Electrical impedance tomography guided positive end-expiratory pressure titration in critically ill and surgical adult patients: a systematic review and meta-analysis

Abstract

Objective

Electrical impedance tomography (EIT) has been used to titrate positive end-expiratory pressure (PEEP). This study aims to develop a comprehensive view of the efficacy and long-term prognosis of EIT-guided PEEP compared to other conventional approaches in various clinical scenarios, including patients with acute respiratory distress syndrome (ARDS), hypoxemic acute respiratory failure (hARF) and patients undergoing surgery under general anesthesia.

Methods

The literature search was conducted in PubMed, Web of Science, Embase, and Cochrane Library, from inception to July 30, 2023 (ARDS/hARF) and October 5, 2023 (surgery). The Cochrane risk of bias assessment and the methodological index for non-randomized studies were used for quality appraisal. The main outcomes were PEEP level, PaO₂/FiO₂ ratio, lung/respiratory system compliance (C_L/C_rs), driving pressure (ΔP), in-hospital mortality, and postoperative pulmonary complications (PPCs) in surgical studies.

Results

Four randomized controlled trials (RCTs), one historical control study, and six before-after studies of ARDS/hARF, as well as eight surgical RCTs, were retrieved. Subgroup analysis has been carried out and analysis of before-after studies was performed separately. Diverse PEEP strategies were adopted in the included studies, such as low/high PEEP-FiO₂-table of ARDS-net, pressure–volume loop, and transpulmonary pressure. In ARDS/hARF studies, the EIT strategy did not result in considerably enhanced respiratory system mechanics, including comparable PaO₂/FiO₂ ratios, comparable ΔP, and increased C_L/C_rs. As for long-term prognosis, the rough overall meta-analysis showed decreased in-hospital mortality (risk ratio RR = 1.54, 95% CI = (1.09, 2.18), P = 0.01). In patients undergoing general anesthesia surgery, the EIT group demonstrated increased PaO₂/FiO₂ ratio, C_L/C_rs, and decreased ΔP versus the fixed 4 or 5 cmH₂O PEEP. In postoperative prognosis, incidence of PPCs was generally comparable between the two groups.

Conclusion

The EIT-derived PEEP setting strategy might be associated with potential benefits in respiratory outcomes and prognosis in ARDS/hARF and surgical patients. Current data is insufficient to provide solid evidence.

Background

Positive end-expiratory pressure (PEEP) is commonly used in critical care to minimize alveolar collapse and improve oxygenation. In the Berlin definition of acute respiratory distress syndrome (ARDS), the optimized PEEP setting has been listed as one of the most crucial approaches in ARDS management [1]. In patients undergoing surgery under general anesthesia, protective mechanical ventilation (MV) with an adequate PEEP setting has also been proven to improve clinical outcomes [2]. However, determining the “ideal” PEEP remains difficult in real-life practice. Overdistension occurs when excessive PEEP is applied, while inadequate PEEP may lead to alveolar collapse, thus resulting in lung derecruitment. The selection of optimized PEEP has long been a major topic of clinical discussion.

The traditional ARDS network PEEP selection table based on the ratio of arterial oxygen partial pressure to fractional inspired oxygen (PaO₂/FiO₂) is a straightforward and convenient method [3]. However, due to significant individual variation in lung recruitability, it may not always provide effective results [4]. Thus, it is necessary to provide individualized PEEP for mechanically ventilated patients. In recent years, many personalized PEEP titration strategies based on various parameters have been proposed, including lung or respiratory system compliance (C_L or C_rs) [5], stress index [6], pressure–volume (PV) curve [7], or transpulmonary pressure (P_L) [8]. These approaches, unfortunately, are all rooted in global parameters that cannot completely represent regional lung status, especially in diseases with heterogeneous representation [9].

Electrical impedance tomography (EIT) is a non-invasive, non-radiological, and bedside lung monitoring tool that provides real-time and continuous information on ventilation distribution [10]. It can access ventilation homogeneity by identifying alveolar recruitment and regional hypoventilated lung units on a regional basis [11]. By visualizing regional responses of lung tissue, EIT would allow clinicians to modify therapeutic interventions and potentially minimize adverse events during MV. Over the past few years, EIT-based PEEP titration strategies have emerged as one of the most popular and valuable topics in MV management. EIT has been evaluated in many studies to guide PEEP titration in multiple clinical conditions such as ARDS, hypoxemic acute respiratory failure (hARF), and surgery under general anesthesia [9]. Despite its promising benefits and wide range of clinical applications, studies to date have not been able to achieve practical results consistently, especially in long-term prognosis [9, 12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29]. Therefore, it is desirable to establish an evidence-based overview to evaluate the clinical effect of the EIT-guided PEEP setting.

