Korean clinical practice guidelines for rehabilitation in critically ill patients and post-intensive care syndrome Authors: Chi Ryang Chung, Hye Ju Yeo, Jong Geol Do, Yoonju Na, Eunjeong Choi, Jiyeon Kang, Jeong-Yi Kwon, Hyun Jung Kim, Jinyoung Park, Hak-Jae Lee, Myung Hun Jang, Seung-Tak Kang, Kyung Hyup Koo, Young Seok Kim, Won Kim, Tae-Yoon Kim, Jiyoun Park, Jaewon Beom, Kyung Cheon Seo, Hyun-woo Sun, Woochul Son, Myung-Jun Shin, So Young Ahn, Song I Lee, Min-Hyung Rhee, Yu Hui Won, Myungeun Yoo, Yu Jin Im, Sun Ju Chang, Ye Rim Chang, Mi-Kyeong Jeon, Jin Hee Jung, Yumi Ha, Sung Eun Hyun, Suk-Kyung Hong, Gee Young Suh Published: 2026 ABSTRACT AbstractBackgroundSurvivors of critical illness commonly experience persistent physical, cognitive, and psychological impairments, collectively known as post-intensive care syndrome. Although intensive care unit (ICU) rehabilitation is now recognized as an important part of recovery after critical illness, rehabilitation practices remain inconsistent, and clinical practice guidelines adapted to the Korean healthcare context have not been available.MethodsThese clinical practice guidelines were developed de novo by a multidisciplinary panel comprising critical care physicians, rehabilitation specialists, nurses, and allied health professionals. The panel prioritized 15 key clinical questions addressing adult, pediatric, and neonatal ICU (NICU) rehabilitation. Systematic literature searches were conducted in Medline via PubMed, Embase, the Cochrane Library, and KoreaMed, and the final searches for all 15 questions were completed by November 30, 2024. The certainty of evidence and strength of recommendations were assessed using the Grading of Recommendations Assessment, Development and Evaluation framework. Final recommendations were approved through a structured consensus process requiring at least 70% panel agreement.ResultsThe panel issued 15 recommendations across major rehabilitation domains, including early structured rehabilitation, criteria for withholding or discontinuing therapy, adjunctive modalities, rehabilitation frequency, inspiratory muscle training, swallowing rehabilitation, cognitive and occupational therapy, nutritional support, neonatal and pediatric ICU rehabilitation, and post-ICU follow-up care. Thirteen recommendations were conditional, whereas two were strong, mainly because they addressed patient safety.ConclusionsThese guidelines provide evidence-informed recommendations for rehabilitation in critically ill patients across adult, pediatric, and NICUs in Korea. They are designed to standardize practice, improve patient safety, support patient-centered functional recovery beyond survival, and identify priorities for future research.Keywords: critical care, early mobilization, intensive care unit, postintensive care syndrome, rehabilitation Infographics INTRODUCTION Advances in critical care have improved survival among patients admitted to intensive care units (ICUs), but survival alone does not indicate full recovery [1,2]. Many ICU survivors have persistent physical, cognitive, and psychological impairments after discharge, collectively referred to as post-intensive care syndrome (PICS) [3,4]. In a Korean multicenter prospective cohort, the prevalence of PICS remained approximately 40%–50% at 3, 6, and 12 months after ICU discharge, and some survivors had impairments in multiple domains [5]. These long-term sequelae are associated with functional dependence, lower quality of life, greater healthcare use, and caregiver burden [4,6]. Accordingly, the goals of critical care have expanded beyond survival to include functional recovery and long-term patient-centered outcomes [7]. Rehabilitation during and after critical illness is now regarded as an essential component of ICU care [8]. Evidence supports selected interventions, including early structured rehabilitation, respiratory and swallowing rehabilitation, cognitive and occupational therapy, adjunctive modalities, nutritional support, and post-ICU follow-up care; however, the certainty of evidence differs by intervention and population [9-21]. Several international guidelines have addressed ICU rehabilitation, but applying them directly to Korean practice is difficult because healthcare systems, staffing models, reimbursement structures, resource availability, and implementation feasibility differ across settings. Existing guidelines have also focused mainly on adult ICU populations and early mobilization, with less detailed guidance on pediatric and neonatal ICU (NICU) rehabilitation or post-ICU recovery. Therefore, we developed Korean clinical practice guidelines for rehabilitation in critically ill patients using de novo systematic reviews, the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework, and multidisciplinary consensus. These guidelines address adult, pediatric, and NICU populations and support safe, standardized, patient-centered rehabilitation practices adapted to local resources and individual patient circumstances. MATERIALS AND METHODS Guideline Development and Scope These clinical practice guidelines were developed de novo by a multidisciplinary panel through systematic evidence synthesis, the GRADE framework, and a structured consensus process [22,23]. The panel included specialists in critical care medicine, rehabilitation medicine, respiratory medicine, surgery, pediatric rehabilitation, nursing, physical therapy, occupational therapy, speech-language pathology, nutrition, and related allied health disciplines. The guideline covered rehabilitation during and after ICU care across adult, pediatric, and neonatal populations, including physical, respiratory, swallowing, cognitive, and occupational rehabilitation; nutritional support; adjunctive rehabilitation modalities; and post-ICU follow-up care. Although these recommendations are presented in a single guideline document, the adult, pediatric, and neonatal recommendations were developed around population-specific key clinical questions (KQs). They should therefore be applied only to the population specified for each recommendation, without extrapolation across age groups. This clinical practice guideline was developed and reported in accordance with the Reporting Items for Practice Guidelines in Healthcare statement. A Korean-language summary of the KQ-specific recommendation statements is provided in Supplementary Appendix 1. KQs and Outcomes Fifteen KQs were selected by panel consensus according to clinical importance, practice variation, relevance to the Korean healthcare context, and gaps in existing evidence. Each KQ was structured using the population, intervention, comparator, and outcome framework. Outcomes were classified a priori as critical, important, or less important for decision-making according to GRADE guidance. Literature Search and Evidence Synthesis Systematic literature searches were conducted in Medline via PubMed, Embase, the Cochrane Library, and KoreaMed from database inception, with final searches for all 15 KQs completed by November 30, 2024 (Supplementary Methods 1). Methodology experts and KQ-specific working group members developed the search strategies. Searches were not restricted by publication year, language, or publication status. For each KQ, at least two reviewers independently screened titles, abstracts, and full-text articles according to predefined eligibility criteria (Supplementary Methods 2). Disagreements were resolved through discussion or adjudication by another panel member (Supplementary Methods 3). Study selection was summarized in Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagrams [24]. Data were extracted at the outcome level. Random-effects meta-analyses were performed when studies were sufficiently homogeneous in population, intervention, comparator, and outcome; otherwise, the evidence was synthesized narratively. Risk of Bias and Certainty of Evidence Two reviewers independently