Skip Navigation
Skip to contents

ACC : Acute and Critical Care

OPEN ACCESS
SEARCH
Search
Advanced Search
mobile menu button

Articles

Page Path
HOME > Acute Crit Care > Volume 41(1); 2026 > Article
Original Article
Pulmonary Predictive value of resting energy expenditure and oxygen consumption of breathing for ventilator weaning in surgical intensive care patients in Viet Nam: a cross-sectional study
Acute and Critical Care 2026;41(1):97-106.
DOI: https://doi.org/10.4266/acc.001150
Published online: February 27, 2026

1Center of Anesthesiology, Critical Care and Pain Management, Hanoi Medical University Hospital, Ha Noi, Viet Nam

2Department of Anesthesia and Critical Care, Hanoi Medical University, Ha Noi, Viet Nam

3Department of Anesthesia and Critical Care, Saint Paul General Hospital, Ha Noi, Viet Nam

4High Service Department of Anesthesiology and ICU, Hanoi Obstetrics and Gynecology Hospital, Ha Noi, Viet Nam

Corresponding author: Tran Viet Duc Center of Anesthesiology, Critical Care and Pain Management, Hanoi Medical University Hospital, No. 1 Ton That Tung Street, Kim Lien Ward, Ha Noi 100000, Viet Nam Tel: +84-33-531-3891 Fax: +84-243-574-6298 Email: ductran.hmu@gmail.com
• Received: April 10, 2025   • Revised: October 24, 2025   • Accepted: November 12, 2025