This systematic review and meta-analysis aims to develop a comprehensive view of the practical implications of EIT-based PEEP titration in ARDS/hARF patients and surgical patients in order to advance future clinical practice and investigations.

Methods

This systematic review was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [30], and the pre-defined study protocol has not been published nor registered on the international PROSPERO.

Literature search

Two authors (Y. Gao and Y. Chi) independently searched the PubMed, Web of Science, Embase, and Cochrane Library databases from inception to July 30, 2023 (ARDS/hARF) and October 5, 2023 (surgery). Literature searches for ARDS/hARF patients and patients undergoing surgery under general anesthesia were conducted separately. The search query for the ARDS/hARF section: ("respiratory distress syndrome"[MeSH Terms] OR "acute respiratory distress syndrome"[Title/Abstract] OR "ARDS"[Title/Abstract]) AND ("positive pressure respiration"[MeSH Terms] OR "positive end-expiratory pressure"[Title/Abstract] OR "PEEP"[Title/Abstract]) AND ("electrical impedance tomography"[Title/Abstract] OR "EIT"[Title/Abstract]). As for the surgical part, the authors used the following keywords and their variations: "Surgery", "Anesthesia", "PEEP", and "EIT". The language was restricted to English in both sections. The researchers also viewed the references of the selected studies for additional articles.

Selection criteria

The inclusion criteria were as follows:

• 1) Participant (P): ARDS/hARF patients with PaO₂/FiO₂ < 300 mmHg or surgical patients receiving general anesthesia.

• 2) Interventions (I): EIT-based PEEP titration strategy.

• 3) Comparisons (C): any other PEEP titration strategy.

• 4) Outcomes (O): Studies contain at least one of the following indicators: PEEP value, PaO2/FiO2 ratio, CL or Crs, driving pressure (ΔP), mortality, or postoperative pulmonary complications (PPCs) in surgical studies.

• 5) Study (S): clinical trials, including randomized controlled trials (RCTs) and other controlled studies.

The exclusive criteria were as follows: 1) pediatric studies (under 18) or animal experiments, 2) case reports, clinical protocols, or reviews, 3) studies with incomplete data 4) full text unavailable.

Data extraction

Two reviewers (Y. Gao and Y. Chi) separately used a pre-defined data extraction form to extract data from the enrolled studies. The third reviewer (H. He) was requested to resolve any conflict. The following data were extracted: first authors, study design, characteristics of patients, sample size, EIT strategy, the duration of EIT guided ventilation, PEEP titration strategy in the control group, outcomes, side effects, etc.

Quality assessment

The categories of the Cochrane Handbook for Systematic Reviews of Interventions were adopted to assess the RCTs: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other bias [31].

Non-randomized studies were assessed by the methodological index for non-randomized studies (MINORS) [32]. MINORS contains 12 items, each of which scores from 0 to 2, with higher total scores suggesting higher quality evidence. All items are applicable for comparative studies [33]. Any discrepancy during the quality assessment process would be evaluated and discussed by all authors.

Outcome measurement

The main outcomes were PEEP value, PaO₂/FiO₂ ratio, C_L or C_rs, ΔP, in-hospital mortality, and PPCs (including respiratory infection, respiratory failure, pleural effusion, atelectasis, pneumothorax, bronchospasm, aspiration pneumonia, pneumonia, ARDS, tracheobronchitis, etc.). Regarding other prognoses, length of mechanical ventilation, length of intensive care unit (ICU) or hospital stay, and weaning success rate were also included in the study.

Statistical analysis

Review Manager software 5.3, provided by the Cochrane Collaboration, was used for meta-analysis. The analysis was conducted based on the recommendations in the PRISMA guidelines [30]. The continuous variables were analyzed in mean difference (MD), estimated at a 95% confidence interval (CI). The risk ratios (RR) were used to process dichotomous variables. Statistical heterogeneity among studies determined the selection of the fixed-effect or random-effect model. The random-effect model would be adopted when the heterogeneity was significant (P ≤ 0.10, I² > 50%). Otherwise, the fixed-effect model would be applied. If the P-value was below 0.05, the difference between the two groups would be considered statistically significant. Funnel plot was used to identify the publication bias. The uniform distribution of study points indicates the absence of publication bias.