assessed risk of bias using validated study design–specific tools: the Cochrane Risk of Bias 2 tool for randomized controlled trials [25], Risk of Bias in Non-randomized Studies of Interventions (ROBINS-I) for nonrandomized studies of interventions [26], and Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) for diagnostic accuracy studies [27]. Disagreements were resolved by consensus (Supplementary Methods 4). The certainty of evidence for each critical and important outcome was assessed using the GRADE framework, with consideration of risk of bias, inconsistency, indirectness, imprecision, publication bias, magnitude of effect, dose–response gradient, and plausible residual confounding, as applicable (Supplementary Methods 5). Formulation of Recommendations Recommendations were formulated using the GRADE Evidence-to-Decision framework [23]. The panel considered desirable and undesirable effects, certainty of evidence, patient values and preferences, the balance of effects, resource requirements, feasibility, equity, acceptability, and applicability to Korean clinical practice (Supplementary Methods 6). Recommendation strength and direction were classified as strong for, conditional for, conditional against, strong against, inconclusive, or expert consensus. Patient and caregiver perspectives were incorporated through a structured experience-based survey of ICU survivors and family caregivers, which was approved by the Institutional Review Board of Samsung Medical Center (No. 2024-02-029). Written informed consent was obtained from each ICU survivor or family caregiver before participation in the survey. Assigned working groups drafted the recommendations, which were then discussed by the full panel and finalized through anonymous blinded electronic voting. A valid vote required participation by at least 70% of eligible panel members, and consensus was predefined as at least 70% agreement among participating voters. Recommendations that did not reach consensus were revised after further discussion and submitted for another vote (Supplementary Methods 7). Detailed search strategies, eligibility criteria, PRISMA flow diagrams, risk-of-bias assessments, summary-of-findings tables, forest plots, Evidence-to-Decision frameworks, and aggregate voting results are provided in Supplementary Appendices 2-8. The guideline will be reviewed every 3 to 5 years, or earlier if new high-quality evidence, major safety concerns, new rehabilitation technologies, or relevant policy changes are expected to alter the recommendations. RESULTS Summary of Recommendations These guidelines addressed 15 KQs on rehabilitation during and after critical illness across adult, pediatric, and NICU populations. Using the GRADE framework and a structured multidisciplinary consensus process, the panel formulated 15 recommendations. Two recommendations were strong: one addressed predefined criteria for withholding or discontinuing physical rehabilitation, and the other addressed swallowing rehabilitation for post-extubation dysphagia. The remaining 13 recommendations were conditional because the evidence was limited or heterogeneous, patient populations and interventions varied, and implementation must be individualized according to clinical stability, patient characteristics, institutional resources, and patient or family preferences. The certainty of evidence was moderate for six recommendations, low for four, and very low for five. Table 1 summarizes all recommendations. Summary-of-findings tables and forest plots for each KQ are provided in Supplementary Tables 1-15 and the Supplementary Figures, respectively. Table 1.Summary of recommendations for rehabilitation in critically ill patientsKQRecommendationStrengthCertainty of evidenceKQ 1We suggest providing structured early rehabilitation soon after ICU admission in critically ill patients, provided that hemodynamic instability or other major contraindications are absent.Conditional forModerateKQ 2We recommend that institutions establish and consistently apply predefined criteria for withholding or stopping physical rehabilitation in critically ill patients.Strong forVery lowKQ 3We suggest using NMES or in-bed cycle ergometry as adjunctive rehabilitation modalities in critically ill patients admitted to the ICU.Conditional forVery lowKQ 4We suggest providing rehabilitation sessions at least twice daily for medically stable critically ill adults at high risk of functional decline who can tolerate repeated sessions.Conditional forVery lowKQ 5We suggest applying IMT in mechanically ventilated critically ill patients.Conditional forModerateKQ 6We suggest not diagnosing dysphagia based solely on bedside swallowing screening after extubation in critically ill patients.Conditional againstVery lowKQ 7We recommend providing swallowing rehabilitation for critically ill patients with dysphagia after extubation.Strong forModerateKQ 8We suggest considering cognitive rehabilitation for critically ill patients admitted to the ICU.Conditional forLowKQ 9We suggest providing occupational therapy as part of multidisciplinary rehabilitation for critically ill patients admitted to the ICU.Conditional forModerateKQ 10We suggest not routinely providing high-dose protein supplementation during ICU rehabilitation in critically ill patients at this time.Conditional againstLowKQ 11We suggest providing developmentally focused physical rehabilitation for neonates admitted to the NICU when clinically appropriate.Conditional forModerateKQ 12We suggest providing feeding and swallowing rehabilitation for neonates admitted to the NICU who have feeding or swallowing difficulties or are expected to have delayed feeding or swallowing development.Conditional forModerateKQ 13We suggest considering early physical rehabilitation selectively for critically ill pediatric ICU patients who meet predefined safety criteria, guided by clinical judgment and patient-specific characteristics.Conditional forLowKQ 14We suggest providing early swallowing evaluation and intervention for pediatric ICU patients when dysphagia is suspected.Conditional forVery lowKQ 15We suggest post-ICU follow-up care for critical illness survivors, considering available clinical resources and individual patient circumstances.Conditional forLowStrength of recommendations and certainty of evidence were assessed using the Grading of Recommendations Assessment, Development and Evaluation framework. Conditional recommendations indicate that different choices may be appropriate depending on patient characteristics, clinical context, patient or family preferences, and resource availability. Strong recommendations may be issued despite low or very low certainty of evidence when patient safety considerations predominate.KQ: key clinical question; ICU: intensive care unit; NMES: neuromuscular electrical stimulation; IMT: inspiratory muscle training; NICU: neonatal intensive care unit. Summary of recommendations for rehabilitation in critically ill patients Strength of recommendations and certainty of evidence were assessed using the Grading of Recommendations Assessment, Development and Evaluation framework. Conditional recommendations indicate that different choices may be appropriate depending on patient characteristics, clinical context, patient or family preferences, and resource availability. Strong recommendations may be issued despite low or very low certainty of evidence when patient safety considerations predominate. KQ: key clinical question; ICU: intensive care unit; NMES: neuromuscular electrical stimulation; IMT: inspiratory muscle training; NICU: neonatal intensive care unit. General Implementation and Population-Specific Considerations These guidelines provide an evidence-informed framework, not a single prescriptive bedside protocol for all critically ill patients. Practical implementation should include daily assessment of clinical stability and rehabilitation eligibility; selection