Characteristics and Progression of Rehabilitation in Patients Undergoing Extracorporeal Membrane Oxygenation: A Retrospective Study

ABSTRACT

Objectives

Early rehabilitation during extracorporeal membrane oxygenation (ECMO) support is increasingly implemented, yet real‐world data describing rehabilitation progression remain limited. This study aimed to characterize rehabilitation patterns during ECMO and descriptively compare physical activity trajectories among patients managed as bridge to transplantation (BTT), bridge to recovery (BTR), and ECMO waitlist death (EWD).

Methods

Adult patients (≥18 years) admitted to a medical ICU between May 2018 and November 2023 who received ECMO for ≥24 h and participated in at least one active rehabilitation session were retrospectively analyzed. Rehabilitation candidacy, weekly rehabilitation frequency, session interruptions, and highest physical activity level were recorded. Between‐group comparisons were exploratory.

Results

A total of 111 patients were included. In ECMO week 1, 58% of rehabilitation candidates received active rehabilitation, increasing to approximately 80% during weeks 2–4. Most patients completed or discontinued rehabilitation within 4 weeks because of transplantation, clinical recovery, or deterioration. Weekly rehabilitation frequency differed among groups and was highest in the BTT group (mean, 4.2 sessions/week; p < 0.05). Other activity‐level comparisons among groups were not statistically significant. Sitting‐level activities were consistently achievable during ECMO support, including support beyond 4 weeks. Among ICU survivors, activity levels were maintained through decannulation and improved afterward.

Conclusions

Among selected patients who entered an active ECMO rehabilitation pathway, rehabilitation was commonly initiated within 2 weeks and completed or discontinued within 4 weeks. Edge‐of‐bed sitting was generally achievable throughout ECMO support.

1. Introduction

In critically ill patients, early mobilization helps prevent post‐intensive care syndrome, including muscle atrophy and intensive care unit (ICU)‐acquired weakness, supports functional recovery, and reduces both the duration of mechanical ventilation and ICU length of stay [1, 2]. Recent studies further demonstrate that, when guided by the patient's clinical condition, early initiation of structured rehabilitation is not only safe but also beneficial across multiple clinical outcomes [3]. As a result, early mobilization has become a central component of ICU rehabilitation.

This approach has increasingly been extended to patients receiving extracorporeal membrane oxygenation (ECMO). Historically, deep sedation and restriction of patient movement were common due to concerns about decannulation, hemodynamic instability, and bleeding. However, a growing body of evidence indicates that, for appropriately screened patients, awake ECMO and active rehabilitation can be performed safely and are associated with favorable outcomes [4, 5, 6]. Active rehabilitation during ECMO support has been shown to reduce ICU length of stay and, particularly for patients awaiting lung transplantation, to preserve pre‐transplant conditioning and accelerate post‐transplant recovery [7, 8]. Maintaining physical activity during ECMO may also prevent muscle atrophy, preserve functional status, and increase the likelihood of successful weaning or transplantation [7, 9, 10].

Despite these encouraging findings, granular data describing how rehabilitation is implemented across the entire ECMO course in real‐world settings remain limited. Most prior studies have focused on the safety, feasibility, or prognostic value, and few have examined, within a sufficiently large cohort, the week‐by‐week progression of rehabilitation stages or the detailed components of physical therapy sessions during ECMO [5, 11, 12, 13].

Accordingly, this study aimed to describe how rehabilitation is applied throughout ECMO support in routine clinical practice and to compare rehabilitation characteristics and physical activity trajectories among three groups: bridge to transplantation (BTT), bridge to recovery (BTR), and ECMO waitlist death (EWD). To accomplish this, we systematically tracked weekly rehabilitation candidacy, participation, session frequency, and the highest physical activity level achieved to delineate real‐world ECMO rehabilitation patterns and identify differences across patient groups.