Results

Literature search and study characteristics

ARDS/hARF

A total of 673 references were retrieved from the databases in the initial search for ARDS/hARF patients. After removing duplicates, 424 studies were screened by titles and abstracts. Ultimately, eleven studies with 463 patients, including four RCTs [9, 12,13,14], one historical control study [15], and six before-after studies [16,17,18,19,20,21], were included in the present review of the ARDS/hARF section based on full-text assessment (Fig. 1).

Fig. 1

Flowchart of study search and inclusion criteria for ARDS/hARF studies

Surgical

For surgical patients, an initial review of 260 references was carried out, and eventually, eight RCTs with a total of 478 patients were included (Fig. 2) [22,23,24,25,26,27,28,29]. The basic characteristics of the selected studies are summarized respectively in Table 1 and Table 2.

Fig. 2

Flowchart of study search and inclusion criteria for surgical studies

Table 1 Basic characteristics of the included studies –- ARDS/hARF Group

Table 2 Basic characteristics of the included studies –- surgical patients under general anesthesia

Quality assessment

The bias risk of the included RCTs is illustrated in Fig. 3. Among the included twelve RCTs, eight were considered as low risk in random sequence generation and six in allocation strategy as they documented the specific methods used. In other studies where there was no detail about the procedures, the risk was judged as unclear. All RCTs were classified as low risk in terms of blinding of participants and personnel, blinding of outcome assessment, and incomplete outcome data. Nine of those were classified as low risk in the selective reporting as the protocol had been registered in advance. Other biases in all RCTs were assessed as having unclear risks. Non-randomized studies, including seven comparative designs, were evaluated according to the MINORS assessment with an average score of 19 ± 0.8. The MINORS scores for each study are presented in Table 1.

Fig. 3

Risk of bias of included RCTs (+ , no bias; -, bias; ?, bias unknown). 1- ARDS/hARF studies; 2- surgical studies

Main outcomes

PEEP value

ARDS/hARF

In the included ARDS/hARF studies, PEEP values were compared between the EIT group and control group with variable PEEP titration strategies such as high and/or low PEEP-FiO₂ table, PV loop, P_L/FiO₂ table. Due to the inconsistency of methods and the lack of data, the meta-analysis of RCTs could not be carried out. Characteristics of optimal PEEP values in ARDS/hARF RCTs [9, 12,13,14] and historical control study [15] are summarized in Table 3a.

Table 3 PEEP value in ARDS/hARF studies

Meta-analysis of the comparison between the high/low PEEP-FiO₂ table and EIT was performed in before-after studies. Characteristics of optimal PEEP values have been listed in Supplementary Table 1. Data from Heines et al. [20] and Eronia et al. [21] was excluded due to mixed or unclear adoption of high and low PEEP-FiO₂ table. In the random-effects analysis (I² = 76%, P = 0.02) of three before-after studies [16, 17, 19] where the high PEEP-FiO₂ table was applied, there was no statistical difference between groups (MD = −2.82, 95% CI: −6.00, 0.37, P = 0.08; Supplemental Fig. 1a). A consistent result was observed in Jimenez et al. [12]. Three before-after studies [17,18,19] used the low PEEP-FiO₂ table in the control group, where PEEP is significantly lower than that in the EIT group (MD = 4.30, 95% CI: 1.71, 6.89, P = 0.001; Supplemental Fig. 1b). The random-effect model was adopted because of the significant heterogeneity (I² = 80%, P = 0.006). Whereas in He et al. [9], PEEP was similar between the EIT group and the low PEEP-FiO₂ table group. When comparing PEEP between the EIT group and PV loop, Hsu et al. [13] and Zhao et al. [15] obtained opposite results. In the study of Scaramuzzo et al. [14], PEEP was comparable between the EIT group and the P_L/FiO₂ table.

Surgical

In trials on surgical patients under general anesthesia [22,23,24,25,26,27,28,29], the PEEP values in control groups were all fixed. Hence, they could not be compared in the meta-analysis. However, in each study, the PEEP value in the EIT group was significantly higher than the fixed PEEP value in the control group (4 or 5 cm H₂O). The weighted mean of the EIT group and control group was 12.00 cm H₂O and 4.91 cm H₂O, respectively. Characteristics of optimal PEEP values in surgical studies are shown in Table 4.