© 2026 The Korean Society of Critical Care Medicine

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

  • 694 Views
  • 48 Download
prev next
Full Article

Download PDF Download PDF

  • Background:
    Patients being weaned from mechanical ventilation are evaluated using portable metabolic carts to measure oxygen consumption (V˙O2) and carbon dioxide production (V˙CO2) to determine resting energy expenditure (REE) and oxygen cost of breathing (OCOB). The purpose of this study was to investigate the cutoff threshold and the predictive values of OCOB and REE variation in such patients at a surgical critical care unit.
  • Methods:
    A cross-sectional study was conducted from March 2023 to February 2024. A total of 58 patients who were at least 18 years old and receiving orotracheal ventilation for more than 48 hours met the criteria for weaning. The relationship among OCOB, variation in resting energy expenditure (∆REE) index, and weaning results was examined. Calculations were performed to determine the receiver operating characteristic (ROC) curve, area under the curve (AUC), sensitivity, specificity, and positive and negative predictive values (PPV and NPV).
  • Results:
    The overall weaning success rate was 72.4% (42/58). Compared to the failure group, the successful weaning group had significantly lower median values of ∆REE and OCOB (6.3% and 6.93% vs. 15.64% and 16.21%, P=0.002, respectively). According to ROC analysis, weaning success was predicted when ∆REE ≤8.28% (AUC, 0.768) or OCOB ≤9.1% (AUC, 0.762).
  • Conclusions:
    Both OCOB and ∆REE can be used to predict a successful outcome of ventilator weaning, with a criterion of OCOB ≤9.1% or ∆REE ≤8.28% in surgical critical care patients.
  • Key Words: energy expenditure; indirect calorimetry; oxygen consumption; surgical intensive care; ventilator weaning
Mechanical ventilation often is necessary in the intensive care unit (ICU) to support ventilation and gas exchange. However, prolonged mechanical ventilation leads to many complications, including lung injury, infection, and neuromuscular dysfunction, further prolonging mechanical ventilation time [1]. For these reasons, ventilator weaning should be performed as early as feasible. However, early extubation and weaning might result in reintubation, which poses several additional risks. Reliable predictors for evaluation of decreased need for ventilator assistance are necessary to effectively implement a weaning program. Numerous indications must be synchronized to maximize the outcomes of the two weaning stages (spontaneous breathing trial and extubation). Indirect calorimetry (IC) measures oxygen consumption (V˙O2) and carbon dioxide production (V˙CO2) to calculate resting energy expenditure (REE) and the respiratory quotient (RQ). The RQ is the proportion of carbon dioxide exhaled to oxygen inhaled (V˙CO2/V˙O2) and can be used for altering a patient's support regimen [2]. REE can be calculated using the abbreviated Weir formula: [3.9 (V̇O2)+1.1 (V̇CO2)]×1,440 (kcal/day) [2]. REE can provide accurate and individualized information on the metabolic and nutritional status of critically ill patients, especially those receiving mechanical ventilation, as well as assessment of weaning success and outcome [3-5]. The oxygen cost of breathing (OCOB) is defined as the percentage of total V˙O2 used by the muscles of respiration and reflects the load of respiratory muscles. Avoiding respiratory muscle overload during the weaning process is crucial for effective weaning from mechanical ventilation, since an excessive load often results in weaning failure [6]. Many studies have shown increased V˙O2 as a predictor of failure to wean from mechanical ventilation. Bellani et al. [7] found that the baseline V˙O2 of patients who failed a decremental pressure support (PS) experiment was greater than that of those who succeeded (mean±standard deviation: 174±44 vs. 215±53 ml/min, P<0.05). Raurich et al. [8] showed that the receiver operating characteristic curve (0.96±0.03), positive likelihood ratio (PLR) (9.5; 95% CI, 1.5–61), and negative likelihood ratio (NLR) (0.1; 95% CI, 0.01–0.4) to differentiate between successful and unsuccessful weaning were highest at a cutoff value of V˙O2 respiratory ≤10%. In surgical intensive care unit (SICU) patients, metabolic and physiological disturbances after major surgery can strongly influence both REE and OCOB. Surgical trauma induces a systemic inflammatory response and hormonal changes that increase protein catabolism, glucose utilization, and overall energy consumption [9]. At the same time, postoperative factors such as pain, residual anesthetic effects, and impaired diaphragmatic function can increase the oxygen requirement for breathing [10]. These processes can lead to an imbalance between metabolic demand and respiratory capacity, especially during the weaning phase. Monitoring REE provides an overview of the patient’s global metabolic load, while OCOB reflects the portion of energy devoted to respiratory effort. In combination, these two parameters can capture both systemic and respiratory components of readiness for spontaneous ventilation, offering a more comprehensive physiological assessment than traditional weaning indices. However, only a few studies have explored the use of IC for this purpose in SICU patients, and further data are needed.
To address the variations in the predictors of weaning from ventilator among the Vietnamese population receiving treatment at various SICUs, this study was conducted to assess the predictive values of REE and OCOB in ventilator weaning protocols in patients receiving treatment at an SICU.
The study was approved by the Ethics Committee of Hanoi Medical University (No. 326/QD-DHYHN, February 15, 2023), which waived the need for direct participant agreement because our study subjects were on mechanical ventilation. Nonetheless, all of the study participants nearest relatives provided written informed consent.
Design of the Study
A cross-sectional study was carried out from March 2023 to February 2024 at Center of Anesthesiology, Critical Care and Pain Management, Hanoi Medical University Hospital, Hanoi, Vietnam. The inclusion criteria were (1) age ≥18 years, (2) orotracheal ventilation for more than 48 hours, and (3) a condition satisfying the criteria for readiness to begin weaning without previous ventilation weaning failure. Patients were excluded if they had (1) severe brain trauma, defined as a Glasgow Coma Scale score less than 8; (2) progressive neuromuscular disease; (3) agitation or convulsions; (4) tracheostomy; (5) endotracheal cuff tube leak, defined as (VTi–VTe)/VTi >10%, where VTi is inspiratory tidal volume, and VTe is expiratory tidal volume; (6) a history of cardiovascular disease or chronic obstructive pulmonary disease; or (7) continuous renal replacement therapy and/or extracorporeal membrane oxygenation.
