Abstract
Purpose
To determine whether indirect calorimetry (IC)-guided nutritional prescriptions improve weaning success and time to liberation from prolonged mechanical ventilation (PMV) compared with predictive-equation-guided feeding.
Methods
In this double-blind randomized controlled trial conducted at Reuth Tel-Aviv Rehabilitation Hospital (Nov 2019-Apr 2023), with 1-year follow-up, adults ventilated ≥ 21 days were randomized to receive IC-guided energy targets or predictive-equation-guided feeding. Primary outcomes were successful weaning (≥ 7 consecutive days of spontaneous breathing) and time to successful weaning. Secondary outcomes included number of weaning cycles and 1-year mortality. Metabolic/nutritional safety parameters were recorded. Analyses followed the intention-to-treat principle, with additional per-protocol analyses. Time-to-event outcomes were evaluated using Kaplan–Meier estimates and Cox proportional hazards models.
Results
Among 211 randomized patients (111 IC; 100 control), successful weaning was achieved in 58% of the IC arm and 50% of the control arm (absolute difference 8%; p = 0.33). Median time to successful weaning was 27 vs. 31 days, respectively (p = 0.58). One-year mortality was 48% in the IC arm and 56% in the control arm (absolute difference of -8%, 95%CI: -21.2%, 5.5%; p = 0.29). In a post hoc analysis of the per-protocol cohort (n = 198), a smaller difference between measured energy expenditure and delivered energy intake (energy-gap) was independently associated with a higher likelihood of successful weaning (HR = 1.02 per 50-kcal/day reduction in the energy-gap; 95%CI: 1.00, 1.05; p = 0.02). IC-guided feeding resulted in closer alignment between measured energy expenditure and delivered energy intake and reduced the tendency toward overfeeding observed with predictive-equation–based prescriptions. Successful weaning was independently associated with lower one-year mortality (HR = 0.54; 95%CI: 0.36–0.82), adjusting for study arm.
Conclusion
The observed rates of successful weaning and one-year survival were numerically higher, and median time to liberation was shorter, with IC-guided nutrition than with predictive-equation-guided feeding, although between-group differences were not statistically significant. Closer alignment between measured energy expenditure and delivered energy intake was independently associated with successful weaning. These findings support further evaluation of energy delivery guided by measured energy expenditure in adequately powered multicenter trials.
Trial registration
ClinicalTrials.gov identifier NCT04825717. Registered March 15, 2021.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s13054-026-06272-7.
Introduction
Prolonged mechanical ventilation (PMV) is a growing clinical challenge and is associated with substantial morbidity, impaired long-term functional outcomes, and high mortality [1–3]. Weaning success among PMV patients varies widely, with reported rates ranging 10–78% across different populations and clinical settings [4, 5]. Even in structured programs, the weaning process may require weeks and is associated with considerable healthcare utilization [6].
Nutritional support is a crucial component of care for mechanically ventilated patients. Underfeeding can exacerbate muscle catabolism, impair respiratory muscle performance, prolong ventilator dependence, and worsen outcomes [7, 8]. Conversely, overfeeding increases carbon dioxide production, elevates ventilatory demand, potentially burdening an already compromised cardiopulmonary system [7, 9]. Accurate determination of energy requirements is therefore essential during the weaning process.
In clinical practice, energy needs are often estimated using predictive equations such as the Schofield or Harris-Benedict–based formulas [10]. However, these equations frequently misestimate metabolic requirements in critically ill or chronically ventilated patients, leading to underfeeding or overfeeding [10, 11]. Indirect calorimetry (IC), which measures oxygen consumption and carbon dioxide production to calculate resting energy expenditure (REE), is considered the gold standard for determining energy needs [12, 13]. Systematic reviews and expert guidelines highlight the limitations of predictive equations and recommend IC when available [10, 12, 14, 15]. In acute ICU populations, IC-guided feeding has been associated with improved nutritional accuracy and better alignment with metabolic demands [13, 16].
Despite these advances, evidence guiding nutritional strategies during the prolonged weaning phase remains limited, particularly in PMV patients treated in rehabilitation settings. No randomized controlled trials have specifically evaluated whether IC-guided nutrition support improves weaning success or survival in this population. Given the increasing burden of chronic critical illness and the importance of nutrition in respiratory muscle function and ventilator liberation [17, 18], such evaluation is warranted. Patients undergoing PMV represent a particularly challenging population because they frequently exhibit prolonged catabolism, sarcopenia, respiratory muscle weakness, recurrent infections, and highly variable metabolic demands. Unlike acute ICU patients, many have transitioned to a phase of chronic critical illness in which the goal of care is no longer stabilization of acute organ failure but restoration of physiological reserve sufficient to permit ventilator liberation. Nutritional management during this phase may therefore have direct implications for respiratory muscle performance, rehabilitation potential, duration of hospitalization, and long-term outcomes. Despite the central role of nutrition in these processes, evidence guiding energy prescription during prolonged weaning remains remarkably limited.
We conducted a double-blind, randomized controlled trial to compare IC-guided nutrition with predictive-equation-guided feeding in PMV patients undergoing structured weaning. We hypothesized that IC-guided feeding would increase weaning success and shorten time to liberation from mechanical ventilation.