Table 4 PEEP value in surgical studies

PaO₂/FiO₂ ratio

ARDS/hARF

Nine ARDS/hARF clinical trials [9, 12,13,14,15,16, 18, 20, 21] including 441 patients reported on PaO₂/FiO₂ ratio. In RCTs [9, 12,13,14] and Zhao et al. [15], comparable PaO₂/FiO₂ was found between the EIT group and the control group in each of the studies. We performed subgroup analysis based on different PEEP strategies in the EIT group and the control group (Forest plot: Fig. 4, Funnel plot: Supplemental Fig. 2a, Summary of results: Supplemental Table 1a). The overall meta-analysis demonstrated that there was no statistical difference between the two groups (pooled MD = 0.06, 95% CI: −18.34, 18.45, P = 1.00; I²: 0%). Similar result was observed in before-after studies [16, 18, 20, 21] (pooled MD = 27.74, 95% CI: −3.40, 58.87, P = 0.08; I²: 76%; Supplemental Fig. 2b).

Fig. 4

Forest plot of PaO₂/FiO₂ in EIT and control groups for ARDS/hARF studies

Surgical

In surgical studies, with six studies [22, 24, 25, 27,28,29] and 382 participants included, the result showed that EIT group was significantly superior to the control group either in each study or in the overall meta-analysis (pooled MD = 103.62, 95% CI: 57.52, 149.72, P < 0.0001; I²: 84%; Forest plot: Fig. 5, Funnel plot: Supplemental Fig. 2c, Summary of results: Supplemental Table 1b).

Fig. 5

Forest plot of PaO₂/FiO₂ in EIT and control groups for surgical studies

Lung and respiratory system compliance

ARDS/hARF

As regards ARDS/hARF patients, nine studies [9, 12, 13, 15,16,17,18, 20, 21] with 438 patients described C_L or C_rs after PEEP titration. The OL/OD strategy was adapted for EIT-guided PEEP titration by all RCTs [9, 12,13,14] and Zhao et al. [15]. Subgroup analysis by PEEP strategy in the control group is shown in Fig. 6. In the studies of Jimenez et al. [12], He et al. [9], and Hsu et al. [13], comparable C_L/C_rs were found between groups. While in Zhao et al. [15], the EIT group presented with better C_L/C_rs. Although in each subgroup, the result between groups was insignificant, the overall meta-analysis showed higher C_L/C_rs in the EIT group (pooled MD = 3.01, 95% CI: 0.13, 5.89, P = 0.04; I²: 39%; Funnel plot: Supplemental Fig. 3a, Summary of results: Supplemental Table 2a). Consistent finding was observed in before-after studies [16,17,18, 20, 21] (pooled MD = 3.84, 95% CI: 0.56, 7.11, P = 0.02; Supplemental Fig. 3b).

Fig. 6

Forest plot of C_L/C_rs in EIT and control groups for ARDS/hARF studies

Surgical

Among seven surgical studies [22,23,24, 26,27,28,29] and 438 patients, the heterogeneity was significant (I² = 95%, P < 0.00001). The comparison between the EIT group and the control group on lung compliance is summarized in Fig. 7. Subgroup analysis according to PEEP strategy in the EIT group is presented. Studies using the CL/OD [22, 24, 27, 28] or RVDI [29] strategy demonstrated significantly better C_L/C_rs in the EIT group, whereas others with Creg [26] and EELI [23] strategy found similar results between groups. In the overall analysis, the EIT group was associated with significantly higher C_L/C_rs (pooled MD = 10.74, 95% CI: 3.43, 18.04, P = 0.004; Funnel plot: Supplemental Fig. 3c, Summary of results: Supplemental Table 2b).