The calculation of sample size was based on research by Huang et al. [11]. The extubation success rate was 63%, and the sample size was determined using the cross-sectional study formula. The minimal sample size was 40, with a type I error of 0.05 and an absolute error of 15%. The ventilator weaning process was undertaken when the patient’s condition fulfilled the following criteria: improvement or reversal of the condition that required mechanical ventilation; minimal or no sedation, with Glasgow Coma Scale score ≥9; axillary temperature <38 °C; and hemoglobin ≥ 8 g/dl. All patients were required to be hemodynamically stable with no vasopressor use or only minimal doses of inotropes/vasopressors at the time of the spontaneous breathing trial, in accordance with our institutional weaning protocol. In addition, patients must demonstrate normalized cardiovascular and hemodynamic status with heart rate < 140 beats/min and systolic blood pressure 90–160 mm Hg without or with vasopressors and inotropes at minimal doses; corrected acid-base balance abnormalities and ventilation with positive end-expiratory pressure (PEEP) ≤5 to 8 cm H2O and fraction of inspired oxygen (FiO2) ≤40%; and a ratio of partial pressure of arterial oxygen (PaO2) and FiO2 ≥250 mm Hg [12,13].
Weaning refers to the process of a spontaneous breathing trial (SBT) and extubation. The SBT stage involved 60 minutes of continuous positive airway pressure (CPAP) mode with settings of PS 7 cm H2O, PEEP 5 cm H2O, and FiO2 40%. In the meantime, the following indicators of intolerance were tracked to assess SBT failure [14]: (1) respiratory failure such as tachypnea, tachycardia, hypertension, hypotension, hypoxemia or acidosis, and arrhythmia and (2) subjective indicators of agitation or distress, depression, diaphoresis, and signs of increasing effort (visible use of accessory respiratory muscles, facial sign of distress, dyspnea). During the SBT, the treating intensivist directly observed the patient for any subjective signs of weaning intolerance. These findings were documented in real time and integrated with objective physiological parameters such as respiratory rate, heart rate, blood pressure, and oxygen saturation. Patients were considered to have had a successful SBT if they showed no symptoms of intolerance. These patients were then extubated and observed for the next 48 hours. The researchers, who are trained intensivists, monitored the patients and evaluated intolerance signals; patients presenting such signs underwent reintubation or tracheostomy. The patients in the successful weaning group were those who demonstrated successful extubation along with successful SBT. The weaning failure group, on the other hand, comprised patients who underwent reintubation, SBT failure, or recommenced ventilatory support. Patients who required non-invasive ventilation (NIV) within 48 hours after extubation due to intolerance were classified as weaning failure, as this represented early extubation failure rather than late therapeutic NIV use [14,15].
Measurement and Data Collection
V˙O2, V˙CO2, and REE were measured directly from GE Carescape R860 ventilator (GE Healthcare) software version 10 SP04 and IC solution components (GE Healthcare E-sCOVX module and GE Healthcare D-Lite++ flow sensor). Additionally, the E-sCOVX module detected gas exchange on a breath-by-breath basis. Samples of gas were collected using a line attached to the flowmeter. For O2 and CO2, the analysis was conducted using paramagnetic and infrared techniques, respectively. The ventilator circuit Y-piece was directly linked to a pneumotach flowmeter to measure flow rate [16].
A steady state is crucial for accurate measurement and is often defined as a single 5-minute interval in which the average RQ varies by less than 5% and the average minute V˙O2 and V˙CO2 vary by less than 10%. The RQ should remain in the normal physiologic range (0.67–1.3) [2,17]. In a test lasting 15 to 20 minutes, stable, interpretable measurements should be acquired: one hour before the IC measurement, avoid intravenous or enteral feeding, use of bicarbonate solution, invasive procedures, and mobility and alter the dosage of any catecholamines.
The average V˙O2, energy expenditure for a stable 20-minute period at rest (REE) were calculated 5 minutes before SBT stage initiation as baseline data. These parameters were also recorded for the last 5 minutes of the trial to determine the working status of the respiratory muscles (at the end of the 60-minute SBT or at failure and subsequent stoppage of the SBT, whichever came first). The following formula was used to determine the OCOB and ∆REE, which show the rate at which each parameter increased during a trial: [(the value of the last 5 minutes of the trial–the baseline value)/the baseline value]×100%.
Clinical and demographic details of gender, Acute Physiology and Chronic Health Evaluation (APACHE) II score, length of hospital stay, length of SICU stay, and length of mechanical ventilation were documented. Upon admission to the SICU, doctors evaluated the APACHE II level. The time between the first day of intubation and ventilator support to extubation and release from ventilators was the mechanical ventilation duration. Patients were further categorized by surgical type (gastrointestinal, neurosurgical, cardiovascular, trauma) and surgical urgency (elective vs. emergency). Comorbidities including hypertension, diabetes mellitus, chronic kidney disease, and malignancy were recorded, as well as vasopressor or inotrope use at the time of the weaning trial.
Statistical Analysis
The statistical analyses were conducted using IBM SPSS Statistics version 26.0 (IBM Corp.) and Medcalc 13.0.6. Categories were shown as percentages, while continuous variables were shown as median (interquartile range [IQR]). To ensure comparability between the weaning success and failure groups, as well as to determine the exact relationships of the REE and OCOB index with weaning process outcomes, univariate analyses of the demographic and clinical factors were performed. Continuous and categorical data were compared using the Mann-Whitney U-test and the chi-square test, respectively. Fisher’s exact test was used when the expected cell count was less than 5. Receiver operating characteristic (ROC) curves were created to calculate the final area under the ROC curve (area under the curve [AUC]) values. The maximum Youden's index (sensitivity+specificity–1) was chosen as the ideal cutoff point by MedCalc, which also calculated the sensitivity and specificity after data entry. The following metrics were computed for the OCOB and ∆REE cutoff values to predict the weaning outcome: sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), PLR, and NLR. The likelihood ratio of any clinical finding, which can vary from 0 to infinity, is calculated by dividing the likelihood of the finding in patients with the disease by the probability of the same finding in patients without the disease. A test that has a high PLR and a low NLR is very accurate in diagnosing conditions. Statistical significance was indicated by a two-sided P-value less than 0.05.