of rehabilitation modality, dose, and intensity according to baseline function, consciousness, participation, and rehabilitation goals; stepwise progression with continuous physiological and device monitoring; predefined criteria for withholding, stopping, reassessing, and resuming rehabilitation; and multidisciplinary documentation, communication, and periodic audit. Institutional protocols should be adapted to available personnel, equipment, monitoring capacity, and other local resources. Additional precautions should be tailored to the clinical population. In medical ICU patients, clinicians should consider respiratory reserve, hemodynamic support, delirium, and ICU-acquired weakness. In surgical patients, rehabilitation planning should incorporate postoperative restrictions, pain control, wound integrity, drains, and procedure-specific precautions. In trauma patients, weight-bearing restrictions, spinal or fracture precautions, neurological status, bleeding risk, and device stability should be reviewed before rehabilitation begins. These considerations are intended to support implementation and do not constitute separate evidence-graded recommendations. Adult, pediatric, and neonatal recommendations should be interpreted as population-specific modules and should not be extrapolated across age groups. Future updates may be organized as separate population-specific modules as the evidence base and implementation experience expand. KEY QUESTIONS AND RECOMMENDATIONS Structured early rehabilitation in critically ill patients admitted to the ICU KQ 1. Is structured early rehabilitation beneficial for critically ill patients admitted to the ICU? Recommendation: We suggest providing structured early rehabilitation soon after ICU admission in critically ill patients, provided that hemodynamic instability or other major contraindications are absent. (Conditional recommendation for; moderate certainty of evidence) Structured rehabilitation refers to a program in which eligibility criteria, patient assessment procedures, and rehabilitation protocols are predefined and delivered through a standardized process. Previous European and Japanese guidelines have recommended initiating early mobilization within 72 hours after ICU admission [11,17,28]. However, the optimal timing, specific components, and safety of rehabilitation remain uncertain across heterogeneous ICU populations. Evidence summary and rationale Nine randomized controlled trials compared structured early rehabilitation initiated within 3 days after ICU admission with usual care [29-37]. At ICU discharge, structured early rehabilitation produced a clinically meaningful improvement in physical function (standardized mean difference [SMD], 1.48; 95% CI, 1.18–1.77; 3 studies; moderate certainty). At hospital discharge, muscle strength was significantly greater in the early rehabilitation group (mean difference [MD], 9.85 points; 95% CI, 8.77–10.94; 3 studies; moderate certainty), whereas activities of daily living did not differ significantly (SMD, 0.13; 95% CI, –0.11 to 0.37; 3 studies; low certainty). ICU and hospital length of stay were statistically shorter (MD, –0.59 days and –0.68 days, respectively), but these absolute reductions did not reach predefined minimally clinically important differences. Hospital mortality did not differ significantly (relative risk [RR], 0.67; 95% CI, 0.35–1.28; 3 studies; moderate certainty). Adverse events were more frequent with early rehabilitation (RR, 1.82; 95% CI, 1.40–2.36; 2 studies; low certainty), but serious events were rare, and most events resolved after temporary interruption or discontinuation. Details are provided in Supplementary Table 1 and Supplementary Figures 1.1-1.6. Implementation considerations The panel judged that the desirable effects probably outweigh potential harms when rehabilitation is delivered to clinically stable patients under predefined safety criteria. Implementation should rely on institutional protocols that address screening, eligibility, contraindications, stepwise mobilization, monitoring, stopping criteria, adverse-event reporting, and multidisciplinary coordination. Criteria for withholding or stopping physical rehabilitation in adult ICU patients KQ 2. Does applying predefined criteria for withholding or stopping physical rehabilitation improve safety outcomes in critically ill adult patients compared with not applying such criteria? Recommendation: We recommend that institutions establish and consistently apply predefined criteria for withholding or stopping physical rehabilitation in critically ill patients. (Strong recommendation for; very low certainty of evidence) This recommendation was graded as strong despite very low certainty because predefined safety criteria address preventable harm, require few additional resources, are broadly acceptable, and align with patient and caregiver values that prioritize safety. Evidence summary and rationale No randomized or observational studies directly compared rehabilitation delivered with predefined withholding or stopping criteria with rehabilitation delivered without such criteria. Evidence was therefore derived indirectly from 13 randomized controlled trials of ICU physical rehabilitation in which physiological thresholds or safety criteria were prespecified and adverse events were systematically reported [30,32,34,35,38-46]. Across the included trials, the pooled incidence of cardiovascular and respiratory adverse events was approximately 1%, patient-related events occurred in approximately 1%–2% of patients, and catheter- or device-related events were rare. Serious adverse events were uncommon, and most events resolved after rehabilitation was paused, modified, or discontinued. However, event definitions, monitoring intensity, rehabilitation dose, intervention type, and patient case mix varied substantially across studies. Because no direct comparator was available and the evidence was indirect, inconsistent, and imprecise, the certainty of evidence was rated as very low. Implementation considerations The panel judged that the safety benefits of predefined criteria outweigh the potential risk of unnecessarily interrupting rehabilitation. These criteria should not be used as rigid cutoffs; instead, they should support structured risk assessment, real-time monitoring, intensity modification, reassessment, and safe resumption. Institutions should adapt criteria to local resources, monitoring capacity, and rehabilitation protocols and incorporate them into daily screening, checklists, electronic records, staff education, adverse-event reporting, and audit. Examples of suggested criteria are provided in Table 2. Details are provided in Supplementary Table 2 and Supplementary Figures 2.1 and 2.2. Table 2.Suggested examples of criteria for withholding or stopping physical rehabilitation in critically ill patientsDomainSuggested criteria (example)RespiratoryPEEP >10 cm H₂O; FiO₂ >0.6; SpO₂ <90% or a sustained drop of ≥20% from baseline; respiratory rate <5 or >35 breaths/min; increasing oxygen or ventilatory support requirementHemodynamicSystolic blood pressure <90 or >200 mm Hg; mean arterial pressure <65 or >110 mm Hg; heart rate <40 or >130 beats/min or a >20% change from baseline; new-onset or uncontrolled arrhythmia; need for vasopressor dose escalationNeurologicalDecreased level of consciousness; agitation or delirium preventing safe participation; new focal neurological deficitPatient toleranceSevere fatigue, dizziness, new chest pain or dyspnea, intolerable pain, or patient request to stopDevice-relatedRisk of dislodgement or accidental removal of endotracheal tube, tracheostomy, vascular catheter, drainage tube, or other critical devicesOthersActive bleeding; inability to cooperate with the rehabilitation program; any other condition judged by the treating team to pose unacceptable riskThese criteria are provided as illustrative examples and should be