Methods
Study design and setting
This study was a prospective, double-blind, randomized controlled trial conducted at the Reuth Tel-Aviv Rehabilitation Hospital between November 2019 and April 2023. The hospital is a national referral center for patients requiring PMV and structured weaning. The full methodological framework has been previously described [19].
Participants
Eligibility criteria
Adults (≥ 18 years) receiving PMV (≥ 21 days) and requiring ventilatory support for ≥ 6 h/day were eligible for inclusion; Patients were required to have relatively stable respiratory and hemodynamic status, defined as positive end-expiratory pressure (PEEP) ≤ 8 cmH₂O; Fraction of inspired oxygen (FiO₂) ≤ 60%; Systolic blood pressure > 80 mmHg with clinically stable hemodynamics (≤ 10% variation in heart rate and blood pressure for at least 1 h prior to assessment and absence of vasopressor support); Body temperature within normal clinical limits; Heart rate 40–140 beats/min; Respiratory rate 10–24 breaths/min, and no clinical evidence of uncontrolled infection.
Exclusion criteria included metabolic acidosis (pH < 7.30); Irreversible neurodegenerative disease or severe musculoskeletal disorders precluding weaning; Active infection treated with antibiotics for < 48 h (patients could be re-evaluated once stabilized).
Because the cohort consisted of chronically ventilated tracheostomized patients undergoing prolonged weaning, more conservative respiratory stability criteria were applied than typically used in acute ICU settings, including exclusion of patients with respiratory rates > 24 breaths/min [20].
Ethical considerations
The study protocol was approved by the institutional review board of Reuth Rehabilitation Hospital (IRB #0014-18-RRH). Written informed consent was obtained from all participants or their legally authorized representatives. An independent Data Safety Committee reviewed adverse events throughout the study.
Randomization
Eligible patients were randomized in a 1:1 ratio to either the IC-guided nutrition group or the predictive-equation-guided feeding group. Randomization was performed using a computer-generated block allocation sequence with a fixed block size of 20 (WinPepi, version 3.26) prepared by an investigator not involved in patient care. Allocation was concealed and communicated only to the dietitians responsible for nutritional prescription.
Blinding
Treating physicians, nursing staff, respiratory therapists, and outcome assessors were blinded to treatment allocation. All participants underwent weekly IC measurements regardless of study arm to preserve blinding. Only the dietitians responsible for prescribing nutritional intake had access to allocation information and were not involved in clinical management or outcome assessment. Patients and families were also unaware of the assigned nutritional strategy.
Study interventions
IC was performed weekly using a metabolic cart to determine measured energy expenditure (mEE), which served as the energy target for the IC-guided nutrition arm. Measurements were conducted under standardized resting conditions by trained personnel. Detailed measurement procedures are provided in Supplementary Methods S1.
Body composition was assessed using bioimpedance at baseline and study completion for descriptive safety monitoring. Detailed measurement procedures are provided in the Supplementary Methods S2.
Nutritional prescription
Before randomization, all participants received nutritional management according to routine clinical practice based on predictive equations. Indirect calorimetry was performed exclusively for the study and was not part of routine clinical care; therefore, IC-guided nutritional support was initiated only after randomization in participants allocated to the intervention group. In the intervention arm, energy intake was prescribed according to mEE obtained by IC, targeting isocaloric feeding with weekly adjustment [10, 14, 21, 22]. In the control arm, energy requirements were estimated using the Schofield predictive equation based on age, sex, and body weight [10, 21]. Protein intake followed institutional PMV guidelines (1.2–1.5 g/kg/day) and was managed similarly in both arms because the intervention addressed energy delivery only. Enteral nutrition was the primary feeding modality. Participants receiving oral nutrition remained eligible for the study provided that their dietary intake could be accurately quantified. Participants whose nutritional intake could not be reliably measured (e.g., because of unrecorded food brought by family members) were withdrawn from the study but remained included in the intention-to-treat analysis (n = 1). Detailed nutritional protocols are described in Supplementary Methods S3.
Safety monitoring
Safety monitoring included regular biochemical, hematologic, and nutritional markers. Detailed baseline and end-of-study laboratory parameters are presented in Supplementary Table S1.
Weaning protocol
Ventilator weaning was conducted using a standardized spontaneous breathing trial (SBT)-based protocol applied uniformly to both study arms and consistent with international consensus recommendations for ventilator weaning [23]. Clinical teams responsible for ventilatory management were blinded to treatment allocation. Patients who met predefined respiratory and hemodynamic stability criteria underwent an SBT to assess tolerance. Patients tolerating the trial entered a structured stepwise weaning program, in which spontaneous breathing periods were progressively extended according to clinical stability.
Weaning success was defined as sustained spontaneous breathing for ≥ 7 consecutive days, either as continuous spontaneous breathing or as ≥ 18 h/day of spontaneous breathing with ≤ 6 h/day of CPAP delivered through the tracheostomy as part of the unit’s standardized weaning protocol.
The complete weaning algorithm, including SBT procedures, failure criteria, and the progressive disconnection protocol, is described in Supplementary Methods S4, S5, S6 and Supplementary Fig. S1.