The study comprised 58 patients in total (Figure 1), with an average age of 59.5 years (standard deviation, 19.3), of whom 55.2% were deemed geriatric (>60 years), and an approximately 2.9:1 male to female ratio. Among the 58 patients, 38 (65.5%) underwent elective procedures, and 20 (34.5%) experienced emergency surgery. The most frequent underlying conditions were gastrointestinal surgery (39.7%) and central nerve disease (24.1%), which was primarily caused by brain tumors. Cardiovascular and mediastinal masses were documented in 20.7% of patients, while 8.6% of the patients had diagnosis of multiple traumas. Major comorbidities were hypertension (31.0%), diabetes mellitus (19.0%), chronic kidney disease (12.1%), and malignancy (13.8%), with no significant differences between the weaning success and failure groups (all P >0.05).
In the SBT stage, intolerance symptoms were seen in 10 instances (17.2%); of the remaining 48 patients (82.8%), 42 achieved successful extubation. Thus, 16 patients (27.6%) experienced failure of ventilator weaning, whereas 42 (72.4%) demonstrated successful weaning outcomes. No death occurred during the study. The proportion of men in the weaning success group (74%) and the weaning failure group (75%) did not vary significantly. Likewise, the median values of age (P=0.093), body mass index (P=0.189), lengths of hospital stay (P=0.583) and SICU stay (P=0.082), and duration of mechanical ventilation (the interval between intubation and the first SBT; P=0.979) did not differ statistically between the two groups. Nonetheless, the APACHE II score of the successful weaning group was statistically considerably lower than that of the unsuccessful weaning group (P=0.000) (Table 1).
There were significant differences in ∆REE and OCOB values between the weaning success and failure groups. The median ∆REE was 6.3% in the weaning success group and 15.64% in the weaning failure group (P=0.002) (Table 2, Figure 2A). The median OCOB was 6.93% in the weaning success group but 16.21% in the weaning failure group (P=0.002) (Table 2, Figure 2B).
The AUC of ∆REE in the whole weaning process was 0.768 (95% CI, 0.630–0.906) (Figure 3A). From the ROC curve, the optimal cutoff point of the ∆REE index was calculated as ≤8.28% to predict weaning success with the highest predictive accuracy (sensitivity: 69%, specificity: 81.3%, positive predictive value: 90.6%, NPV: 50%, PLR: 3.7, NLR: 0.4). The AUC and ROC curves of the OCOB index in the entire process of ventilator cessation were also created. An OCOB ≤9.1% had good accuracy (sensitivity, 71%; specificity, 81.3%; positive predictive value, 90.9%; NPV, 51.7%; PLR, 3.8; NLR, 0.4), with an AUC of 0.762 (95% CI, 0.621–0.903) (Figure 3B).
The success rate of weaning from a ventilator in our study was 72.4% (95% CI, 60.6%–84.3%). This result was similar to many studies on weaning from mechanical ventilation in surgical intensive care patients, such as the research by Puetpaiboon et al. [18], with a success rate of 423/553 patients (76.5%). In addition, 73.2% of patients were successfully weaned from mechanical ventilation after gastrointestinal surgery as published by Yildirim et al. [19]. In our study, among the 16 patients who failed to wean from mechanical ventilation, 10 failed in the first SBT phase, four had to be reintubated within 24 hours after extubation, and two had to be supported by NIV due to intolerance after extubation. Our study results showed no difference in median age, body mass index, duration of mechanical ventilation before weaning, length of stay in ICU or length of hospital stay, but the APACHE II score in the failed weaning group (median, 12.0; IQR, 11.0–14.8) was significantly higher than in the successful weaning group (median, 9.0; IQR, 6.8–10.0; P<0.001). The APACHE II score is widely used to predict mortality and severity in patients treated in the ICU as well as the outcome of weaning from mechanical ventilation [20-22]. McConville and Kress [23] also showed that APACHE II ≥12 at the time of ICU admission was predictive of failed weaning. The higher APACHE II scores in the failed weaning group suggest that overall illness severity influenced the metabolic and respiratory responses. In surgical ICU patients, postoperative inflammation, pain, and transient diaphragmatic dysfunction can further amplify this metabolic stress. ∆REE and OCOB can reflect both the general severity captured by APACHE II and surgery-related physiological factors that transiently increase the work of breathing and energy demand.
In this study, surgical types and comorbidity profiles were comparable between the weaning success and failure groups (Table 1). This balanced distribution indicates that the differences observed in ∆REE and OCOB were unlikely to result from disparities in baseline illness severity or surgical complexity. Most patients had undergone elective gastrointestinal or cardiovascular operations, and the overall prevalence of major comorbidities such as diabetes, chronic kidney disease, or malignancy was low. These findings suggest that variations in metabolic and respiratory responses during weaning primarily reflected postoperative physiological changes rather than chronic disease burden.
Our study noted that the rate of increased respiratory V˙O2 (or OCOB) was higher in the group with failed weaning (6.93% vs. 16.21%, P=0.002). Miwa et al. [24] also found that OCOB was lower in the group with successful weaning (median, 15.2%; IQR, 8.6%–23.3%) than in the group with failed weaning (median, 32%; IQR, 30.2%–40.3%) (P<0.001) [24]. The cutoff predicting successful weaning according to this author was ≤30% with sensitivity 96%, specificity 82%, PPV 0.92, NPV 0.89, and an AUC of 0.919 [24]. Compared with our study, this was a noticeable difference since the AUC of OCOB for the entire ventilator weaning process was 0.762 (95% CI, 0.621–0.903) and the cutoff predicting successful ventilator weaning was ≤9.1% with sensitivity 71%, specificity 81.3%, PPV 90.9%, and NPV 51.7%. These differences might be due to the variance in study design. Miwa et al. [24] selected medical ICU patients with long-term mechanical ventilation (mean duration of mechanical ventilation was 23.2 days), and each patient underwent multiple mechanical ventilation weaning trials (mean, 10.4 times) with two selected ventilation modes, synchronized intermittent mandatory ventilation and CPAP, to evaluate OCOB for efforts to reduce ventilator support, not for the entire weaning process (including extubation). A study with a design similar to that of Miwa et al. [24] was that of Mitsuoka et al. [25], who found that the cutoff value for OCOB to predict successful reducing ventilator support was ≤15% with sensitivity 96.6%, specificity 85.7%, PPV 98.5%, and NPV 72% [25]. The cutoff point in our study was lower than that in other studies, possibly because the duration of mechanical ventilation before weaning was not as long, and more importantly, the ventilator support level changed from assist/control ventilation mode to CPAP/PS with PS 7 cm H2O instead of gradually decreasing the support level from 20 cm H2O as in most previous studies. Despite differences in study design, OCOB appears to be a useful indicator of respiratory muscle function. Muñoz et al. [26] found work of breathing to be a crucial factor in assessing respiratory muscle function. Work of breathing values between 0.8 and 1.3 J/L were considered indicative of high respiratory demand, such as in patients with airway and pulmonary abnormalities. Values between 1.3 and 1.8 J/L were deemed elevated, potentially hindering adequate spontaneous ventilation over time [26]. This finding suggests that elevated indicators of respiratory muscle strain can impair a patient's ability to tolerate spontaneous ventilation. For surgical patients, this index is a reliable predictor of successful weaning from mechanical ventilation; however, as with most new modalities, more research is necessary to determine its true effectiveness and limitations.