Fig. 7

Forest plot of C_L/C_rs in EIT and control groups for surgical studies

Driving pressure

ARDS/hARF

ΔP was documented in nine ARDS/hARF studies [9, 12,13,14,15,16,17,18, 22] (419 patients). Subgroup analysis was carried out on the PEEP strategy in the EIT and control groups (Fig. 8). Studies with PV loop in the control group [13, 15] or CL/OD in the EIT group [9, 12, 13, 15] resulted in significantly lower ΔP in the EIT group compared to the control group. Yet there was no statistical difference between groups in the overall analysis of both RCTs [9, 12,13,14] and Zhao et al. [15] (pooled MD = −1.12, 95% CI: −2.57, 0.34, P = 0.13; I²: 79%; Funnel plot: Supplemental Fig. 4a, Summary of results: Supplemental Table 3a) or before-after studies [16,17,18, 21] (pooled MD = −0.49, 95% CI: −1.15, 0.16, P = 0.14; I²: 0%; Supplemental Fig. 4b).

Fig. 8

Forest plot of ΔP in EIT and control groups for ARDS/hARF studies

Surgical

Regarding the surgical studies, the heterogeneity test indicated a significant heterogeneity between the seven included trials [22, 23, 25,26,27,28,29] (I² = 91%, P < 0.00001), which contained 398 patients. Subgroup analysis was performed based on the PEEP strategy in the EIT group (Fig. 9). Significantly lower ΔP was found in the EIT group in each subgroup except the Creg strategy. The overall random-effects meta-analysis revealed that ΔP in the EIT group was significantly lower compared to the control group (pooled MD = −3.01, 95% CI: −4.63, −1.39, P = 0.0003; I²: 91%; Funnel plot: Supplemental Fig. 4c, Summary of results: Supplemental Table 3b).

Fig. 9

Forest plot of ΔP in EIT and control groups for surgical studies

In-hospital mortality

ARDS/hARF

Three ARDS trials [9, 13, 15] involving 259 patients recorded in-hospital mortality or survival, which was converted to mortality by the reviewers. He et al. [9] collected 28-day mortality, and the follow-up duration of Hsu et al. [13] and Zhao et al. [15] was until hospital discharge only. No heterogeneity was found between the studies (I² = 0, P = 0.70). Subgroup analysis was conducted according to the PEEP strategy in the control group for the CL/OD strategy adopted by all three studies in the EIT group (Fig. 10). The results in the PV loop subgroup and total meta-analysis demonstrated that mortality in the EIT group was significantly reduced compared with the control group (pooled RR = 1.54, 95% CI: 1.09, 2.18, P = 0.01; I²: 0%; Funnel plot: Supplemental Fig. 5, Summary of results: Supplemental Table 4).

Fig. 10

Forest plot of in-hospital mortality in EIT and control groups for ARDS/hARF studies

Incidence of postoperative pulmonary complications

Surgical

Five surgical studies [22,23,24, 27, 29] with 350 patients investigated the incidence of PPCs. Ma et al. [22] and Xiao et al. [23] investigated PPCs within 7 days after surgery. The follow-up period for PPCs in Zha et al. [24] and Liu et al. [27] was until hospital discharge. Nestler et al. [29] recorded PPCs within 28 postoperative days. Heterogeneity between studies was not significant (I² = 11%, P = 0.34). In each study, the incidence of PPCs was comparable between the EIT group and the control group. Similar finding was obtained by the overall analysis (pooled RR = 1.46, 95% CI: 0.99, 2.14, P = 0.05; Forest plot: Fig. 11, Funnel plot: Supplemental Fig. 6, Summary of results: Supplemental Table 5).

Fig. 11

Forest plot of incidence of PPCs in EIT and control groups for surgical studies

Long-term prognosis

Only a limited number of included trials reported on long-term prognosis.

ARDS/hARF

The length of mechanical ventilation and length of ICU stay were presented in two ARDS studies [9, 13]. In both studies, patients in the EIT group showed longer MV and ICU stay periods, yet there were no statistical differences between groups. In addition, two studies [13, 15] have found higher but not significant weaning success rates in the EIT group.

Surgical

Length of hospital stay, reported in four surgical trials [22, 24, 27, 28], was also not statistically different between the groups (MD = −0.03, 95% CI: −0.44, 0.337, P = 0.88; Supplemental Fig. 7). Pereira et al. [29] assessed post-operative atelectasis by using whole-lung computed tomography after extubation and anesthesia recovery. The outcome revealed that the EIT group reduced atelectasis with a significantly lower proportion of collapsed lung tissue.

Adverse events

ARDS/hARF

Four ARDS studies [9, 13, 15, 20] have reported adverse events. Of those, the incidence of new barotrauma in the EIT group was all zero, with one study [15] reporting a 6.5% incidence of barotrauma in the control group (no statistical difference between groups).