adapted according to institutional protocols, available monitoring, rehabilitation intensity, and individual patient conditions. Suggested numeric thresholds are not intended as mandatory cutoffs and should be interpreted in conjunction with overall clinical judgment. Clinicians may choose to withhold, modify, continue, pause, or stop rehabilitation depending on the overall clinical context.PEEP: positive end-expiratory pressure; FiO₂: fraction of inspired oxygen; SpO₂: peripheral oxygen saturation. Suggested examples of criteria for withholding or stopping physical rehabilitation in critically ill patients These criteria are provided as illustrative examples and should be adapted according to institutional protocols, available monitoring, rehabilitation intensity, and individual patient conditions. Suggested numeric thresholds are not intended as mandatory cutoffs and should be interpreted in conjunction with overall clinical judgment. Clinicians may choose to withhold, modify, continue, pause, or stop rehabilitation depending on the overall clinical context. PEEP: positive end-expiratory pressure; FiO₂: fraction of inspired oxygen; SpO₂: peripheral oxygen saturation. Neuromuscular electrical stimulation (NMES) and in-bed cycle ergometry as adjunctive rehabilitation modalities KQ 3. Should NMES or in-bed cycle ergometry be used as adjunctive rehabilitation modalities in critically ill patients admitted to the ICU? Recommendation: We suggest using NMES or in-bed cycle ergometry as adjunctive rehabilitation modalities in critically ill patients admitted to the ICU. (Conditional recommendation for; very low certainty of evidence) Evidence summary and rationale Thirty-one randomized controlled trials evaluated NMES, in-bed cycle ergometry, or both in adult ICU patients [38,39,42,47-74]. NMES was associated with a significantly shorter hospital length of stay (MD, –5.13 days; 95% CI, –8.84 to –1.41; 4 studies; low certainty). However, its effects were uncertain for muscle strength, assessed using the Medical Research Council (MRC) sum score (SMD, 0.49; 95% CI, –0.09 to 1.07; 9 studies); muscle mass (SMD, 0.71; 95% CI, –0.36 to 1.78; 8 studies); ICU length of stay (MD, –3.15 days; 95% CI, –8.88 to 2.59; 5 studies); and duration of mechanical ventilation (MD, –5.45 days; 95% CI, –10.92 to 0.02; 6 studies), with very low certainty for most outcomes. In-bed cycle ergometry improved 6-minute walking distance (MD, 83.85 m; 95% CI, 67.63–100.14; 3 studies; very low certainty) and shortened hospital length of stay (MD, –3.31 days; 95% CI, –5.76 to –0.85; 2 studies; very low certainty), but it did not significantly affect ICU length of stay or duration of mechanical ventilation. Combined NMES and in-bed cycle ergometry did not significantly improve the MRC sum score. Adverse events were generally mild and transient, and serious adverse events were rare; however, adverse-event definitions and monitoring methods varied across studies. Outcome-specific certainty ranged from low to very low. Because most outcomes, including the MRC sum score and 6-minute walking distance, were rated as very low certainty, the overall certainty of evidence for KQ 3 was rated as very low. The observed benefits for hospital length of stay and 6-minute walking distance, together with the rarity of serious adverse events, supported using these interventions as adjunctive modalities. However, the absence of consistent benefit for muscle strength, muscle mass, ICU length of stay, and duration of mechanical ventilation supported a conditional rather than strong recommendation. Implementation considerations The panel judged that NMES and in-bed cycle ergometry may be useful adjuncts, particularly for patients who cannot participate fully in active mobilization because of weakness, sedation, delirium, or ventilatory support. These modalities should supplement, rather than replace, conventional active, goal-directed rehabilitation once patients are able to engage. Implementation requires appropriate patient selection, staff training, monitoring, skin and electrode checks for NMES, line and tube management during cycle ergometry, and consideration of equipment availability. Details are provided in Supplementary Table 3 and Supplementary Figures 3.1-3.10. Frequency of rehabilitation sessions in adult ICU patients KQ 4. Is providing multiple rehabilitation sessions per day effective in critically ill adult patients admitted to the ICU? Recommendation: We suggest providing rehabilitation sessions at least twice daily for medically stable critically ill adults at high risk of functional decline who can tolerate repeated sessions. (Conditional recommendation for; very low certainty of evidence) Evidence summary and rationale Twenty-two randomized controlled trials compared once-daily rehabilitation with two or more rehabilitation sessions per day in adult ICU patients [30,32,38,40,41,43,52,53,67,74-86]. Across most short-term clinical outcomes, multiple daily rehabilitation sessions showed no clear benefit. Duration of mechanical ventilation did not differ significantly (MD, –0.50 days; 95% CI, –1.68 to 0.68; 15 studies; low certainty). ICU length of stay showed a nonsignificant reduction (MD, –0.66 days; 95% CI, –1.94 to 0.62; 20 studies; very low certainty), and hospital length of stay did not differ significantly (MD, 0.11 days; 95% CI, –0.48 to 0.70; 17 studies; very low certainty). ICU mortality and in-hospital mortality were also unaffected. For functional outcomes, the MRC sum score did not differ significantly at ICU discharge, hospital discharge, or 3 months after ICU discharge, although a small improvement was observed at 6 months (MD, 1.47; 95% CI, 0.55–2.38; 2 studies; very low certainty). Health-related quality of life did not differ significantly at hospital discharge, but modest improvements were observed at 6 months (MD, 2.74; 95% CI, 0.99–4.48; 6 studies; very low certainty) and 12 months (MD, 3.34; 95% CI, 1.05–5.63; 4 studies; very low certainty). The clinical importance of these long-term differences remains uncertain. Adverse events, including respiratory, cardiovascular, catheter-related, and patient-related events, were not significantly increased, and no serious adverse events were reported. Implementation considerations Given the very low certainty of evidence, uncertain short-term benefit, and greater resource requirements, multiple daily sessions should not be implemented routinely for all ICU patients. They may be considered selectively for medically stable adults at high risk of functional decline who can tolerate repeated sessions, such as patients with severe weakness, prolonged mechanical ventilation, extracorporeal membrane oxygenation support, or insufficient response to a single daily session. Practical alternatives include shorter divided sessions, longer or more intense single sessions, nurse-led mobilization, and safe self-directed exercise when appropriate. Table 3 provides examples of patient groups in whom multiple daily sessions may be considered. Details are provided in Supplementary Table 4 and Supplementary Figures 4.1-4.7. Table 3.Examples of patient groups in whom multiple daily rehabilitation sessions may be conditionally consideredCategoryExampleMedically stable patientsHemodynamically stable patients; FiO₂ <0.6; receiving low-dose and non-escalating vasopressorsHigh risk of functional declineOlder age; mechanical ventilation; ECMO support; after major surgery, including organ transplantationAble to participate in rehabilitationAble to follow simple commands; no uncontrolled agitation or delirium preventing safe participation; analgesia and sedation appropriately controlledPatients insufficiently responsive to a single daily sessionSevere weakness allowing only short exercise duration; patients who may benefit from splitting sessions because of fatiguePatients with rapid recovery between sessionsPatients who tolerated previous sessions