Criteria for investigator-initiated withdrawal
Patients were withdrawn from the intervention if any of the following occurred: Violation of the nutritional protocol (receipt of unrecorded food preventing accurate quantification of caloric intake); Medical deterioration requiring acute hospital transfer > 7 days; Excessive or persistent weight loss posing clinical risk; Any condition judged by clinicians to compromise safety.
Withdrawn participants remained in the ITT analyses.
Primary outcomes
Weaning success: Achieving ≥ 7 consecutive days of spontaneous breathing, defined above.
Time to successful weaning: Days from randomization to achieving weaning success, with censoring at 120 days.
Secondary outcomes
Number of weaning cycles required (up to 5); Three-month and one-year all-cause mortality, analyzed by ITT and stratified by weaning status; Safety outcomes, including metabolic markers, anthropometry, phase angle, and adverse events; Mean daily energy-gap (i.e., differences between IC mEE and delivered energy intake).
Protocol violations
No major changes were made to the study design, eligibility criteria, interventions, or primary outcomes after trial commencement. Temporary recruitment suspension during the COVID-19 pandemic, occasional interruption of calorimetry measurements because of equipment malfunction (n = 2), and withdrawal of participants whose caloric intake could not be reliably quantified (n = 1) were handled according to predefined study procedures and are presented in Fig. 1.
Fig. 1.
Study CONSORT diagram. aWithdrawn - calorimeter unavailable (n = 2). bWithdrawn - nutritional protocol violation: unrecorded oral intake (n = 1)
Statistical analysis
Sample size calculations assumed a 15% absolute difference in weaning success (30% vs. 45%), requiring 326 patients (α = 0.05, power = 80%) including 10% anticipated dropout. Analyses followed the intention-to-treat principle with additional per-protocol analyses.
Time-to-event outcomes were evaluated using Kaplan–Meier estimates and Cox proportional hazards models. Participants who died before achieving successful weaning were censored at the time of death in the time-to-event analyses. Weekly caloric delivery relative to mEE (energy-gap) was included as a time-dependent covariate. The energy-gap analysis was exploratory and performed post hoc. The mean daily energy-gap was calculated as the average of the weekly differences between measured energy expenditure and delivered caloric intake across follow-up. Cox proportional hazards models were adjusted for age, sex, and randomized treatment assignment. The per-protocol analysis included participants with sufficient indirect calorimetry and caloric intake data to assess adherence to the assigned nutritional intervention; participants with incomplete calorimetry or nutritional intake data were excluded from this analysis.
Continuous variables were compared using Kruskal–Wallis tests and categorical variables using χ² or Fisher’s exact tests. Proportional hazards assumptions were verified using Schoenfeld residuals.
Analyses were performed using R (version 3.6.2) and SAS (version 9.4).
The trial was conducted and reported in accordance with CONSORT guidelines for randomized clinical trials.
Results
Study Population: Between November 2019 and April 2023, 211 eligible patients were enrolled and randomized: 111 to the IC-guided arm and 100 to the predictive-equation-guided feeding control arm (Fig. 1). Recruitment was slower than anticipated, primarily because of the COVID-19 pandemic, and the trial concluded prior to recruiting the 326 participants specified in the original sample size calculation. All randomized participants were included in the intention-to-treat (ITT) analysis.
Baseline demographic and clinical characteristics were comparable between groups (Table 1). Mean age was 67 ± 14 vs. 68 ± 14 years and 46% vs. 39% were female in the IC and control arms, respectively. The duration of mechanical ventilation before enrolment exceeded two months in both groups.
Table 1.
Baseline characteristics of the cohort according to study arm
| Baseline characteristics | Study arm | ||
|---|---|---|---|
| Intervention n = 111 |
Control n = 100 |
P | |
| Age, years, mean ± SD | 67.4 ± 13.8 | 68.2 ± 14.4 | 0.68 |
| Female, n (%) | 46 (41) | 39 (39) | 0.83 |
| Comorbidities, n (%) | |||
| Hypertension | 72 (65) | 56 (56) | 0.24 |
| Diabetes | 54 (49) | 47 (47) | 0.87 |
| COPD | 19 (17) | 19 (19) | 0.87 |
| CHF | 13 (12) | 13 (13) | 0.94 |
| State of consciousness, n (%) | 0.44 | ||
| Full conscious | 28 (25) | 25 (25) | |
| Minimally conscious state | 66 (60) | 53 (53) | |
| Vegetative state | 17 (15) | 22 (22) | |
| Ventilation method, n (%) | 0.37 | ||
| Volume-SIMVa | 32 (29) | 34 (34) | |
| PSV-SPONTb | 47 (43) | 46 (46) | |
| Pressure-SIMV | 31 (28) | 20 (20) | |
| APACHE2c index, median [IQR] | 17 [14, 22] | 19 [15, 24] | 0.15 |
| FIMd, median [IQR] | 23 [18, 31] | 21 [18, 34] | 0.42 |
| Ventilation duration (days) at study entry, median [IQR] | 47 [32, 73] | 40 [29, 63] | 0.15 |
| Mean ± SD | 66.9 ± 76.7 | 60.6 ± 81.8 | 0.57 |
aSIMV- Synchronized Intermittent Mandatory Ventilation, bPSV -SPONT- Pressure Support Ventilation - Spontaneous breathing on ventilation; cApache2- Acute Physiology and Chronic Health Evaluation score, dFIM- Functional Independence Measure [24]
Mean IC mEE was similar between groups (1482 ± 407 vs. 1415 ± 377 kcal/day; data not shown). Nevertheless, participants managed using the Schofield predictive equation consistently received more energy than those in the IC-guided arm (1733 ± 231 vs. 1580 ± 326 kcal/day), resulting in a persistent positive energy-gap and lower rates of achieving the predefined energy target (Fig. 2, Supplementary Tables S2 and S10). These findings indicate that predictive equations overestimated energy requirements in this cohort.