Surgical

None of the included studies on surgical patients have observed intervention-induced adverse effects.

Obese patients

Surgical

One surgical study [29] (Nestler et al.) has focused on obese patients (BMI > 35 kg/m²). The findings of PEEP value, PaO₂/FiO₂ ratio, C_L/C_rs, and ΔP were generally similar between Nestler et al. and non-obese surgical studies. However, Nestler et al. appears to provide more powerful results in each outcome indicator in the EIT group.

Discussion

This systematic review comprehensively describes the efficacy of different EIT PEEP titration methods versus various conventional PEEP setting methods in several clinical scenarios including ARDS, hARF, surgical, and obese patients. In ARDS/hARF patients, the EIT method did not lead to notably improved respiratory system mechanics, except for increased PaO₂/FiO₂ ratio. Yet regarding long-term prognosis, the meta-analysis of three trials roughly resulted in decreased in-hospital mortality. In patients undergoing general anesthetic surgery, the EIT group exhibited superior PaO₂/FiO₂ ratio, C_L/C_rs, and ΔP over the fixed 4 or 5 cmH₂O PEEP in the control group in most studies. Concerning postoperative prognosis, the overall meta-analysis of five trials showed comparable results between the EIT group and the control group.

Meta-analysis including subgroup analysis based on different PEEP strategies in the EIT group and control group has been introduced in the present review as statistical evidence. However, unfortunately, due to the considerable heterogeneity in PEEP setting methods in either the EIT or control groups of the included trials, the strength of the current findings has been largely undermined. We suggest that the results of meta-analyses should be interpreted cautiously and only for rough reference.

ARDS and hARF are common critical respiratory diseases that are characterized by severe hypoxemia and massive non-cardiogenic pulmonary edema [34]. Despite considerable progress, ARDS remains life-threatening and a major concern in clinical practice, especially after the COVID-19 pandemic. Although it has been suggested in the Berlin definition that low tidal volume, adequate PEEP, and prone positioning should be adopted in the management of severe ARDS, the optimal way of PEEP titration is still ambiguous [1]. An inappropriately high PEEP can lead to alveolar overdistention and consequently serious lung injury. In this meta-analysis, we compared different EIT-based PEEP titration strategies with other methods including PEEP-FiO₂ table, PV loop, and P_L/FiO₂ table. Rough overall meta-analysis indicated generally similar respiratory function and improved prognosis, including higher C_L/C_rs, comparable PaO₂/FiO₂ ratio, comparable ΔP, and lower in-hospital mortality.

Regarding mortality, although the EIT group showed a significant improvement in the meta-analysis, the included studies are limited to providing solid evidence due to their considerable heterogeneity and insufficient number. Among three trials that reported mortality, Hsu et al. was the only study found significantly reduced mortality in the EIT group. Notably, they also performed the longest duration of EIT-guided ventilation for 48 h. Theoretically, longer periods of EIT-guided ventilation are more likely to provide positive results. Yet the current evidence is inadequate to draw a firm conclusion.

Despite the potential benefit of EIT, studies to date seldom found significant improvement in long-term outcomes. It is noticeable that in most of the current studies the duration of EIT-guided ventilation was short, with only a few hours. It is doubtful that a short period of EIT ventilation management would lead to significant improvement in prognosis. Future multicenter trials are necessary to further evaluate this issue. We encourage future studies to perform longer period of EIT-guided ventilation for at least 48 h. Moreover, the severity of ARDS/hARF could also affect the results. Patients with moderate to severe hypoxemia are more likely to achieve greater improvement in prognosis.

Lung-protective MV with low tidal volume and low ΔP is widely applied in patients undergoing surgery under general anesthesia [2, 35]. Inappropriate management of MV could lead to a high risk of postoperative complications, even ARDS [36]. The main goal for this scenario focuses on enhancing prognosis and minimizing postoperative complications. In most included surgical studies in the present review, the EIT-based PEEP strategy resulted in enhanced pulmonary outcomes such as PaO₂/FiO₂ ratio, C_L/C_rs, and ΔP. Regarding the incidence of PPCs and other long-term prognosis, benefits in the EIT group was insignificant. A previous meta-analysis has confirmed that changes in PEEP levels that result in increased ΔP are correlated with more PPCs [35]. We may infer from this finding that the reduced ΔP in the EIT group could potentially improve post-operative prognosis.