well and recovered quickly enough to permit an additional sessionThese examples are provided to support clinical decision-making and should not be interpreted as mandatory criteria. Application should be individualized based on overall clinical judgment, patient tolerance, rehabilitation goals, safety criteria, and institutional resources.FiO₂: fraction of inspired oxygen; ECMO: extracorporeal membrane oxygenation. Examples of patient groups in whom multiple daily rehabilitation sessions may be conditionally considered These examples are provided to support clinical decision-making and should not be interpreted as mandatory criteria. Application should be individualized based on overall clinical judgment, patient tolerance, rehabilitation goals, safety criteria, and institutional resources. FiO₂: fraction of inspired oxygen; ECMO: extracorporeal membrane oxygenation. Inspiratory muscle training (IMT) in mechanically ventilated critically ill patients KQ 5. Is IMT beneficial for mechanically ventilated critically ill patients? Recommendation: We suggest applying IMT in mechanically ventilated critically ill patients. (Conditional recommendation for; moderate certainty of evidence) Evidence summary and rationale Fourteen randomized controlled trials compared IMT with usual care alone in mechanically ventilated critically ill patients [87-100]. IMT significantly improved maximal inspiratory pressure compared with control (MD, 6.86 cm H₂O; 95% CI, 5.84–7.88; 13 studies; moderate certainty), although the magnitude of improvement did not exceed the predefined minimally clinically important difference. IMT also improved several respiratory and weaning-related outcomes. The rapid shallow breathing index decreased significantly (MD, –6.06; 95% CI, –10.55 to –1.57; 5 studies; low certainty), duration of mechanical ventilation was shorter (MD, –1.92 days; 95% CI, –2.79 to –1.05; 7 studies; moderate certainty), and weaning duration was reduced (MD, –0.73 days; 95% CI, –1.15 to –0.30; 6 studies; low certainty). ICU length of stay was also shorter in the IMT group (MD, –4.46 days; 95% CI, –5.35 to –3.56; 3 studies; low certainty). IMT increased the weaning success rate (RR, 1.32; 95% CI, 1.05–1.66; 5 studies; low certainty), whereas in-hospital mortality and reintubation rate did not differ significantly. Adverse events were not meta-analyzed because definitions and reporting methods varied across studies. No serious IMT-related adverse events were reported, although one trial described interrupted sessions because of paradoxical breathing, tachypnea, desaturation, hemodynamic instability, or supraventricular tachycardia. Implementation considerations The panel judged that IMT may be considered for clinically stable mechanically ventilated patients with respiratory muscle weakness, delayed weaning, or difficulty achieving ventilator liberation. IMT should be delivered by trained clinicians who can assess alertness, cooperation, airway safety, secretion burden, oxygenation, hemodynamic stability, and contraindications such as suspected pharyngeal injury or markedly increased intracranial pressure. Details are provided in Supplementary Table 5 and Supplementary Figures 5.1-5.4. Bedside swallowing screening after extubation in critically ill patients KQ 6. Is bedside swallowing screening alone sufficient to diagnose dysphagia after extubation in critically ill patients? Recommendation: We suggest not diagnosing dysphagia based solely on bedside swallowing screening after extubation in critically ill patients. (Conditional recommendation against; very low certainty of evidence) Evidence summary and rationale Seven diagnostic accuracy studies evaluated bedside swallowing screening after extubation in critically ill patients [101-107]. All included studies used fiberoptic endoscopic evaluation of swallowing as the reference standard. Two studies evaluated the Gugging Swallowing Screen for the intensive care unit (GUSS-ICU), whereas the remaining studies used other bedside screening approaches, including bedside swallowing evaluation, cough reflex testing, modified volume-viscosity swallow testing, and other clinical screening methods. In most studies, the reference standard was performed within 24 hours after the screening test. In the pooled analysis, bedside swallowing screening had limited and variable diagnostic accuracy, with a pooled sensitivity of 0.78 (95% CI, 0.60–0.89) and pooled specificity of 0.70 (95% CI, 0.44–0.88). Diagnostic performance varied substantially across studies according to the screening method, patient population, assessment timing, and post-extubation clinical status. In a subgroup analysis of two studies using GUSS-ICU, diagnostic performance appeared better, with a sensitivity of 0.86 (95% CI, 0.76–0.93) and specificity of 0.89 (95% CI, 0.65–0.99) [102,103]. However, this subgroup included few studies, and the certainty of evidence was rated as very low because of risk-of-bias concerns, indirectness, inconsistency, and imprecision. Implementation considerations Bedside swallowing screening may be useful as an initial safety check or triage tool after extubation, but it should not be used as the sole basis for confirming or excluding dysphagia. Patients with positive screening results, high aspiration risk, weak cough, altered mental status, neurological disease, persistent respiratory instability, or unclear findings should undergo further clinical evaluation and, when appropriate, instrumental assessment such as fiberoptic endoscopic evaluation of swallowing or a videofluoroscopic swallowing study. Details are provided in Supplementary Tables 6.1 and 6.2, and Supplementary Figures 6.1-6.4. Swallowing rehabilitation for post-extubation dysphagia in critically ill patients KQ 7. Does swallowing rehabilitation improve clinical outcomes in critically ill patients with dysphagia after extubation compared with no swallowing rehabilitation? Recommendation: We recommend providing swallowing rehabilitation for critically ill patients with dysphagia after extubation. (Strong recommendation for; moderate certainty of evidence) Evidence summary and rationale Four studies evaluated swallowing rehabilitation in critically ill patients after endotracheal extubation: three randomized controlled trials and one nonrandomized controlled study [108-111]. The included studies enrolled patients who had been successfully extubated after prolonged endotracheal intubation, most commonly for at least 48 hours, and assessed outcomes within 7–30 days after extubation. In the three randomized controlled trials, swallowing rehabilitation significantly shortened the time to resumption of oral feeding compared with no swallowing rehabilitation (MD, –0.85 days; 95% CI, –1.37 to –0.33; 198 participants; 3 studies; moderate certainty). One randomized controlled trial reported improved salivary flow 7 days after extubation (MD, 7.56 mm; 95% CI, 5.67–9.45; 90 participants; low certainty). Another randomized controlled trial showed a lower incidence of pneumonia in the swallowing rehabilitation group than in the control group (odds ratio [OR], 0.24; 95% CI, 0.10–0.56; 117 participants; moderate certainty). The nonrandomized study also suggested earlier resumption of oral feeding with swallowing and oral care intervention (hazard ratio, 1.77; 95% CI, 1.08–2.91), although baseline imbalance and study-design limitations should be considered. Evidence was downgraded mainly for imprecision and, for salivary flow, risk-of-bias concerns and reliance on a single study. Implementation considerations The panel issued a strong recommendation because swallowing rehabilitation improved clinically important outcomes, may reduce aspiration-related complications, is generally acceptable to patients and clinicians, and requires relatively limited resources when speech-language pathology or trained rehabilitation personnel are available. Rehabilitation should be individualized