2. Materials and Methods

2.1. Study Design and Clinical Characteristics

This single‐center retrospective descriptive study included adult patients (≥18 years) admitted to the medical intensive care unit (MICU) of a tertiary center in Seoul, Republic of Korea, between May 2018 and November 2023. Eligible patients received ECMO support for ≥24 h and participated in at least one active rehabilitation session during ECMO support. Patients who underwent no active rehabilitation during ECMO were excluded, as were those who initiated ECMO after lung transplantation, required ECMO continuation after transplantation, or began rehabilitation only after transplantation (Figure 1). Therefore, this cohort represents selected patients who entered an active ECMO rehabilitation pathway and was not intended to estimate the overall early mobilization rate among all ICU patients receiving ECMO from ICU admission. Because patients were classified according to clinical indication, disease trajectory, and outcome rather than randomization, this study was designed to characterize real‐world rehabilitation delivery and progression. Between‐group comparisons were considered exploratory and descriptive rather than causal. Collected variables included age, sex, reason for ICU admission, ICU length of stay, APACHE II and SOFA scores at MICU admission, ECMO duration, and ICU mortality. The study was approved by the Institutional Review Board of Asan Medical Center (IRB No. 2025‐1216).

FIGURE 1.

Flow diagram of study patients. ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit.

2.2. Rehabilitation during ECMO Support

Rehabilitation for patients receiving ECMO was implemented according to the MICU's structured rehabilitation program. Potential candidates were screened daily, and eligibility for rehabilitation was reassessed throughout the ECMO course. If rehabilitation was temporarily paused due to clinical instability or if the activity level was downgraded to passive range of motion (PROM) only, patients remained within the rehabilitation cohort and continued to be tracked. When clinical status improved sufficiently to permit active range of motion or active respiratory rehabilitation, the rehabilitation level was upgraded and active therapy was resumed.

If rehabilitation was discontinued because of clinical improvement, the patient was categorized as “graduated”. If rehabilitation was no longer feasible due to clinical deterioration or death, the patient was classified as “terminated.” Only rehabilitation data collected up to the point of graduation or termination were included in the analysis.

2.3. Active Rehabilitation and Safety Criteria

Patients receiving ECMO for ≥24 h were screened daily according to the institutional MICU Early Rehabilitation Protocol. Active rehabilitation was considered when the patient had a Richmond Agitation–Sedation Scale (RASS) score between −2 and +2 and limb muscle strength ≥ grade 2, allowing cooperative respiratory or active limb exercise. The safety checklist was adapted from published ICU mobilization safety recommendations, the institutional MICU protocol, and ECMO‐specific bedside assessment by the attending physician, dedicated ICU nurse, and physical therapist [3, 14, 15].

Active rehabilitation was performed only when all minimum safety criteria were met: SpO₂ ≥80%, arterial pH ≥7.25, respiratory rate <45 breaths/min, mean arterial pressure ≥60 and <150 mmHg, heart rate <130 beats/min, absence of hemodynamically unstable arrhythmia, no vasopressor escalation within 2 h, and norepinephrine or epinephrine dose <0.2 µg/kg/min. Fixed FiO₂ or PEEP thresholds were not used; instead, respiratory trend, ECMO flow stability, cannula security, bleeding risk, neurological status, sedation level, and overall clinical stability were reviewed before and during each session.

Active rehabilitation was withheld or downgraded to passive range‐of‐motion exercise if any exclusion criterion was present, including new arrhythmia or thrombosis, new hemorrhagic or ischemic stroke, increased intracranial pressure, uncontrolled seizure, uncontrolled bleeding, fever ≥38°C or severe chills, platelet count <20,000/µL, prone positioning, early postoperative status after major surgery, major fracture, neuromuscular blockade, patient refusal, or physician concern. Postoperative day 3 was used as a general reference point after major surgery according to institutional ICU practice; however, this criterion did not indicate absolute immobility, and limited positional changes or basic nursing care were allowed. Patients were reassessed at the next scheduled rehabilitation time point, and active rehabilitation was resumed once all eligibility criteria were satisfied.