Fig. 2.
The time dependent difference between Indirect Calorimeter-measured energy expenditure (mEE) and delivered energy intake (energy-gap) over the study weeks according to intervention (left) and control (right) arm. Positive values indicate caloric delivery exceeding measured energy expenditure
Primary outcomes (Table 2)
Table 2.
Study outcomes (ITT approach) according to study arm
| Study arm | |||
|---|---|---|---|
| Intervention n (%) |
Control n (%) |
p | |
| N | 111 | 100 | |
| Successful weaning | 64 (58) | 50 (50) | 0.33 |
| Time to weaning of successfully weaned patients, days | |||
| Median [IQR] | 27 [15, 46] | 31 [20, 44] | 0.58 |
| Mean ± SD | 34.5 ± 23.9 | 35.5 ± 22.8 | 0.82 |
| Number of weaning cycles required for successful weaning | 0.56 | ||
| 1 | 49 (77) | 35 (70) | |
| 2 | 11 (17) | 10 (20) | |
| 3–5 | 4 (6) | 5 (10) | |
| Number of cycles among patients not weaneda | 0.88 | ||
| 1 | 26 (55) | 31 (62) | |
| 2 | 6 (13) | 7 (14) | |
| 3–5 | 15 (32) | 12 (24) | |
| Time in follow-up, days | |||
| Median [IQR] | 25 [15, 46] | 31 [14, 45] | 0.73 |
| Mean ± SD | 31 ± 22 | 34 ± 26 | 0.52 |
| Three-month mortality | 30 (27) | 28 (28) | 0.87 |
| One-year mortality | 53 (48) | 56 (56) | 0.29 |
aReasons for discontinuation of weaning attempts are presented in Supplementary Table S6
Weaning success
Overall, 58% of participants in the IC arm achieved successful weaning compared with 50% in the control arm, corresponding to an absolute difference of 8% points (95%CI: -6 to 21% points; p = 0.33).
Time to successful weaning
Median time to successful weaning among the 118 patients who achieved liberation from ventilation was 27 days [IQR 15–46] in the IC group and 31 days [IQR 20–44] in the control group (p = 0.58). Kaplan–Meier estimates are presented in (Supplementary Figure S2).
Secondary outcomes
Number of Weaning Cycles: Among participants who were successfully liberated from mechanical ventilation, the number of required weaning cycles did not differ significantly between study arms (p = 0.56). The proportion of IC-arm patients weaned after a single cycle was 77% and 70% in the control arm, corresponding to an absolute difference of 7%.
Neurological status distribution among successfully weaned patients is shown in Supplementary Table S3.
Three-month and one-year mortality
Three-month mortality was similar between the study groups. Within three months of randomization, 30 of 111 participants (27.0%) in the IC-guided nutrition group and 28 of 100 participants (28.0%) in the predictive-equation-guided feeding group had died (absolute risk difference − 1.0%, 95% CI − 13.0% to 11.0%; P = 0.87). One-year all-cause mortality was 48% in the IC arm and 56% in the control arm, corresponding to an absolute difference of -8% points (95%CI: -21.2%, 5.5%; p = 0.29). Survival curves by study arm and multivariable Cox regression analyses of 1-year mortality of the total cohort are presented in Supplementary Figure S3 and Table S4. Successful weaning was independently associated with lower 1-year mortality (HR = 0.54, 95%CI: 0.36–0.82), adjusting for age, sex, and study arm.
Predictors of weaning success
In the per-protocol cohort (n = 198), Cox proportional hazards analysis identified the mean daily energy-gap as an independent predictor of successful weaning (Table 3). Each 50-kcal decrease in the energy-gap was associated with a higher likelihood of weaning success (HR = 1.024; 95%CI: 1.003–1.047; p = 0.024). Younger age also showed a graded association with weaning success; patients < 50 years were more than twice as likely to be successfully weaned compared with those ≥ 80 years (HR = 2.3, 95%CI: 1.04–5.07). Sex and treatment arm were not independent predictors.
Table 3.