according to swallowing status, airway protection, cough strength, mental status, respiratory stability, secretion burden, aspiration risk, and nutritional route. Instrumental assessment should be considered when aspiration risk is high or bedside findings are unclear. Details are provided in Supplementary Table 7 and Supplementary Figures 7.1-7.3. Cognitive rehabilitation in adult ICU patients KQ 8. Does cognitive rehabilitation improve cognitive and delirium-related outcomes in critically ill adult patients admitted to the ICU? Recommendation: We suggest considering cognitive rehabilitation for critically ill patients admitted to the ICU. (Conditional recommendation for; low certainty of evidence) Evidence summary and rationale Six randomized controlled trials evaluated cognitive rehabilitation or cognitive training interventions in adult ICU patients [112-117], and five were included in the meta-analysis [112-116]. Interventions included cognitive rehabilitation alone, occupational therapist–guided cognitive intervention, physical-cognitive therapy, and combined cognitive and physical rehabilitation. Cognitive rehabilitation was associated with a small but statistically significant improvement in overall cognitive function 3 months after ICU discharge, as assessed using tools such as the Mini-Mental State Examination and Montreal Cognitive Assessment (MD, 1.11; 95% CI, 0.32–1.89; 2 studies; low certainty) [113,115]. However, this improvement was smaller than the predefined minimally clinically important difference. Executive function, health-related quality of life, delirium occurrence, and delirium duration did not differ significantly between groups. No serious adverse events directly attributable to cognitive rehabilitation were reported. Evidence was downgraded mainly because the evidence base was small, risk-of-bias concerns were present, interventions and outcome measures were heterogeneous, and estimates were imprecise. Implementation considerations The panel judged that cognitive rehabilitation may be considered selectively for ICU patients at risk of cognitive impairment, particularly when patients are alert, cooperative, recovering from delirium, and able to communicate or engage in structured tasks. Because the observed benefits were small and evidence for delirium, executive function, and quality of life was inconclusive, routine use in all ICU patients is not supported. Interventions should be tailored to consciousness level, delirium status, fatigue, communication ability, sensory impairment, and rehabilitation goals. Details are provided in Supplementary Table 8 and Supplementary Figures 8.1-8.5. Occupational therapy in adult ICU patients KQ 9. In adult patients admitted to the ICU, does occupational therapy improve cognitive and physical recovery compared with usual care or no occupational therapy? Recommendation: We suggest providing occupational therapy as part of multidisciplinary rehabilitation for critically ill patients admitted to the ICU. (Conditional recommendation for; moderate certainty of evidence) Evidence summary and rationale Eight studies evaluated occupational therapy, or rehabilitation programs that included occupational therapy, in adult ICU patients: six randomized controlled trials and two nonrandomized before-and-after studies [32,34,112,114,118-121]. Occupational therapy was associated with significant improvements in several patient-centered outcomes at hospital discharge. Activities of daily living improved modestly in the intervention group (SMD, 0.37; 95% CI, 0.03–0.72; 2 studies; moderate certainty), although the effect size did not clearly exceed the predefined minimally clinically important difference. Physical function, assessed mainly using grip strength or functional performance measures, also improved (SMD, 0.68; 95% CI, 0.36–1.00; 2 studies; moderate certainty), and cognitive function was better in the intervention group (SMD, 0.50; 95% CI, 0.15–0.84; 2 studies; moderate certainty). Occupational therapy was associated with shorter ICU length of stay (MD, –1.32 days; 95% CI, –2.40 to –0.24; 3 studies; low certainty), but hospital length of stay and mortality did not clearly decrease. One trial suggested improved quality of life at hospital discharge, and another suggested a lower risk of cognitive impairment at 1 year; however, both findings were based on limited evidence. Adverse-event reporting was incomplete, but serious occupational therapy–related adverse events were rare. Implementation considerations The panel judged that occupational therapy may provide meaningful benefits when incorporated into multidisciplinary ICU rehabilitation, particularly for patients with impaired activities of daily living, cognitive dysfunction, delirium risk, weakness, or difficulty resuming self-care tasks. Interventions should be tailored to arousal level, delirium status, communication ability, upper-limb function, fatigue, hemodynamic and respiratory stability, and rehabilitation goals. Details are provided in Supplementary Table 9 and Supplementary Figures 9.1-9.10. High-dose protein supplementation during ICU rehabilitation KQ 10. Does high-dose protein supplementation during ICU rehabilitation improve outcomes in critically ill patients compared with standard protein provision? Recommendation: We suggest not routinely providing high-dose protein supplementation during ICU rehabilitation in critically ill patients at this time. (Conditional recommendation against; low certainty of evidence) This recommendation does not discourage appropriate nutritional assessment or standard protein provision; rather, it indicates that routine high-dose supplementation beyond standard targets during ICU rehabilitation is not currently supported. Evidence summary and rationale Two randomized controlled trials evaluated high-dose protein supplementation in critically ill patients receiving rehabilitation [122,123]. One trial compared a high-protein target of 2.0–2.2 g/kg/day with a standard-protein target of 1.4–1.5 g/kg/day in patients receiving early rehabilitation, whereas the other compared 1.8 g/kg/day with 0.9 g/kg/day in patients undergoing ICU rehabilitation. High-dose protein supplementation did not consistently improve clinically important outcomes, including ICU length of stay, hospital length of stay, duration of mechanical ventilation, or functional status; its effects on mortality remain uncertain. One trial suggested less loss of femoral muscle volume in the high-protein group, but the clinical importance of this finding remains uncertain because no minimally clinically important difference has been established. Overall, the certainty of evidence was rated as low because the evidence came from few trials, risk-of-bias concerns were present, protein targets varied, and estimates were imprecise. Implementation considerations Adverse events were not directly evaluated in the two included trials [122,123]. However, indirect evidence from trials and meta-analyses comparing higher versus lower protein delivery in critically ill patients suggests potential concerns, including increased urea levels and possible harm in patients with acute kidney injury [124,125]. Therefore, protein provision should be tailored to illness phase, nutritional risk, renal function, metabolic tolerance, feeding route, and rehabilitation goals. High-dose supplementation should be used cautiously, particularly in patients with renal dysfunction or poor metabolic tolerance. Details are provided in Supplementary Table 10 and Supplementary Figure 10. Developmentally focused physical rehabilitation in neonates admitted to the NICU KQ 11. Does developmentally focused physical rehabilitation improve clinical and developmental outcomes in neonates admitted to the NICU? Recommendation: We suggest providing developmentally focused physical rehabilitation for neonates admitted to the NICU when clinically appropriate. (Conditional