We also found that the variation in REE after the SBT or ∆REE between the two groups of successful weaning (median, 6.3%; IQR, 3.75%–10.08%) and failed weaning (median, 15.64%; IQR, 8.37%–30.59%) was significantly different (P=0.002). This result was compared to those of Savchenko et al., showing a smaller increase in REE in the group that successfully weaned from the ventilator (median, 16%; IQR, 12%–29%), but the difference was not significant compared to the group that failed to wean (median, 22%; IQR, 15%–31%) (P=0.570) [27]. The reason for this difference compared to our study might be due to several reasons. Savchenko et al. conducted a study on patients after craniotomy, with a duration of mechanical ventilation >7 days, a small sample size (33 patients), and before weaning from the ventilator using the PS reduction method for 1 day. This allowed the patients to acclimate to the gradual reduction of ventilator support, hindering expression of the difference in energy expenditure between patients who successfully or unsuccessfully withdrew from the ventilator. However, that study emphasized that the absence of a decrease in energy expenditure would predict successful SBT with sensitivity 44% and specificity 100% [27]. Santos and Vieira [28] also conducted weaning by gradually reducing ventilator support via T-tube and measuring the increase in energy expenditure. The results showed that EE increased by 14.43% in the group with signs of intolerance when weaning from the ventilator (P<0.001), similar to our study results. When constructing the ROC curve for ∆REE to predict the outcome of weaning from mechanical ventilation, we obtained results similar to the OCOB index when the AUC of the ∆REE index was 0.768 (95% CI, 0.630–0.906). The optimal cutoff point was ∆REE ≤8.28% which predicted successful weaning from mechanical ventilation with sensitivity 69%, specificity 81.3%, PPV 90.6%, NPV 50%, PLR of 3.7, and NLR of 0.4. Miwa et al. [24] found that a cutoff point for ∆REE of ≤20% predicted successful weaning from mechanical ventilation with sensitivity 74%, SP 70%, PPV 85%, NPV 54%, and AUC=0.807 [24]. This difference was due to the study design of Miwa et al. [24], who evaluated the outcome of SBT stage or repeated reduction of ventilator support on a small group of medical intensive care patients on long-term mechanical ventilation. Therefore, patients in Miwa et al. [24] would likely have required a large REE change (20% or greater) to show symptoms of intolerance in the SBT phase.
SICU patients have distinct physiological characteristics compared with medical ICU populations. Following major surgery, systemic inflammation, hormonal stress responses, and postoperative factors such as pain, sedation, and transient diaphragmatic dysfunction can modify both metabolic demand and respiratory workload [29]. These conditions can lower the observed ∆REE and OCOB thresholds compared with studies in medical ICU settings, where chronic respiratory or cardiac disease often is prevalent. Therefore, the present findings likely reflect postoperative metabolic adaptation and transient respiratory effort rather than underlying chronic illness, highlighting the specific relevance of calorimetric assessment in SICU patients.
Although ∆REE and OCOB showed good predictive performance, traditional weaning indices such as the rapid shallow breathing index, maximum inspiratory pressure, and vital capacity were not included because they were not part of our protocol, which focused on metabolic and oxygen-cost parameters measured by IC. These findings suggest that ∆REE and OCOB can complement rather than replace conventional predictors. Future studies should examine different surgical subgroups and include markers of nutritional status and muscle mass to better characterize metabolic diversity among postoperative patients. Predictive models integrating calorimetric, physiological, and clinical parameters will further enhance the accuracy and clinical applicability of weaning assessments.
There were some limitations on our study. First, because of the short research period, there were very few study participants. Our sample size enabled close monitoring of patients, especially for post-extubation discomfort, which minimized the risk of measurement errors. Additionally, due to the small number of weaning failures (n=16), we did not perform multivariable penalized regression or bootstrapping validation to avoid model instability and overfitting. Instead, we applied univariate and ROC-based analyses, which are statistically appropriate for small-sample diagnostic studies, and demonstrated robust discriminatory performance (AUC, 0.76–0.77). Second, we were unable to compare our results with studies that used manometers or traditional respiratory indices such as rapid shallow breathing index, maximum inspiratory pressure, and vital capacity, because our data were obtained solely from ventilators with IC. This limitation prevented us from evaluating whether ∆REE and OCOB offer additional or superior predictive value compared with conventional weaning parameters. Last, several baseline characteristics that might potentially operate as confounders in this study might have been missed. Serum albumin or prealbumin levels were not consistently available for all patients, which limited our ability to assess the potential contribution of nutritional status to weaning outcomes. Future studies with larger sample sizes should further validate these findings using penalized or resampling-based multivariable approaches.
In conclusion, the success of the entire weaning process can be predicted using both OCOB and REE. For surgical critical care patients, OCOB ≤9.1% and/or ∆REE ≤8.28% can be indicative of successful ventilator weaning.
▪ The research was carried out on surgical intensive care patients who had been receiving mechanical breathing for at least 48 hours.
▪ Indirect energy expenditure and oxygen cost of respiration were non-invasively measured using a ventilator-attached module.
▪ Variation in resting energy expenditure ≤8.28% or oxygen cost of breathing ≤9.1% predicts successful weaning from mechanical ventilation with good prognostic value.