Cox regression model for time to weaning success, adjusted for age, sex, treatment arm, and mean daily energy-gap (n = 198 patients - per protocol cohort)
| Variable | Reference | HR (95% CI) | p |
|---|---|---|---|
| Age category, years | 80+ | ||
| < 50 | 2.297 (1.041, 5.067) | 0.039 | |
| 50–69 | 1.993 (1.051, 3.777) | 0.034 | |
| 70–79 | 1.889 (0.984, 3.627) | 0.055 | |
| Sex, male | Female | 1.023 (0.689, 1.521) | 0.908 |
| Arm, intervention | Control | 1.017 (0.679, 1.523) | 0.935 |
| Energy-gapa | 50 Calories difference decrements | 1.024 (1.003, 1.047) | 0.024 |
aEnergy-gap – Mean daily difference between calorimetry-measured energy expenditure (mEE) and delivered energy intake
Intervention protocol reliability: energy-gap distribution
Post-hoc analysis of weekly energy-gaps demonstrated narrower distributions in the IC arm and wider deviations in the control arm, including frequent overfeeding, consistent with the known inaccuracy of predictive equations (Fig. 2). IC-guided feeding more frequently achieved the predefined energy target (± 50 kcal/day) throughout follow-up than predictive-equation–guided feeding (Supplementary Table S10), confirming improved alignment between delivered energy intake and measured energy expenditure.
Adverse events and safety
Three deaths occurred during the active weaning phase (one IC, two control), none related to the intervention. Recurrent infections were the most frequent adverse events (15 – IC-arm, 11- control). Detailed list of AEs/SAEs is provided in Supplementary Table S5.
Among patients who were not successfully weaned, cessation of weaning attempts most commonly followed infection, respiratory failure events, neurological deterioration, or repeated unsuccessful weaning attempts. Event distribution was similar between groups. Detailed reasons for censoring among participants who did not achieve ventilator liberation are provided in Supplementary Table S6.
Clinical biochemistry, hematologic indices, inflammatory markers, body mass index, fat-free mass, and phase angle trajectories were comparable between groups (Supplementary Tables S1, S2, S7). No safety signals associated with IC-guided feeding were identified.
Subgroup analysis
Exploratory subgroup analyses within the per-protocol cohort demonstrated broadly consistent associations between study variables and weaning outcomes across clinical subgroups. Detailed results are provided in Supplementary Table S8.
Discussion
In this randomized controlled trial of patients undergoing weaning from prolonged mechanical ventilation, the observed rates of successful weaning and one-year survival were higher and median time to ventilator liberation was shorter in the IC-guided group than in the Schofield predictive-equation guided group. The between-group differences were not statistically significant, and the trial did not reach its planned sample size. In the per-protocol cohort, closer alignment between mEE and delivered energy intake (energy-gap) emerged as an independent predictor of successful weaning. Together, these findings identify metabolic alignment as a potentially relevant treatment target while leaving the comparative clinical effectiveness of IC-guided feeding uncertain. The randomized comparisons showed an 8-percentage-point higher weaning rate, a 4-day shorter median time to liberation, and an 8-percentage-point lower one-year mortality in the IC group. These estimates were imprecise and compatible with both no treatment effect and a clinically relevant effect. Recruitment was curtailed by the COVID-19 pandemic, and the final sample of 211 participants was below the planned 326. Moreover, the control-group weaning rate was substantially higher than anticipated in the sample-size calculation. The trial therefore lacked sufficient precision to determine whether the observed differences reflect a true treatment effect.
PMV is characterized by prolonged catabolism, skeletal muscle wasting, and ventilator-induced diaphragm dysfunction, factors that may impair ventilator liberation and contribute to adverse outcomes [25]. Both underfeeding and overfeeding can further compromise respiratory function through muscle catabolism or increasing carbon dioxide production and ventilatory demand [7, 9]. Accurate estimation of energy requirements is therefore central to nutritional management during weaning, and IC remains the gold standard for determining energy needs in critically ill patients [10, 14, 15].
In the present study, predictive-equation-guided feeding consistently resulted in higher caloric delivery than measured metabolic demand, consistent with previous reports demonstrating the limited accuracy of predictive equations in critically ill patients [10, 11, 26]. Overfeeding has been associated with increased carbon dioxide production, worsened respiratory mechanics, and delayed ventilator liberation [7, 9]. Accordingly, IC-guided nutrition achieved closer alignment between measured energy expenditure and delivered energy intake, reducing the tendency toward overfeeding despite active dietitian monitoring in both groups. These observations provide a physiological rationale for the association between closer metabolic alignment and successful ventilator liberation observed in our study.
Together, the weekly energy-gap distributions (Fig. 2), the higher proportion of patients achieving the predefined energy target (Supplementary Table S10), and the independent association between a smaller energy-gap and successful weaning observed in the post hoc per-protocol analysis support the hypothesis that closer alignment between delivered energy intake and mEE, rather than the specific method used to estimate mEE, may be the biologically relevant determinant of ventilator liberation. Because this analysis was not prespecified, it should be regarded as hypothesis-generating and requires confirmation in future studies.
These findings align with prior literature demonstrating that predictive equations frequently misestimate metabolic requirements in critically ill patients [10, 11]. Systematic reviews and clinical guidelines from ESPEN and ASPEN therefore recommend IC when available to guide nutritional therapy in critical illness [10, 21, 22]. However, most randomized studies evaluating calorimetry-guided nutrition have been conducted in acute ICU populations, where results regarding clinical outcomes have been heterogeneous [12, 13, 16, 27]. Patients undergoing PMV represent a distinct clinical population characterized by chronic critical illness, prolonged catabolism, and variable metabolic demands. Evidence guiding nutritional strategies in this phase of care remains limited.