recommendation for; moderate certainty of evidence) Evidence summary and rationale Sixty-two randomized controlled trials evaluated developmentally focused physical rehabilitation in the NICU [126-187], and 59 trials were included in the meta-analysis. Interventions included kangaroo mother care, skin-to-skin contact, individualized developmental care, tactile–kinesthetic or multisensory stimulation, therapist-led rehabilitation, and caregiver-assisted interventions. Compared with standard NICU care, developmentally focused physical rehabilitation shortened hospital or NICU length of stay (MD, –5.89 days; 95% CI, –6.20 to –5.58; 19 studies; 2,166 participants; moderate certainty), exceeding the predefined minimally clinically important difference of 3 days. Duration of mechanical ventilation did not differ significantly between groups (MD, –1.23 days; 95% CI, –3.51 to 1.05; 3 studies; low certainty). Developmentally focused physical rehabilitation was also associated with lower caregiver stress (SMD, –0.87; 95% CI, –1.05 to –0.68; 8 studies; moderate certainty). Developmental and growth outcomes generally favored the intervention. Cognitive development improved significantly (SMD, 0.58; 95% CI, 0.40–0.75; 13 studies; low certainty), and motor development showed a small improvement (SMD, 0.20; 95% CI, 0.06–0.35; 16 studies; very low certainty). Growth outcomes, including body weight, height, and head circumference, also favored the intervention. However, because neonatal growth is influenced by multiple factors and the included studies were clinically heterogeneous, the observed effect sizes did not exceed the predefined minimally clinically important differences, limiting assessment of clinical significance. Safety outcomes did not suggest increased physiological instability; respiratory rate and apnea incidence favored the intervention group, suggesting possible benefits for cardiorespiratory stability, whereas heart rate did not differ significantly. However, most studies excluded neonates with marked hemodynamic instability, increasing ventilatory requirements, major congenital anomalies, uncontrolled infection, or other high-risk conditions. Therefore, these safety findings should be applied primarily to clinically stable neonates. Implementation considerations The panel judged that benefits probably outweigh harms when developmentally focused rehabilitation is applied to clinically appropriate neonates in the NICU. Implementation should include standardized training, institution-specific standard operating protocols with clear eligibility and safety-monitoring criteria, and multidisciplinary collaboration among rehabilitation physicians, neonatologists, NICU nurses, therapists, and caregivers. When therapist-led care is limited, structured caregiver education may be a practical complement. Details are provided in Supplementary Table 11 and Supplementary Figures 11.1-11.12. Feeding and swallowing rehabilitation in neonates admitted to the NICU KQ 12. Does feeding and swallowing rehabilitation improve clinical outcomes in neonates admitted to the NICU compared with no feeding or swallowing rehabilitation? Recommendation: We suggest providing feeding and swallowing rehabilitation for neonates admitted to the NICU who have feeding or swallowing difficulties or are expected to have delayed feeding or swallowing development. (Conditional recommendation for; moderate certainty of evidence) This recommendation should be applied only when neonates are medically stable and feeding or swallowing intervention is judged to be safe. Evidence summary and rationale Nineteen randomized controlled trials evaluated feeding and swallowing rehabilitation in neonates admitted to the NICU [188-206]. Interventions included oral motor stimulation, non-nutritive sucking, pre-feeding oral stimulation, premature infant oral motor intervention, pacifier-based interventions, and multistimulation approaches. Compared with standard NICU feeding care, feeding and swallowing rehabilitation significantly shortened hospital or NICU length of stay (MD, –4.39 days; 95% CI, –4.93 to –3.85; 14 studies; moderate certainty) and time to independent oral feeding (MD, –4.36 days; 95% CI, –4.66 to –4.07; 12 studies; moderate certainty). Both effects exceeded the predefined minimally clinically important differences. Feeding and swallowing rehabilitation also improved feeding performance, including overall intake (SMD, 0.87; 95% CI, 0.51–1.22; 2 studies; moderate certainty) and rate of transfer (MD, 0.48; 95% CI, 0.25–0.71; 3 studies; low certainty). Body weight at discharge did not differ significantly between groups. The lower body weight observed at the time of independent oral feeding was interpreted as reflecting earlier achievement of oral feeding rather than impaired growth. Available safety outcomes suggested no increased risk: desaturation, bradycardia, and feeding intolerance did not differ significantly, whereas gastrointestinal motility disturbance was less frequent in the intervention group. However, most studies excluded neonates with marked hemodynamic instability, increasing ventilatory requirements, major congenital anomalies, or other high-risk conditions. Therefore, these safety findings should be applied primarily to clinically stable neonates. Evidence was downgraded mainly for risk of bias, imprecision, and heterogeneity in intervention protocols and outcome definitions. Implementation considerations Feeding and swallowing rehabilitation should be tailored to gestational or corrected age, physiological stability, respiratory support, oral readiness, sucking–swallowing–breathing coordination, aspiration risk, fatigue, stress cues, and feeding tolerance. Trained clinicians should deliver these interventions in collaboration with neonatologists, NICU nurses, therapists, and caregivers. Details are provided in Supplementary Table 12 and Supplementary Figures 12.1-12.6. Early physical rehabilitation in pediatric ICU patients KQ 13. In critically ill pediatric ICU patients, does early physical rehabilitation improve clinical outcomes compared with usual care? Recommendation: We suggest considering early physical rehabilitation selectively for critically ill pediatric ICU patients who meet predefined safety criteria, guided by clinical judgment and patient-specific characteristics. (Conditional recommendation for; low certainty of evidence) Evidence summary and rationale Six studies evaluated early physical rehabilitation in pediatric ICU patients: one randomized controlled trial and five nonrandomized studies [207-212]. Interventions included early mobilization, out-of-bed activity, protocolized rehabilitation, and multidisciplinary rehabilitation programs. Comparator care consisted of usual pediatric ICU care, including basic passive range-of-motion exercise when applicable. In the randomized trial, early physical rehabilitation did not reduce pediatric ICU length of stay (MD, 1.50 days; 95% CI, –4.07 to 7.07; 1 study; low certainty), hospital length of stay (MD, 0.00 days; 95% CI, –6.73 to 6.73; 1 study; low certainty), or functional score (MD, –1.00; 95% CI, –2.11 to 0.11; 1 study; low certainty). Functional outcomes favored early rehabilitation directionally, but the estimate was imprecise and did not reach statistical significance. In the nonrandomized studies, pooled effects on functional outcomes were small and uncertain, and duration of mechanical ventilation did not differ significantly (MD, 0.45 days; 95% CI, –0.23 to 1.12; very low certainty). Risk of bias was a major limitation because the studies had small samples, limited blinding, loss to follow-up, limitations in comparator selection, and potential residual confounding. Implementation considerations The panel judged that early physical rehabilitation should be considered for selected pediatric ICU patients rather than applied routinely to all such patients. Patient selection should account for