CONFLICT OF INTEREST

No potential conflict of interest relevant to this article was reported.

FUNDING

None.

ACKNOWLEDGMENTS

We sincerely thank the medical staff of the Center of Anesthesiology, Critical Care and Pain Management, Hanoi Medical University Hospital for conducting this study.

AUTHOR CONTRIBUTIONS

Conceptualization: NHT. Data curation: VHP, TVD, HTH, DTD. Formal analysis: VHP, TVD, DTD, PQM.

Methodology: NHT, VHP, TVD. Project administration: NHT. Visualization: VHP, TVD, HTH. Writing – original draft: TVD, DTD, PQM. Writing – review & editing: VHP, TVD, HTH, PQM. All authors read and agreed to the published version of the manuscript.

Figure 1.
Flow diagram of study participants. SICU: surgical intensive care unit.
acc-001150f1.jpg
Figure 2.
Box plot of variation in resting energy expenditure (∆REE) (A) and oxygen cost of breathing (OCOB) (B) between success and failure groups in the weaning process
acc-001150f2.jpg
Figure 3.
Area under the curve (AUC) of variation in resting energy expenditure (A) and oxygen cost of breathing (B) in the entire ventilator weaning process.
acc-001150f3.jpg
Table 1.
Patient characteristics
Characteristics Total (n=58) Successful group (n=42) Failure group (n=16) P-value
Male sex 43 (74.1) 31 (73.8) 12 (75) 0.926a)
Age (yr) 63 (49–73) 59 (45–71) 71 (51–82) 0.093b)
BMI (kg/m2) 21.2 (17.7–22.6) 21.3 (18.5–22.7) 18.0 (17.6–22.2) 0.189b)
APACHE II score 10.0 (7.0–11.3) 9.0 (6.8–10.0) 12.0 (11.0–14.8) <0.001b)
Surgical type
 Gastrointestinal 23 (39.7) 17 (40.5) 6 (37.5) 0.836a)
 Neurosurgical 14 (24.1) 9 (21.4) 5 (31.3) 0.435a)
 Cardiovascular 12 (20.7) 9 (21.4) 3 (18.8) 0.822c)
 Trauma 5 (8.6) 4 (9.5) 1 (6.2) 0.691c)
 SSTI 4 (6.9) 3 (7.2) 1 (6.2) 1.000c)
Surgical urgency 0.765a)
 Elective 38 (65.5) 28 (66.7) 10 (62.5)
 Emergency 20 (34.5) 14 (33.3) 6 (37.5)
Comorbidity
 Hypertension 18 (31.0) 12 (28.6) 6 (37.5) 0.511a)
 Diabetes mellitus 11 (19.0) 8 (19.0) 3 (18.8) 1.000c)
 CKD 7 (12.1) 5 (11.9) 2 (12.5) 1.000c)
 Malignancy 8 (13.8) 6 (14.3) 2 (12.3) 1.000c)
Length of hospital stay (day) 15.5 (11–22.3) 14.5 (11.8–23.5) 16.0 (10.3–21.0) 0.583b)
Length of SICU stay (day) 4.0 (3.0–6.0) 4.0 (3.0–5.0) 5.5 (3.0–8.5) 0.082b)
Ventilation duration (day) 3.0 (2.0–5.0) 4.0 (3.0–5.0) 3.5 (2.0–6.0) 0.979b)

Values are presented as number (%) or median (interquartile range).

BMI: body mass index; APACHE: Acute Physiology and Chronic Health Evaluation; SSTI: skin and soft tissue infection; CKD: chronic kidney disease; SICU: surgical intensive care unit.

a)Chi-square test;

b)Mann-Whitney U-test;

c)Fisher’s exact test.

Table 2.
The difference of ∆REE and OCOB index between success and failure groups in the ventilator weaning process
Index Success group (n=42) Failure group (n=16) P-valuea)
Baseline
 V˙O2 (ml/min) 207.5 (186.0–240.8) 183.0 (167.3–243.0) 0.128
 V˙CO2 (ml/min) 167.0 (153.5–189.8) 147.0 (136.5–186.0) 0.080
 RQ 0.815 (0.775–0.840) 0.810 (0.737–0.840) 0.979
 REE (kcal/day) 1,449.0 (1,300.3–1,675.3) 1,255.0 (1,180.8–1,693.5) 0.095
Last 5 minutes of the trial
 V˙O2 (ml/min) 230.5 (199.8–275.3) 223.5 (195.0–299.8) 0.938
 V˙CO2 (ml/min) 176.5 (160.0–201.0) 169.5 (154.3–222.0) 0.903
 RQ 0.79 (0.72–0.84) 0.80 (0.682–0.855) 0.944
 REE (kcal/day) 1,578.5 (1,370.5–1,892.0) 1,515.5 (1,366.8–2,104.0) 0.896
∆REE (%) 6.30 (3.75–10.08) 15.64 (8.37–30.59) 0.002
OCOB (%) 6.93 (3.81–10.92) 16.21 (9.27–29.76) 0.002

Values are presented as median (interquartile range).

∆REE: variation in resting energy expenditure; OCOB: oxygen cost of breathing; V˙O2: oxygen consumption; V˙CO2: carbon dioxide production; RQ: respiratory quotient; REE: resting energy expenditure.

a)Mann-Whitney U-test.