Surgical patients tend to have better global lung conditions and respiratory system mechanics compared to ARDS/hARF patients. It is therefore more difficult to find significant benefits, especially in studies with small samples. In addition, different types of surgery could also influence the outcome. Among the included surgical studies, most of them focused on patients undergoing laparoscopic surgery. Compared to other surgeries such as laparotomy and non-abdominal surgery, laparoscopic surgery imposes additional intra-abdominal pressure on patients, which means that these patients may present with greater individual heterogeneity and demand for higher levels of PEEP. This makes patients undergoing laparoscopic surgery more likely to benefit from EIT-guided ventilation.

In recent years, thoracic EIT has gained increasing attention in PEEP titration. For the selection of “optimal” PEEP, many EIT-based methods have been proposed, such as CL/OD strategies [9, 12, 13, 15,16,17, 19, 20, 22, 24, 27, 28] (the intercept point of the cumulative percentage lung collapse [CL] and overdistension [OD] curves or a percentage of collapsed lung tissue less than 15% with the least overdistension), Creg strategy [18, 26] (the highest regional compliance [Creg]), GI strategy [37] (the lowest global inhomogeneity [GI] index), RVDI strategy [25, 29] (the minimum regional ventilation delay index [RVDI]), EELI strategy [21, 23] (trend of end-expiratory lung impedance [EELI]), and SStot strategy [14] (the lowest pressure with a total percentage of total silent spaces [SStot] less than or equal to 15%). In this review, due to the lack of existing clinical trials, meta-analysis or subgroup analysis for comparisons among EIT strategies could not be carried out.

Despite its potential advantages, EIT remains an advanced technique that has not been widely adopted in general practice. It requires experienced clinicians and specialized and expensive equipment and is therefore challenging for system-wide implementation, especially in under-resourced regions and developing countries. Further investigation and development of EIT technology is still needed to overcome this gap. As the technical progress of EIT continues, the cost of machine production might be significantly reduced in the future. Meanwhile, guidance and expert consensus in EIT-based operational standards, diagnosis, and ventilation strategy should be established to promote the popularization of the EIT technique. Moreover, the EIT database, machine learning, and EIT-based predictive modeling are recommended to be conducted in future research.

This study is the first meta-analysis that focuses on long-term prognosis, including mortality in ARDS/hARF patients and PPCs in surgical patients. To date, studies focusing on EIT-based PEEP strategy are heterogeneous in methodology and lack participants, therefore not enabling high-quality evidence. With the development of multi-center studies with large sample sizes, this issue might be addressed. Moreover, the efficacy of the EIT-based method compared to conventional methods has not been clarified, especially in terms of long-term prognosis such as pulmonary function, mortality, length of hospital stay, etc. For further research, longer follow-up duration and long-term functional assessment are encouraged.

The limitations are as follows: Firstly, the evidence in this meta-evidence is not sufficient with a lack of adequate RCTs. In addition, part of the included studies is before-and-after or historical controlled studies. Secondly, EIT strategies and PEEP setting methods in control groups varied, including the conventional ARDS-net and individual settings based on global parameters. Thirdly, the duration of EIT-guided ventilation in the included ARDS/hARF trials are notably differed, which also increases the heterogeneity between studies. Fourthly, this study did not distinguish between mild and moderate to severe ARDS. Differences in the severity of hypoxemia may lead to certain deviations in clinical response. Fifthly, the ARDS patients included in this analysis might have exhibited various extrapulmonary and pulmonary origins, including COVID-19. Evidence from an earlier investigation indicated that COVID-19-related ARDS exhibited higher EIT-guided PEEP and different lung recruitability [38]. Sixthly, evidence for long-term prognosis is inadequate, especially in surgical patients. Seventhly, there is a widespread recognition that obese patients benefit from higher PEEP. In the obese study included in this review, the control groups adopted fixed lower PEEP, which would have resulted in an increased benefit in oxygenation and respiratory mechanics in the EIT group compared to other studies, and therefore may have influenced the overall outcomes to a certain extent.

Conclusion

This systematic review and meta-analysis of current existing studies indicates that the EIT-derived PEEP setting strategy may be associated with potential benefits over conventional PEEP strategies in ARDS/hARF patients and patients receiving surgery under general anesthesia. Further studies investigating the long-term prognosis involving lung complications and mortality are recommended.