organ function. Additional analyses of non-pulmonary SOFA changes also revealed significant between-group differences, indicating improvements beyond oxygenation and suggesting potential benefits in extra-pulmonary organ function. While the absolute differences were modest, these findings are consistent with prior evidence linking protective ventilation with reduced systemic inflammation and improved organ outcomes. Notably, He et al. also reported early reductions in SOFA with EIT-guided PEEP, supporting the potential benefit of individualized PEEP in preserving extra-pulmonary organ function27. It is plausible that individualized PEEP reduces not only pulmonary stress and strain but also systemic inflammatory responses, thereby positively influencing extra-pulmonary systems. Future studies should investigate biomarkers of inflammation and perfusion to clarify these mechanisms.

Regarding clinical endpoints, our study observed a lower, though not statistically significant, 28-day mortality in the EIT group compared with controls (28.6% vs. 44.2%; p = 0.090). Although underpowered to detect a mortality difference, this trend is clinically relevant given the established association between driving pressure and survival. Large retrospective analyses have shown that lower driving pressure correlates with improved outcomes in ARDS, while these were not interventional trials33,34. Similarly, the success rate of weaning from mechanical ventilation was numerically higher in the EIT group (42.9 vs. 36.5%; p = 0.503), although this did not reach statistical significance. Length of ICU stay and duration of mechanical ventilation were comparable between groups, in line with other trials evaluating recruitment maneuvers and PEEP titration strategies35,36. Importantly, the incidence of barotrauma, ECMO use, tracheostomy, prone positioning, and adjunctive therapies (corticosteroids, neuromuscular blockade) was similar between groups, indicating that the EIT-guided approach did not increase complications.

Our results align with previous work emphasizing the limitations of fixed PEEP-FiO₂ tables and the potential benefits of individualized PEEP titration. The ART trial, which tested aggressive recruitment maneuvers and high PEEP titration, was associated with increased mortality and barotrauma36. In contrast, current international guidelines recommend against routine or prolonged high-pressure recruitment maneuvers because of the risk of hemodynamic instability and barotrauma5,6. However, in our protocol, recruitment was performed only once before PEEP titration, using a brief and closely monitored maneuver to standardize baseline lung conditions while minimizing hemodynamic and barotrauma risk. In our approach, which used real-time regional ventilation monitoring by EIT, achieved better compliance and reduced driving pressures without increasing adverse events, as detailed in the Supplementary Material. This underscores the importance of tailoring recruitment to the individual patient’s physiology rather than applying universal high-PEEP strategies. Moreover, previous physiological studies have demonstrated that EIT can identify recruitable lung regions and prevent overdistension22. Our trial extends these findings into clinical practice, showing measurable improvements in lung mechanics and short-term organ outcomes. Importantly, although mortality differences were not statistically significant, the observed trend is consistent with the hypothesis that lowering driving pressure improves survival33.

Limitations

This study has several limitations. First, it was a single-center study with a relatively small sample size and was not powered to detect differences in mortality or long-term outcomes. Second, blinding of treating clinicians was not feasible and may have introduced bias. Third, although EIT provides valuable real-time information, it has technical limitations, including limited spatial resolution and dependency on electrode placement. Fourth, Fifth, we acknowledge that the low PEEP/FiO₂ strategy used in the control group is no longer the most up-to-date standard, but it was an intentional design choice in our study. Finally, our protocol incorporated recruitment maneuvers before PEEP titration, which may have influenced outcomes and complicated attribution of effects solely to EIT-guided titration. Despite these limitations, our findings provide meaningful evidence supporting EIT as a practical bedside tool for optimizing mechanical ventilation in ARDS.

Clinical implications and future directions

The results of this study suggest that EIT-guided recruitment maneuver and PEEP titration can improve oxygenation, reduce driving pressure, and enhance respiratory system compliance without increasing adverse events. These physiological benefits may translate into improved survival, as reflected by the trend toward lower 28-day mortality. Given the heterogeneity of ARDS, individualized approaches are urgently needed, and EIT offers a non-invasive, radiation-free, real-time method for tailoring ventilator settings. Larger randomized trials are warranted to determine whether these physiological improvements lead to clinically significant reductions in mortality and morbidity. Furthermore, integration of EIT with other bedside tools, such as transpulmonary pressure monitoring and advanced imaging, may further refine personalized ventilation strategies.

Conclusion

EIT-guided PEEP adjustment provided superior oxygenation and more favorable respiratory mechanics compared with low PEEP/FiO₂ strategy in moderate-to-severe ARDS. While the survival benefit was not statistically significant, the physiological improvements highlight the potential of EIT as a bedside tool to individualize lung-protective ventilation. Integration of EIT into clinical protocols may enhance decision-making in the management of ARDS.