Data on the reasons for delayed or non‐initiation of rehabilitation were not systematically available for the entire study period, and available 2023 institutional audit data are summarized in Table S1.

An interrupted rehabilitation session was defined as any session that was stopped because of a potential or actual safety concern. (Table S2).

2.4. Details of Rehabilitation Sessions

The rehabilitation team comprised the attending physician, a dedicated ICU nurse, and a physical therapist. During the study period, the medical ICU environment enabled rehabilitation for approximately 4–5 patients per day, with 30 min allocated per patient.

A daily rehabilitation stage goal was set individually according to each patient's condition, and rehabilitation was delivered in a stepwise manner, progressing from low‐ to high‐intensity activities.

2.5. Evaluation and Stratification of Activity Levels

Rehabilitation was individualized according to each patient's physical activity level, which was classified into six stages. These stages were adapted from the Medical ICU Early Rehabilitation Protocol (MICU‐ERP), developed collaboratively by the pulmonology and rehabilitation medicine departments at our institution. (Supplementary Materials) Activity levels were defined as follows: stage 1, PROM; stage 2, active range of motion in bed; stage 3, sitting activities such as bedside dangling or sitting at the edge of the bed; stage 4, standing activities; and stages 5 and 6, bedside or in‐room ambulation and ward ambulation, respectively [15]. This classification closely aligns with the ICU Mobility Scale, with higher stages indicating greater functional mobility [16]. The rehabilitation stage recorded for each patient was regarded as a representation of their actual mobility level during the session. For secondary descriptive analyses, activity level was dichotomized at stage 4 because this stage represents the first standing‐level, weight‐bearing activity, corresponding conceptually to a standing‐level mobility milestone in the ICU Mobility Scale [16, 17]. This threshold was used to distinguish patients who achieved verticalized mobility from those whose activity remained limited to bed‐level or sitting‐level rehabilitation.

2.6. Patient Classification and Rehabilitation Outcomes

Patients undergoing ECMO rehabilitation were first classified by ECMO indication as BTT or BTR. Among patients initially managed as BTT, those who died before transplantation were described separately as the ECMO waitlist death (EWD) group to characterize the rehabilitation trajectory of patients who deteriorated before transplantation. Therefore, the EWD group was used as an outcome‐defined descriptive subgroup rather than a baseline treatment category.

Rehabilitation‐related variables included time from ECMO initiation to the first rehabilitation session, total and weekly session numbers, and the occurrence of interrupted sessions. Rehabilitation intensity was represented by the highest activity level achieved during each session. Discontinuation of rehabilitation was classified as graduation due to clinical improvement or transplantation, or termination due to deterioration or death.

2.7. Statistical Analysis

Continuous variables were summarized as means ± standard deviations or medians with interquartile ranges, and categorical variables as frequencies and percentages. Clinical characteristics and ECMO rehabilitation features were compared across predefined groups using one‐way ANOVA or the Kruskal–Wallis test for continuous variables, and chi‐square or Fisher's exact tests for categorical variables, as appropriate.

To evaluate temporal changes in rehabilitation exposure during ECMO support, ECMO week was used as the analytic time unit. For each ECMO week, the number of rehabilitation candidates, the number of patients who underwent rehabilitation, and the number of sessions performed were descriptively recorded. Highest achieved physical activity levels were compared at ECMO initiation, immediately before decannulation, and ICU discharge using repeated‐measures ANOVA.

All between‐group analyses were exploratory. Given the retrospective observational design and absence of random group assignment, statistical comparisons were interpreted descriptively to characterize rehabilitation delivery and progression across clinical trajectories, rather than to infer causal effects.