Body composition measurements obtained through bioimpedance analysis showed stable trajectories of fat-free mass and phase angle across both study arms, supporting the metabolic safety of IC-guided feeding (Supplementary Table S7). Phase angle has been proposed as a surrogate marker of cellular integrity and nutritional status and has been associated with outcomes in critically ill and surgical populations [28–30]. However, these measurements do not directly reflect respiratory muscle strength or diaphragm function. Objective assessments such as maximal inspiratory pressure or diaphragm ultrasound were not collected in the present study, limiting mechanistic interpretation of the relationship between nutritional support and respiratory muscle performance.
Successful weaning from mechanical ventilation was strongly associated with improved one-year survival, consistent with prior studies demonstrating the prognostic importance of ventilator liberation in chronically ventilated patients [31]. Patients who achieved successful weaning had approximately half the mortality risk of those who remained ventilator-dependent. These findings underscore the importance of identifying modifiable determinants of ventilator liberation in this vulnerable population. Beyond survival, successful ventilator liberation has important implications for long-term care pathways. In patients with severe neurological impairment, liberation from mechanical ventilation may permit transfer from specialized ventilator units to lower-acuity care settings, reduce exposure to ventilator-associated complications, and allow patients to follow the natural course of their underlying disease rather than remaining dependent on long-term ventilatory support.
Exploratory subgroup analyses showed generally consistent patterns of weaning outcomes across clinical subgroups (Supplementary Table S8), although these analyses were not powered for definitive inference.
The study population reflects the clinical context of PMV units in Israel, where patients are typically hemodynamically stable but remain ventilator-dependent, often with severe neurological impairment. Structured weaning attempts are therefore frequently continued even in patients with poor neurological prognosis (Supplementary Table S3). Although ventilator liberation does not alter the neurological condition of these patients, successful weaning may permit transfer from specialized ventilator units to lower-acuity supportive care settings, substantially reducing healthcare resource utilization. In addition, liberation from mechanical ventilation reduces exposure to complications associated with long-term ventilator dependence, including recurrent respiratory infections, airway complications, and prolonged immobilization. Importantly, ventilator liberation allows patients to follow the natural course of their underlying illness rather than remaining dependent on chronic life-sustaining ventilatory support. This setting therefore provides a unique opportunity to evaluate modifiable factors influencing ventilator liberation, including nutritional management. Chronically ventilated patients undergoing prolonged weaning remain underrepresented in randomized trials, and the present study contributes prospective evidence in this population.
This trial has several strengths. It employed a randomized double-blind design and a standardized weaning protocol, reducing variability in clinical decision-making. Indirect calorimetry was performed weekly in all participants irrespective of treatment allocation, allowing consistent physiological assessment across study arms. Safety monitoring included biochemical, anthropometric, and body composition parameters, with no evidence of adverse effects associated with IC-guided feeding. The study also introduces the concept of the energy-gap as a clinically measurable indicator of metabolic alignment and demonstrates its association with successful ventilator liberation.
Several limitations should be acknowledged. First, the trial did not reach its planned sample size, reducing statistical power to detect moderate treatment effects. Recruitment was substantially affected by the COVID-19 pandemic, which prolonged enrolment and limited the achievable cohort size. Second, calorimetry measurements were performed weekly rather than daily; more frequent measurement might have improved the alignment between metabolic needs and delivered nutrition. In addition, nutritional intake data before study enrolment was not available. However, indirect calorimetry was not part of routine clinical care; therefore, all participants received standard predictive equation–based nutritional management before randomization. Third, the control group was managed using the Schofield predictive equation; therefore, the findings may not be generalizable to other predictive equations. Fourth, respiratory muscle function was not directly assessed. Fifth, we did not collect detailed data on rehabilitation intensity or other potentially modifiable factors, such as mobilization, which may also influence ventilator liberation and should be considered in future studies. Sixth, energy expenditure was measured using the Quark RMR indirect calorimeter, which has been widely validated but has largely been superseded by newer-generation devices. Nevertheless, the same device and standardized measurement protocol were used throughout the trial, making differential measurement bias between study arms unlikely. Finally, the study was conducted at a single center, which may limit generalizability, although the standardized clinical environment allowed consistent implementation of the intervention.
The trial estimates were compatible with both no effect and potentially relevant differences in weaning and survival outcomes. The independent association between closer metabolic alignment and successful weaning provides a rationale for further investigation but does not establish that IC-guided feeding improves clinical outcomes. Adequately powered multicenter trials should evaluate whether reducing discrepancies between measured requirements and delivered energy improves ventilator liberation and long-term outcomes.
Conclusions
In patients undergoing prolonged mechanical ventilation weaning, observed weaning success and one-year survival were numerically higher and time to liberation was shorter with IC-guided than with the Schofield predictive-equation–guided feeding, but the trial lacked the precision to establish a between-group treatment effect. Closer alignment between measured energy expenditure and delivered intake was independently associated with successful weaning and warrants evaluation as a modifiable nutritional target in adequately powered multicenter trials.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
We thank the clinical teams of the Prolonged Mechanical Ventilation Unit at Reuth Rehabilitation Medical Center for their dedicated care of participating patients and their invaluable support throughout the trial. We also thank the dietitians and research staff for their meticulous work in conducting calorimetry and bioimpedance measurements and maintaining study fidelity.