age, developmental stage, diagnosis, baseline function, hemodynamic and respiratory stability, sedation and delirium status, neuromuscular weakness, device burden, ability to participate, and family involvement. Programs should use predefined safety criteria, stepwise progression, close monitoring, and multidisciplinary coordination. Details of the evidence profile and forest plots are provided in Supplementary Table 13 and Supplementary Figures 13.1-13.5. Early swallowing evaluation and intervention in pediatric ICU patients KQ 14. In pediatric ICU patients with suspected dysphagia, does early swallowing evaluation and intervention improve clinical outcomes compared with usual care without swallowing rehabilitation? Recommendation: We suggest providing early swallowing evaluation and intervention for pediatric ICU patients when dysphagia is suspected. (Conditional recommendation for; very low certainty of evidence) Evidence summary and rationale Two nonrandomized studies evaluated early swallowing evaluation and intervention in pediatric ICU patients with suspected dysphagia [213,214]. One retrospective observational study enrolled children with oral-to-esophageal injury after chemical or button battery ingestion, and one quasi-experimental prospective study enrolled children with suspected post-extubation dysphagia. Pediatric ICU length of stay did not differ between groups (MD, –0.50 days; 95% CI, –3.95 to 2.95; 1 study; very low certainty). Hospital length of stay was shorter in the intervention group (MD, –5.00 days; 95% CI, –6.71 to –3.29; 1 study; very low certainty). Feeding status at discharge favored early swallowing evaluation and intervention but did not reach statistical significance (RR, 1.14; 95% CI, 0.84–1.55; 1 study; very low certainty). In contrast, fewer patients were discharged with limited oral intake or nothing by mouth in the intervention group (RR, 0.04; 95% CI, 0.00–0.60; 1 study; very low certainty). No adverse events were directly attributed to the intervention, although safety outcomes were not systematically assessed. Overall certainty was very low because all evidence came from nonrandomized studies with serious risk of bias, small samples, indirectness, and imprecision. Implementation considerations Early swallowing evaluation and intervention should be considered for pediatric ICU patients with suspected dysphagia after extubation or upper aerodigestive injury. Patient selection should incorporate airway protection, respiratory stability, cough strength, mental status, secretion burden, nutritional route, aspiration risk, and access to speech-language pathologists or other trained rehabilitation clinicians. Instrumental swallowing assessment should be considered when aspiration risk is high or bedside findings are unclear. Details of the evidence profile and forest plots are provided in Supplementary Table 14 and Supplementary Figures 14.1-14.2. Post-ICU follow-up care for critical illness survivors KQ 15. Does post-ICU follow-up care improve outcomes in critical illness survivors compared with usual care? Recommendation: We suggest post-ICU follow-up care for critical illness survivors, considering available clinical resources and individual patient circumstances. (Conditional recommendation for; low certainty of evidence) Evidence summary and rationale Thirteen randomized controlled trial reports, representing 12 studies, evaluated post-ICU follow-up care for survivors of critical illness [215-227]. Intervention models varied substantially and included shared-care or case-management clinics, ICU recovery or multidisciplinary follow-up programs, counseling programs, telephone- or web-based interventions, telemedicine clinics, and rehabilitation-oriented programs. Comparator care was usual care after ICU discharge. Outcomes were assessed for up to 12 months after discharge. Across pooled analyses, post-ICU follow-up care did not produce clear improvements in psychological outcomes. Anxiety symptoms were not significantly reduced (SMD, –0.05; 95% CI, –0.20 to 0.09; 3 studies; low certainty), nor were depression symptoms (SMD, –0.03; 95% CI, –0.16 to 0.10; 4 studies; low certainty) or post-traumatic stress disorder symptoms (SMD, –0.09; 95% CI, –0.23 to 0.05; 4 studies; low certainty). The incidence of anxiety, depression, and post-traumatic stress disorder also did not differ significantly between groups. Cognitive function, physical function, and health-related quality of life showed no significant improvement with follow-up care. Evidence for healthcare utilization remained uncertain, with no significant differences in emergency department visits or hospital readmissions. Mortality up to 12 months was not reduced (RR, 0.92; 95% CI, 0.75–1.11; 11 studies; low certainty). The evidence was downgraded because intervention models were heterogeneous, outcome measures varied, long-term attrition was common, risk-of-bias concerns were present, and effect estimates were imprecise. Implementation considerations Although pooled trial evidence did not show clear outcome improvement, the panel judged that structured post-ICU follow-up may help identify unmet needs. Follow-up care can help identify persistent physical, cognitive, psychological, nutritional, and social needs; coordinate referrals; support medication and symptom review; and guide patient-centered recovery planning. Implementation should be individualized according to local resources, survivor risk profile, persistent symptoms, functional limitations, caregiver burden, and availability of multidisciplinary follow-up services. Details of the evidence profile and forest plots are provided in Supplementary Table 15 and Supplementary Figures 15.1-15.6. LIMITATIONS These guidelines should be interpreted in light of several limitations. First, adult, pediatric, and neonatal critically ill patients are clinically heterogeneous populations. Although recommendations were developed separately for each population, presenting them in a single document may reduce bedside usability for clinicians caring for any one group. Second, the available evidence was insufficient to support separate recommendations for medical, surgical, and trauma ICU subgroups. Third, most recommendations were conditional, and more than half were supported by low- or very-low-certainty evidence. Future updates may be better organized as separate population-specific guidelines, with subgroup-specific protocols added as evidence and implementation experience accumulate. KEY MESSAGES ▪ These guidelines provide the first evidence-based recommendations for rehabilitation in critically ill patients across adult, pediatric, and neonatal intensive care units (ICUs) in Korea. ▪ Most recommendations are conditional because the supporting evidence is limited or heterogeneous; the strong recommendations focus on patient safety and standardized criteria for rehabilitation practice. ▪ The guidelines support safe, patient-centered, interdisciplinary rehabilitation practices that promote functional recovery beyond survival across diverse ICU settings. Footnotes CONFLICT OF INTEREST No potential conflict of interest relevant to this article was reported. FUNDING This research was supported by a grant from the Patient-Centered Clinical Research Coordinating Center (PACEN), funded by the Ministry of Health & Welfare, Republic of Korea (grant No. RS-2024-00397935). ACKNOWLEDGMENTS None. AUTHOR CONTRIBUTIONS Conceptualization: JGD, YN, GYS. Data curation: all authors. Formal analysis: all authors. Methodology: JGD, YN, CRC, HJY. Project administration: JGD, YN. Supervision: GYS, HJY. Funding acquisition: GYS. Writing – original draft: JGD, HJY, CRC, YN. Writing – review & editing: all authors. All authors read and agreed to the published version of the manuscript. SUPPLEMENTARY MATERIALS Supplementary materials can be found via https://doi.org/10.4266/acc.006630. REFERENCES Associated Data Supplementary Materials