  • 1. Potchileev I, Doroshenko M, Mohammed AN. Positive pressure ventilation [Internet]. StatPearls Publishing; 2023 [cited 2025 Dec 8]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK560916
  • 2. Matarese LE. Indirect calorimetry: technical aspects. J Am Diet Assoc 1997;97:S154-60.ArticlePubMed
  • 3. De Waele E, Jonckheer J, Wischmeyer PE. Indirect calorimetry in critical illness: a new standard of care? Curr Opin Crit Care 2021;27:334-43.ArticlePMC
  • 4. Priem S, Jonckheer J, De Waele E, Stiens J. Indirect calorimetry in spontaneously breathing, mechanically ventilated and extracorporeally oxygenated patients: an engineering review. Sensors (Basel) 2023;23:4143.Article
  • 5. Shikora SA, Bistrian BR, Borlase BC, Blackburn GL, Stone MD, Benotti PN, et al. Work of breathing: reliable predictor of weaning and extubation. Crit Care Med 1990;18:157-62.
  • 6. Epstein SK. Weaning parameters. Respir Care Clin N Am 2000;6:253-301.Article
  • 7. Bellani G, Foti G, Spagnolli E, Milan M, Zanella A, Greco M, et al. Increase of oxygen consumption during a progressive decrease of ventilatory support is lower in patients failing the trial in comparison with those who succeed. Anesthesiology 2010;113:378-85.ArticlePDF
  • 8. Raurich JM, Ibáñez J. Oxygen cost of breathing in the prediction of successful weaning from mechanical ventilation. Med Intensiva 2007;31:172-8.ArticlePubMed
  • 9. Finnerty CC, Mabvuure NT, Ali A, Kozar RA, Herndon DN. The surgically induced stress response. JPEN J Parenter Enteral Nutr 2013;37:21S-29S.ArticlePubMedPMCPDF
  • 10. Suzuki S. Oxygen administration for postoperative surgical patients: a narrative review. J Intensive Care 2020;8:79.ArticlePubMedPMCPDF
  • 11. Huang CT, Tsai YJ, Lin JW, Ruan SY, Wu HD, Yu CJ, et al. Application of heart-rate variability in patients undergoing weaning from mechanical ventilation. Crit Care 2014;18:R21.ArticlePubMedPMCPDF
  • 12. Bien Udos S, Souza GF, Campos ES, Farah de Carvalho E, Fernandes MG, Santoro I, et al. Maximum inspiratory pressure and rapid shallow breathing index as predictors of successful ventilator weaning. J Phys Ther Sci 2015;27:3723-7.ArticlePubMedPMC
  • 13. Zein H, Baratloo A, Negida A, Safari S. Ventilator weaning and spontaneous breathing trials; an educational review. Emerg (Tehran) 2016;4:65-71.PubMedPMC
  • 14. Boles JM, Bion J, Connors A, Herridge M, Marsh B, Melot C, et al. Weaning from mechanical ventilation. Eur Respir J 2007;29:1033-56.ArticlePubMed
  • 15. Heunks LM, van der Hoeven JG. Clinical review: the ABC of weaning failure--a structured approach. Crit Care 2010;14:245.ArticlePubMedPMC
  • 16. Rehal MS, Fiskaare E, Tjäder I, Norberg Å, Rooyackers O, Wernerman J, et al. Measuring energy expenditure in the intensive care unit: a comparison of indirect calorimetry by E-sCOVX and Quark RMR with Deltatrac II in mechanically ventilated critically ill patients. Crit Care 2016;20:54.ArticlePubMedPMC
  • 17. Metabolic measurement using indirect calorimetry during mechanical ventilation—2004 revision & update. Respir Care 2004;49:1073-9.ArticlePDF
  • 18. Puetpaiboon A, Chatmongkolchart S, Oofuvong M. The outcomes and predictors associated with prolonged weaning of postoperative patients requiring mechanical ventilation in general wards. Anaesth Pain Intensive Care 2022;26:161-7.ArticlePDF
  • 19. Yıldırım F, Karabacak H, Kaya İO. Factors affecting weaning failure in critically-ill patients undergoing emergency gastrointestinal surgery. J Crit Intensive Care 2020;11:8-14.
  • 20. Matic I, Titlic M, Dikanovic M, Jurjevic M, Jukic I, Tonkic A, et al. Effects of APACHE II score on mechanical ventilation; prediction and outcome. Acta Anaesthesiol Belg 2007;58:177-83.PubMed
  • 21. Meade M, Guyatt G, Cook D, Griffith L, Sinuff T, Kergl C, et al. Predicting success in weaning from mechanical ventilation. Chest 2001;120:400S-24S.Article
  • 22. Carpenè N, Vagheggini G, Panait E, Gabbrielli L, Ambrosino N. A proposal of a new model for long-term weaning: respiratory intensive care unit and weaning center. Respir Med 2010;104:1505-11.Article
  • 23. McConville JF, Kress JP. Weaning patients from the ventilator. N Engl J Med 2012;367:2233-9.Article
  • 24. Miwa K, Mitsuoka M, Takamori S, Hayashi A, Shirouzu K. Continuous monitoring of oxygen consumption in patients undergoing weaning from mechanical ventilation. Respiration 2003;70:623-30.ArticlePDF
  • 25. Mitsuoka M, Kinninger KH, Johnson FW, Burns DM. Utility of measurements of oxygen cost of breathing in predicting success or failure in trials of reduced mechanical ventilatory support. Respir Care 2001;46:902-10.ArticlePDF
  • 26. Muñoz IC, Hernández AM, Mañanas MÁ. Estimation of work of breathing from respiratory muscle activity in spontaneous ventilation: a pilot study. Appl Sci 2019;9:2007.Article
  • 27. Savchenko YV, Goryachev AS, Savin IA, Krylov KY, Polupan AA, Birg TM, et al. Changes in energy expenditures while reducing pressure support as a predictor of successful spontaneous breathing trial. Vestn Anesteziol Reanimatol 2020;17:24-31.ArticlePDF
  • 28. Santos L, Vieira S. Metabolic evaluation during weaning from mechanical ventilation using indirect calorimetry. Crit Care 2008;12:P140.Article
  • 29. Soop M, Ljungqvist O. Metabolic responses to surgical stress. Clin Nutr ESPEN 2025;67:178-83.Article