3. Results

Table 1 summarizes the clinical characteristics of ECMO rehabilitation patients across the three groups. Patients in the BTR group were significantly older than those in the BTT and EWD groups. In both the BTT and EWD groups, interstitial lung disease (ILD) was the predominant underlying condition, accounting for 79.2% and 100% of cases, respectively. In contrast, the BTR group exhibited more heterogeneous etiologies, with pneumonia being the most common (40%). The duration of ECMO support was significantly longer in the EWD group compared with the other groups. There were no significant differences among groups in illness severity scores, including APACHE II and SOFA. Among the ECMO‐supported patients, VV ECMO was the predominant mode, and most patients were lung transplant recipients or candidates; however, VA, VAV, VVA, central ECMO, and non‐transplant patients were also included. (Table S3) Most ECMO rehabilitation sessions were performed while patients were receiving concomitant invasive mechanical ventilation.

TABLE 1.

Clinical characteristics of patients undergoing ECMO rehabilitation.

Variables	Total (n = 111)	BTT group (n = 77)	ECMO‐Waitlist Death (EWD) group (n = 19)	BTR group (n = 15)	p	Bonferroni
Age (years)	58.7 ± 11.3	57.0 ± 11.2	59.2 ± 8.5	66.5 ± 11.9	<0.05 b	a,b<c
Sex, n (%)
Male	63 (56.8)	43 (55.8)	12 (63.2)	8 (53.3)	0.813
Female	48 (43.2)	34 (44.2)	7 (36.8)	7 (46.7)
Etiology, n (%)					<0.05 [Link] , [Link]
ARDS	6 (5.4)	1 (1.3)	0 (0.0)	5 (33.3)
BO	4 (3.6)	3 (3.9)	0 (0.0)	1 (6.7)
COVID‐19	8 (7.2)	8 (10.4)	0 (0.0)	0 (0.0)
ILD	81 (73.0)	61 (79.2)	19 (100.0)	1 (6.7)
Pneumonia	7 (6.3)	1 (1.3)	0 (0.0)	6 (40.0)
Others	5 (4.5)	3 (3.9)	0 (0.0)	2 (13.3)
Length of ICU stay (days)	49.7 ± 30.4	51.9 ± 30.2	48.4 ± 34.2	39.9 ± 25.5	0.2	—
Initial APACHE II score	26 (11‐40)	28 (14‐40)	23 (13‐38)	26 (11‐40)	0.175	—
Initial SOFA score	9 (1‐17)	10 (1‐16)	9 (2‐13)	10 (4‐17)	0.397	—
ECMO days	25.8 ± 23.7	21.3 ± 17.7	42.0 ± 36.8	28.4 ± 22.1	<0.05 b	a < b
ICU mortality, n (%)	35 (31.5)	7 (9.1)	19 (100.0)	9 (60.0)	<0.05 b

^aFishers's exact test. Chi‐square test or Fisher's exact test was used for categorical variables, and one‐way ANOVA or Kruskal‐Wallis test was used for continuous variables. Data are presented as mean ± standard deviation, median (IQR) or number (percentage).

^b p‐value <0.05, Fisher's exact test, one‐way ANOVA or Kruskal‐Wallis test.

Abbreviations: APACHE, Acute physiology and chronic health evaluation; ARDS, Acute respiratory distress syndrome; BO, Bronchiolitis obliterans; ECMO, Extracorporeal membrane oxygenation; ICU, Intensive care unit; ILD, Interstitial lung disease; IQR, Interquartile range; SOFA, Sequential organ failure assessment.

Table 2 compares ECMO rehabilitation characteristics across groups. The mean interval from ECMO initiation to the first rehabilitation session was 8.7 days and did not differ significantly among groups. Weekly rehabilitation frequency differed significantly, with the BTT group receiving a mean of 4.2 sessions per week. In contrast, the number of active rehabilitation sessions per patient, the proportion of patients with interrupted sessions, and activity‐level comparisons at ECMO initiation, pre‐decannulation, and ICU discharge did not differ significantly among groups. When activity level was dichotomized at stage 4 (Standing), the proportion of patients achieving standing‐level or higher activity also did not differ significantly.

TABLE 2.

Comparison of ECMO rehabilitation between the groups.