Author contributions
R.D. M.E. M.F. and P.S. wrote the main manuscript text; N.A. and M.B. analyzed the data and N.A. contributed to drafting the statistical analysis Methods section; M.E., R.D., M.B., M.G.K. and P.S. contributed to the conception and design of the work; M.F., M.E., L.B. and T.M. contributed to the acquisition of data; R.D., M.E., P.S. and N.A. contributed to the interpretation of the findings. All authors reviewed the manuscript and approved the submission.
Funding
This study was supported by the Israel General Custodian (Grant #20200311) and by an internal research grant from the Reuth Rehabilitation Hospital. The funding bodies had no role in study design, data collection, analysis, interpretation, or manuscript preparation.
Data availability
The data that support the findings of this study are available from Reuth Rehabilitation Hospital, but restrictions apply to the availability of these data, which were used under patient’s privacy restrictions per Reuth’s IRB committee decision and so are not publicly available. The data are, however, available upon reasonable request and with the permission of the Reuth Rehabilitation Hospital.
Declarations
Ethics approval and consent to participate
The study was approved by the Institutional Review Board of Reuth Tel-Aviv Rehabilitation Hospital (IRB #0014-18-RRH). Written informed consent was obtained from all participants or their legally authorized representatives.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Mila Fradkin and Maya Elyashiv are joint first authors.
Pierre Singer and Rachel Dankner are joint last authors.
References
- 1.McConville JF, Kress JP. Weaning patients from the ventilator. N Engl J Med. 2012;367:2233–9. [DOI] [PubMed] [Google Scholar]
- 2.Fradkin M, Elyashiv M, Camel A, Agay N, Brik M, Singer P, et al. A historical cohort study on predictors for successful weaning from prolonged mechanical ventilation and up to 3-year survival follow-up in a rehabilitation center. Respir Med. 2024;227:107636. [DOI] [PubMed] [Google Scholar]
- 3.Damuth E, Mitchell JA, Bartock JL, Roberts BW, Trzeciak S. Long-term survival of critically ill patients treated with prolonged mechanical ventilation: a systematic review and meta-analysis. Lancet Respir Med. 2015;3(7):544–53. [DOI] [PubMed] [Google Scholar]
- 4.Shin H, Chang J, Ahn S, Kim T, Park C, Lim J, et al. Clinical factors associated with weaning failure in patients requiring prolonged mechanical ventilation. J Thorac Dis. 2017;9(1):143–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Rose L, Messer B. Prolonged mechanical ventilation, weaning, and the role of tracheostomy. Crit Care Clin. 2024;40(2):409–27. [DOI] [PubMed] [Google Scholar]
- 6.Dolinay T, Hsu L, Maller A, Walsh BC, Szűcs A, Jerng JS, et al. Ventilator weaning in prolonged mechanical ventilation - a narrative review. J Clin Med. 2024;13(7):1909. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Pham T, Heunks LM, Bellani G, Madotto F, César Aragao I, et al. Weaning from mechanical ventilation in intensive care units across 50 countries (WEAN SAFE): a multicentre, prospective, observational cohort study. Lancet Respir Med. 2023;11(5):465–76. [DOI] [PubMed] [Google Scholar]
- 8.Van Zanten AR, De Waele E, Wischmeyer PE. Nutrition therapy and critical illness: practical guidance for the ICU, post-ICU, and long-term convalescence phases. Crit Care. 2019;23:368. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.McClave SA, Taylor BE, Martindale RG, Warren MM, Johnson DR, Braunschweig C, et al. Guidelines for the provision and assessment of nutrition support therapy in the adult critically ill patient: Society of Critical Care Medicine (SCCM) and American Society for Parenteral and Enteral Nutrition (A.S.P.E.N). JPEN J Parenter Enter Nutr. 2016;40(2):159–211. [DOI] [PubMed] [Google Scholar]
- 10.Bendavid I, Lobo DN, Barazzoni R, Cederholm T, Coëffier M, de van der Schueren M, et al. The centenary of the Harris-Benedict equations: how to assess energy requirements best? Recommendations from the ESPEN expert group. Clin Nutr. 2021;40(3):690–701. [DOI] [PubMed] [Google Scholar]
- 11.Pertzov B, Bar-Yoseph H, Mendel Y, Bendavid I, Kagan I, Glass YD, et al. The effect of indirect calorimetry guided isocaloric nutrition on mortality in critically ill patients - a systematic review and meta-analysis. Eur J Clin Nutr. 2022;76(1):5–15. [DOI] [PubMed] [Google Scholar]