Figure & Data

References

    Citations

    Citations to this article as recorded by  

      • PubReader PubReader
      • ePub LinkePub Link
      • Cite
        CITE
        export Copy
        Close
        Download Citation
        Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.
        Format:
        • RIS — For EndNote, ProCite, RefWorks, and most other reference management software
        • BibTeX — For JabRef, BibDesk, and other BibTeX-specific software
        Include:
        • Citation for the content below
        Predictive value of resting energy expenditure and oxygen consumption of breathing for ventilator weaning in surgical intensive care patients in Viet Nam: a cross-sectional study
        Acute Crit Care. 2026;41(1):97-106.   Published online February 27, 2026
        DOI: https://doi.org/10.4266/acc.001150
        Close
      • XML DownloadXML Download
      Figure
      • 0
      • 1
      • 2
      Related articles
      Predictive value of resting energy expenditure and oxygen consumption of breathing for ventilator weaning in surgical intensive care patients in Viet Nam: a cross-sectional study
      Image Image Image
      Figure 1. Flow diagram of study participants. SICU: surgical intensive care unit.
      Figure 2. Box plot of variation in resting energy expenditure (∆REE) (A) and oxygen cost of breathing (OCOB) (B) between success and failure groups in the weaning process
      Figure 3. Area under the curve (AUC) of variation in resting energy expenditure (A) and oxygen cost of breathing (B) in the entire ventilator weaning process.

      Figure 1.

      Figure 2.

      Figure 3.

      Predictive value of resting energy expenditure and oxygen consumption of breathing for ventilator weaning in surgical intensive care patients in Viet Nam: a cross-sectional study
      Characteristics Total (n=58) Successful group (n=42) Failure group (n=16) P-value
      Male sex 43 (74.1) 31 (73.8) 12 (75) 0.926a)
      Age (yr) 63 (49–73) 59 (45–71) 71 (51–82) 0.093b)
      BMI (kg/m2) 21.2 (17.7–22.6) 21.3 (18.5–22.7) 18.0 (17.6–22.2) 0.189b)
      APACHE II score 10.0 (7.0–11.3) 9.0 (6.8–10.0) 12.0 (11.0–14.8) <0.001b)
      Surgical type
       Gastrointestinal 23 (39.7) 17 (40.5) 6 (37.5) 0.836a)
       Neurosurgical 14 (24.1) 9 (21.4) 5 (31.3) 0.435a)
       Cardiovascular 12 (20.7) 9 (21.4) 3 (18.8) 0.822c)
       Trauma 5 (8.6) 4 (9.5) 1 (6.2) 0.691c)
       SSTI 4 (6.9) 3 (7.2) 1 (6.2) 1.000c)
      Surgical urgency 0.765a)
       Elective 38 (65.5) 28 (66.7) 10 (62.5)
       Emergency 20 (34.5) 14 (33.3) 6 (37.5)
      Comorbidity
       Hypertension 18 (31.0) 12 (28.6) 6 (37.5) 0.511a)
       Diabetes mellitus 11 (19.0) 8 (19.0) 3 (18.8) 1.000c)
       CKD 7 (12.1) 5 (11.9) 2 (12.5) 1.000c)
       Malignancy 8 (13.8) 6 (14.3) 2 (12.3) 1.000c)
      Length of hospital stay (day) 15.5 (11–22.3) 14.5 (11.8–23.5) 16.0 (10.3–21.0) 0.583b)
      Length of SICU stay (day) 4.0 (3.0–6.0) 4.0 (3.0–5.0) 5.5 (3.0–8.5) 0.082b)
      Ventilation duration (day) 3.0 (2.0–5.0) 4.0 (3.0–5.0) 3.5 (2.0–6.0) 0.979b)
      Index Success group (n=42) Failure group (n=16) P-valuea)
      Baseline
       V˙O2 (ml/min) 207.5 (186.0–240.8) 183.0 (167.3–243.0) 0.128
       V˙CO2 (ml/min) 167.0 (153.5–189.8) 147.0 (136.5–186.0) 0.080
       RQ 0.815 (0.775–0.840) 0.810 (0.737–0.840) 0.979
       REE (kcal/day) 1,449.0 (1,300.3–1,675.3) 1,255.0 (1,180.8–1,693.5) 0.095
      Last 5 minutes of the trial
       V˙O2 (ml/min) 230.5 (199.8–275.3) 223.5 (195.0–299.8) 0.938
       V˙CO2 (ml/min) 176.5 (160.0–201.0) 169.5 (154.3–222.0) 0.903
       RQ 0.79 (0.72–0.84) 0.80 (0.682–0.855) 0.944
       REE (kcal/day) 1,578.5 (1,370.5–1,892.0) 1,515.5 (1,366.8–2,104.0) 0.896
      ∆REE (%) 6.30 (3.75–10.08) 15.64 (8.37–30.59) 0.002
      OCOB (%) 6.93 (3.81–10.92) 16.21 (9.27–29.76) 0.002
      Table 1. Patient characteristics

      Values are presented as number (%) or median (interquartile range).

      BMI: body mass index; APACHE: Acute Physiology and Chronic Health Evaluation; SSTI: skin and soft tissue infection; CKD: chronic kidney disease; SICU: surgical intensive care unit.

      Chi-square test;

      Mann-Whitney U-test;

      Fisher’s exact test.

      Table 2. The difference of ∆REE and OCOB index between success and failure groups in the ventilator weaning process

      Values are presented as median (interquartile range).

      ∆REE: variation in resting energy expenditure; OCOB: oxygen cost of breathing; V˙O2: oxygen consumption; V˙CO2: carbon dioxide production; RQ: respiratory quotient; REE: resting energy expenditure.

      Mann-Whitney U-test.

      Table 1.

      Table 2.

      ACC : Acute and Critical Care
      © The Korean Society of Critical Care Medicine.
      TOP