Variables		Total	BTT group (n = 77)	ECMO‐Waitlist Death (EWD) group (n = 19)	BTR group (n = 15)	p
	n	mean ± SD or n (%)
Time to first rehabilitation after ECMO initiation (Day)	111	8.7 ± 13.3	8.2 ± 13.5	7.9 ± 11.7	12.2 ± 14.7	0.060
Weekly frequency of ECMO rehabilitation	111	3.5 ± 2.3	4.2 ± 2.2	1.8 ± 1.3	2.0 ± 1.6	<0.05 b
Number of active rehabilitation sessions per person	111	8.8 ± 6.8	8.9 ± 6.9	9.2 ± 6.4	7.3 ± 7.2	0.400
Number of patients with interrupted sessions	111	16 (14.4)	10 (13.0)	5 (26.3)	1 (6.7)	0.981 a
Highest mean activity level per session
at ECMO Initiation	111	2.7 ± 1.0	2.7 ± 1.0	2.4 ± 1.0	2.9 ± 1.0	0.359
Stage < 4 (PROM‐Sitting)		76 (68.5)	53 (68.8)	14 (73.7)	9 (60.0)	0.690
Stage ≥ 4 (Standing‐Ambulation)		35 (31.5)	24 (31.2)	5 (26.3)	6 (40.0)
at pre‐decannulation	87	2.8 ± 1.2	2.8 ± 1.2		3.1 ± 1.0	0.346
Stage < 4 (PROM‐Sitting)		54 (62.1)	49 (63.6)		5 (50.0)	0.494 a
Stage ≥ 4 (Standing‐Ambulation)		33 (37.9)	28 (36.4)		5 (50.0)
at ICU discharge	78	3.4 ± 1.2	3.3 ± 1.2		4.3 ± 1.4	0.056
Stage < 4 (PROM‐Sitting)		37 (47.4)	36 (50.0)		1 (16.7)	0.204 a
Stage ≥ 4 (Standing‐Ambulation)		41 (52.6)	36 (50.0)		5 (83.3)

^aFishers's exact test. Chi‐square test or Fisher's exact test was used for categorical variables, and Kruskal‐Wallis test was used for continuous variables. Data are presented as mean ± standard deviation or number (percentage).

^b p‐value <0.05, Kruskal‐Wallis test.

Abbreviations: ECMO, Extracorporeal membrane oxygenation; PROM, Passive range of motion.

Figure 2 illustrates, by ECMO week, the number of patients who underwent rehabilitation and the number of sessions performed, relative to the number of patients remaining in the ECMO rehabilitation pathway. In week 1, only approximately 58% of candidates received rehabilitation; however, by weeks 2–4, approximately 80% of eligible patients participated in rehabilitation, indicating that most patients received rehabilitation within the first 4 weeks of ECMO support.

FIGURE 2.

Number of ECMO‐rehabilitated patients and rehabilitation sessions during ECMO. Number of patients who received rehabilitation during ECMO (left) and the number of rehabilitation sessions per ECMO week (right). ECMO, extracorporeal membrane oxygenation.

Figure 3 presents the weekly distribution of patients who graduated from or terminated rehabilitation, categorized by outcome: lung transplantation, clinical improvement, or clinical deterioration/death. As with Figure 2, most patients completed ECMO rehabilitation within 4 weeks. The largest proportion comprised patients who discontinued ECMO after lung transplantation, followed by those with improved clinical status and those who deteriorated or died.

FIGURE 3.

Number of patients who graduated or terminated ECMO rehabilitation at each ECMO week.

Figures 4 and 5 summarize the highest physical activity level achieved per rehabilitation session. As shown in Figure 4, sitting‐level rehabilitation (stage 3) was generally achievable during the first 4 weeks of ECMO support, and no statistically significant intergroup differences were observed. In within‐group comparisons (Figure 5), the BTT group demonstrated significant improvement in mobility stage from ECMO initiation to ICU discharge and from pre‐decannulation to ICU discharge (p < 0.05). The BTR group showed similar patterns, although differences were not statistically significant.