- 12.Delsoglio M, Achamrah N, Berger MM, Pichard C. Indirect calorimetry in clinical practice. J Clin Med. 2019;8(9):1387. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Oshima T, Berger MM, De Waele E, Guttormsen AB, Heidegger CP, Hiesmayr M, et al. Energy expenditure and indirect calorimetry in critical illness and convalescence: current evidence and practical considerations. J Intensive Care. 2021;9(1):8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Duan JY, Zheng WH, Zhou H, et al. Energy delivery guided by indirect calorimetry in critically ill patients: A systematic review and meta-analysis. Crit Care. 2021;25:88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Tatucu-Babet OA, Ridley EJ, Tierney AC. Is energy delivery guided by indirect calorimetry associated with improved clinical outcomes in the ICU? A systematic review and meta-analysis. Nutr Clin Pract. 2020;35(3):527–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Singer P, Anbar R, Cohen J, Shapiro H, Shalita-Chesner M, Lev S, et al. The tight calorie control study (TICACOS): a prospective, randomized, controlled pilot study of nutritional support in critically ill patients. Intensive Care Med. 2011;37(4):601–9. [DOI] [PubMed] [Google Scholar]
- 17.Kahn JM, Le T, Angus DC, Cox CE, Hough CL, White DB, et al. The epidemiology of chronic critical illness in the United States. Crit Care Med. 2015;43(2):282–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kahn JM, Davis BS, Le TQ, Yabes JG, Chang CH, Angus DC, et al. Variation in mortality rates after admission to long-term acute care hospitals for ventilator weaning. J Crit Care. 2018;46:6–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Fradkin M, Elyashiv M, Brik M, Kait MG, Albukrek D, Singer P, et al. The efficacy of calorimetry-based nutritional support versus recommended dietary intake-based nutritional support in the successful weaning of chronically ventilated patients: a study protocol for a double-blind randomized controlled trial. Clin Nutr ESPEN. 2023;54:94–7. [DOI] [PubMed] [Google Scholar]
- 20.Janssen ML, Endeman H, Yang Z, Elderman JH, Goeijenbier M, Dongelmans T, et al. Trach-Wean study group. Dyspnea is related to clinical outcomes in patients weaning from invasive mechanical ventilation with tracheostomy: a multicenter prospective study. Crit Care. 2026;30(1):16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Singer P, Blaser AR, Berger MM, Alhazzani W, Calder PC, Casaer MP, et al. ESPEN guideline on clinical nutrition in the intensive care unit. Clin Nutr. 2019;38(1):48–79. [DOI] [PubMed] [Google Scholar]
- 22.Singer P, Blaser AR, Berger MM, et al. ESPEN practical and partially revised guideline: Clinical nutrition in the intensive care unit. Clin Nutr. 2023;42(9):1671–89. [DOI] [PubMed] [Google Scholar]
- 23.Boles JM, Bion J, Connors A, Herridge M, Marsh B, Melot C, et al. Weaning from mechanical ventilation. Eur Respir J. 2007;29(5):1033–56. [DOI] [PubMed] [Google Scholar]
- 24.Keith RA, Granger CV, Hamilton BB, Sherwin FS. The Functional Independence Measure: a new tool for rehabilitation. In: Eisenberg MG, Grzesiak RC, editors. Advances in Clinical Rehabilitation. Volume 1. New York: Springer Publishing Company; 1987. pp. 6–18. [PubMed] [Google Scholar]
- 25.Powers SK, Wiggs MP, Sollanek KJ, Smuder AJ. Ventilator-induced diaphragm dysfunction: cause and effect. Am J Physiol Regul Integr Comp Physiol. 2013;305(5):R464–77. [DOI] [PubMed] [Google Scholar]
- 26.Zusman O, Theilla M, Cohen J, Kagan I, Bendavid I, Singer P. Resting energy expenditure, calorie and protein consumption in critically ill patients: a retrospective cohort study. Crit Care. 2016;20(1):367. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Singer P, De Waele E, Sanchez C, et al. TICACOS international: A multi-center, randomized, prospective controlled study comparing tight calorie control versus liberal calorie administration study. Clin Nutr. 2021;40(2):380–7. [DOI] [PubMed] [Google Scholar]
- 28.Powers SK, Veenstra JA, Koopmans M, Boerma CE. Phase angle is related to outcome after ICU admission; an observational study. Clin Nutr ESPEN. 2018;23:61–6. [DOI] [PubMed] [Google Scholar]
- 29.Stellingwerf F, Beumeler LFE, Rijnhart-de Jong H, Boerma EC, Buter H. The predictive value of phase angle on long-term outcome after ICU admission. Clin Nutr. 2022;41(6):1256–9. [DOI] [PubMed] [Google Scholar]
- 30.Norman K, Stobäus N, Pirlich M, Bosy-Westphal A. Bioelectrical phase angle and impedance vector analysis—clinical relevance and applicability of impedance parameters. Clin Nutr. 2012;31:854–61. [DOI] [PubMed] [Google Scholar]
- 31.Scheinhorn DJ, Hassenpflug MS, Votto JJ, Chao DC, Epstein SK, Doig GS, et al. Ventilator-dependent survivors of catastrophic illness transferred to 23 long-term care hospitals for weaning from prolonged mechanical ventilation. Chest. 2007;131(1):76–84. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from Reuth Rehabilitation Hospital, but restrictions apply to the availability of these data, which were used under patient’s privacy restrictions per Reuth’s IRB committee decision and so are not publicly available. The data are, however, available upon reasonable request and with the permission of the Reuth Rehabilitation Hospital.