FIGURE 4.

Highest activity level achieved during rehabilitation by ECMO week (Left) The highest physical activity level achieved during each ECMO week; (right) a comparison of the highest activity levels across the three patient groups. ECMO, extracorporeal membrane oxygenation.

FIGURE 5.

Comparison of the highest activity level achieved per session among ICU survivors. Differences over time between groups were analyzed using repeated measures ANOVA.* Significant difference between ECMO initiation and ICU discharge. † Significant difference between before the ECMO removal and ICU discharge. Abbreviation: BTT, bridge to transplantation; BTR, bridge to recovery.

4. Discussion

This single‐center retrospective descriptive study characterized how active rehabilitation was delivered and progressed during ECMO support in routine MICU practice. The main contribution of this study is not to establish superiority among BTT, BTR, and EWD groups, but to provide granular longitudinal benchmarks of ECMO rehabilitation, including weekly rehabilitation candidacy, actual participation, session frequency, interrupted sessions, completion or termination, and activity‐level progression. Previous studies and recent guidelines have mainly addressed the feasibility, safety, predictors, or general principles of ECMO rehabilitation, whereas detailed week‐by‐week descriptions across the ECMO course remain limited [5, 12, 14, 18, 19]. Our findings therefore add real‐world data on how rehabilitation is implemented and sustained in a high‐acuity ECMO population.

As shown in Table 2, rehabilitation characteristics differed across groups. The interval to first rehabilitation did not differ significantly among groups. The BTT group received significantly more weekly rehabilitation sessions, whereas other rehabilitation timing and activity‐level outcomes should be interpreted descriptively. This likely reflects the clear transplant goal in BTT patients, whereas BTR patients were generally older and more often required deep sedation or neuromuscular blockade [4, 11]. When comparing the highest activity level achieved during sessions, no statistically significant between‐group difference was observed in peak activity levels, including analyses dichotomized at stage 4, consistent with prior reports that BTT patients have lower baseline function due to end‐stage lung disease [11, 20]. Across groups, the proportion of patients achieving standing‐level or higher activity increased from ECMO insertion to pre‐decannulation and to ICU discharge. This trajectory was steeper in the BTR group, suggesting that functional status can be preserved through structured rehabilitation during ECMO and that meaningful improvements can occur after decannulation. The comparatively slower trajectory in the BTT group may reflect the additional recovery burden posed by major transplant surgery.

In Figure 2, both the proportion of rehabilitation candidates who received rehabilitation and the number of sessions were higher in ECMO weeks 2–3 than in week 1. This likely reflects early physiological instability after ECMO initiation, during which patients require time to achieve sufficient stability for active rehabilitation [21, 22]. Although week 1 included the largest number of candidates, participation increased thereafter, while interrupted sessions decreased and stabilized after week 2, becoming rare after week 5. This pattern indicates that patient status is highly dynamic in the first 1–2 weeks, requiring vigilant monitoring and individualized adjustment of rehabilitation intensity. Importantly, limited or interrupted rehabilitation early in ECMO should not preclude continued attempts, as safe and consistent rehabilitation became feasible with improved stability even during prolonged ECMO support.

Meanwhile, the numbers of active sessions and rehabilitation candidates declined sharply after ECMO week 3 (Figure 2), a pattern that is also visible in Figure 3. In the BTT group, the mean duration of ECMO support was approximately 21.3 days, consistent with previous reports that ECMO is typically maintained for 2–3 weeks as a BTT [7, 20]. Thus, most BTT patients were likely to undergo lung transplantation within 4 weeks. In the BTR group, the mean duration of ECMO support was also approximately 28 days (Table 1). Consequently, within this 4‐week window, most patients either progressed to transplantation, improved sufficiently to graduate from rehabilitation, or deteriorated to the point of treatment withdrawal [4, 11, 22]. These findings underscore the importance of systematic and intensive management during the first 4 weeks of ECMO rehabilitation.

When examining the highest physical activity level achieved per rehabilitation session by ECMO week, patients generally maintained at least stage 3 activity (Figure 4). The EWD group, however, demonstrated lower mean activity levels than the BTT and BTR groups. Sitting activities require both trunk and upper extremity strength and therefore serve as indicators of relatively preserved functional capacity [23]. Yoshigawa et al. similarly identified edge‐of‐bed sitting as an independent predictor of subsequent walking recovery [24]. In our cohort, even patients requiring ECMO support for more than 4 weeks remained, on average, capable of stage 3 activity. This suggests that prolonged ECMO support does not necessarily lead to functional decline and that meaningful functional maintenance is achievable. Accordingly, if patients meet initial screening criteria, early initiation of active rehabilitation aimed at maintaining at least stage 3 function may positively influence longer‐term outcomes such as ambulation. This pattern is further reflected in the comparison of mean activity levels at ECMO initiation, pre‐decannulation, and ICU discharge among survivors in the BTT and BTR groups (Figure 5): functional status was preserved through the ECMO period and improved substantially after decannulation. These findings align with the primary goals of BTT—to prevent skeletal muscle loss and maintain physical activity—and further support the value of active rehabilitation during ECMO [7].

5. Limitation

This study has several limitations. First, because the cohort included only patients who participated in at least one active rehabilitation session during ECMO support, we could not estimate the overall early mobilization rate among all ICU patients receiving ECMO from ICU admission. In addition, the exact reasons for exclusion of the six patients who did not enter active rehabilitation could not be fully ascertained, as the reasons for non‐participation were not systematically captured in this retrospective dataset. Therefore, the possibility of selection bias cannot be excluded, and our findings should be interpreted as reflecting patients who were considered eligible for and received ECMO rehabilitation at our center, rather than all patients supported with ECMO. Second, subgroup sizes were imbalanced; in particular, the relatively small numbers of BTR and EWD patients and the limited number of deaths may have reduced statistical power and warrant caution when interpreting between‐group comparisons. Finally, our analysis focused on short‐term physical activity levels up to ICU discharge and did not evaluate long‐term functional recovery, survival, or post‐ICU trajectories. As a result, the long‐term impact of ECMO rehabilitation on functional or clinical outcomes cannot be directly inferred.

6. Conclusion

In this single‐center retrospective cohort of patients who received active rehabilitation during ECMO support, rehabilitation was commonly initiated within 2 weeks and completed or discontinued within 4 weeks. Sitting‐level or edge‐of‐bed activity was generally achievable throughout ECMO support, suggesting that basic functional capacity can be maintained in selected patients. These findings provide practical descriptive benchmarks for planning stepwise ECMO rehabilitation in routine ICU practice.

Author Contributions

Woo Chul Son and Hyeon Jeong Moon contributed equally as first authors and were primarily responsible for study conception, data acquisition, data analysis, and drafting of the manuscript. Eun Young Choi participated in data extraction and contributed to manuscript revision. Won Kim supervised the overall study design, data interpretation, and critical revision of the manuscript. Jee Hwan Ahn, Dong‐gon Hyun, Roo Ma Lee, Hyo Jin Lim, Young Hoon Mo, and Sang‐Bum Hong provided clinical support, assisted with data verification, and contributed to the refinement of the final manuscript. All authors reviewed and approved the final version of the manuscript.

Funding

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. RS‐2024‐00346048), and by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (No. RS‐2024‐00408722).

Ethics Statement

The study protocol was approved by the Institutional Review Board of Asan Medical Center (IRB No. 2025‐1216). The requirement for informed consent was waived due to the retrospective nature of the study. All research procedures adhered to the Declaration of Helsinki and followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.

Conflicts of Interest

All authors declare that they have no other competing financial interests or personal relationships that could have influenced the work reported in this paper.

Supporting information

Supporting File: ctr70662‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The datasets generated and analyzed during the current study are not publicly available due to institutional privacy regulations. However, de‐identified data may be made available from the corresponding author upon reasonable request.