Effect of an Integrated Early Enteral Nutrition Management Protocol on Nutritional Target Attainment in Mechanically Ventilated ICU Patients: A Quasi‐Experimental Study Guided by the MRC Framework Authors: Jie Zhang, Nan Wang, Yongqing Hang, Jinya Huang, Ping Yang, Peibei Duan Published: 2026 ABSTRACT ABSTRACTBackgroundAchieving early enteral nutrition (EN) targets remains challenging in mechanically ventilated intensive care unit (ICU) patients. Evidence for nursing‐led multicomponent protocols incorporating traditional Chinese medicine (TCM) approaches is limited.AimsTo develop an integrated early EN management protocol using the Medical Research Council (MRC) framework and conduct a preliminary evaluation in mechanically ventilated ICU patients.Study DesignA single‐centre quasi‐experimental before‐and‐after study was conducted in a tertiary ICU in China. Adults at nutritional risk receiving mechanical ventilation (MV) and EN were enrolled during sequential control and intervention phases. The intervention group received the MRC‐guided protocol and the control group received routine EN care. The primary outcome was EN target attainment on Days 3 and 7.ResultsEighty‐two patients were analysed (intervention, n = 40; control, n = 42). EN target attainment was higher in the intervention group on Day 3 (70.0% ± 17.5% vs. 53.5% ± 18.7%; p < 0.001; Cohen's d = 0.91) and Day 7 (83.7% ± 9.4% vs. 64.1% ± 15.4%; p < 0.001; Cohen's d = 1.53). Both differences remained significant after adjustment for calendar time. Earlier EN initiation and post‐pyloric feeding were associated with higher attainment in exploratory group‐adjusted analyses. Albumin showed a significant group‐by‐time interaction (p = 0.002). MV duration, ICU length of stay and feeding intolerance did not differ significantly between groups.ConclusionsIn this single‐centre quasi‐experimental study, protocol implementation was associated with higher early EN target attainment and was feasible in the study setting, without a detected increase in feeding intolerance. Causal inference is limited by the before‐and‐after design, and component‐specific effects require evaluation in multicentre randomised controlled trials.Relevance to Clinical PracticeThe protocol provides critical care nurses with a structured approach to early EN delivery, tolerance monitoring and supportive care. Implementation elsewhere requires context‐specific adaptation and further evaluation.Keywords: enteral nutrition, intensive care unit, mechanical ventilation, Medical Research Council framework, traditional Chinese medicine Introduction Respiratory failure is one of the most common life‐threatening conditions encountered in the intensive care unit (ICU), and mechanical ventilation (MV) remains the cornerstone of life support for these patients [1]. Under the combined effects of disease‐related stress, hypermetabolism, prolonged sedation and analgesia, and impaired gastrointestinal perfusion, critically ill patients are highly vulnerable to malnutrition, with a reported prevalence of 30%–50% [2, 3]. If not corrected in a timely manner, malnutrition may further compromise intestinal barrier integrity, increase the risks of infection and organ dysfunction and ultimately worsen patient outcomes [4]. Early enteral nutrition (EN) has been recommended by multiple guidelines as the preferred or first‐line route of nutrition support for critically ill patients [3, 5]. Previous studies have shown that EN helps maintain intestinal mucosal integrity, reduce the translocation of microorganisms and their products, preserve immune function and promote gastrointestinal motility [6]. However, feeding intolerance remains one of the major barriers to EN delivery. Its reported incidence varies considerably across studies, ranging from 30.5% to 75% [7, 8]. Feeding intolerance frequently leads to feeding interruption, rate reduction or delayed achievement of nutritional targets, thereby resulting in actual energy delivery substantially below prescribed requirements and limiting the effectiveness of nutrition support. In recent years, traditional Chinese medicine (TCM)‐based supportive techniques have attracted increasing attention in the management of gastrointestinal dysfunction in critically ill patients because of their relative simplicity, bedside feasibility and generally acceptable safety profile [9]. Existing studies suggest that Zusanli (ST36) acupoint stimulation, auricular acupressure and related auricular stimulation techniques may improve gastrointestinal motility and facilitate gastrointestinal functional recovery by modulating autonomic nervous activity and gastrointestinal hormone secretion [10, 11]. Nevertheless, TCM interventions for early EN management in mechanically ventilated patients have largely remained at the level of isolated single‐component applications. A standardised management protocol developed through the integration of evidence‐based recommendations is still lacking, and its clinical effectiveness has yet to be adequately evaluated. Moreover, interventions involving multiple components, multiple implementers and multilevel outcomes require a rigorous methodological framework for systematic development and evaluation. The 2021 updated Medical Research Council (MRC) framework for complex interventions provides such guidance [12]. It emphasises the iterative and nonlinear nature of the research process and highlights the roles of context, theoretical underpinning and process evaluation in intervention development, implementation and assessment. Against this background, the present study was guided by the MRC framework for complex interventions and informed by relevant national and international guidelines as well as high‐quality research evidence. We developed an integrated early EN management protocol combining conventional and TCM‐based approaches for mechanically ventilated ICU patients and conducted a quasi‐experimental study to evaluate EN target attainment and related clinical outcomes. In addition, a logic model was constructed to specify the hypothesised pathways linking the intervention components to implementation and outcomes. This study aimed to provide preliminary evidence to inform the refinement and future evaluation of the protocol. Methods Framework for Intervention Development This study was guided by the 2021 updated MRC framework for complex interventions [12]. The Joanna Briggs Institute model of evidence‐based healthcare was used for evidence synthesis [13]. Guided by these frameworks, the present study comprised three study‐specific activities: intervention development, logic model construction with expert consultation and preliminary effectiveness evaluation. The overall methodological framework and study positioning are illustrated in Figure 1. The logic model depicting the hypothesised pathways between intervention components and outcomes is presented in Figure 2. The intervention was developed from core guidelines, systematic reviews and high‐quality clinical studies on EN management in the ICU [3, 5]. Evidence synthesis focused on nutritional risk screening and assessment, timing of EN initiation, feeding route and delivery mode, nutritional requirements and targets, formula selection, prevention and management of complications and team/process management. The preliminary protocol was refined through two rounds of multidisciplinary expert consultation and finalised as a standard operating procedure (SOP). In line with the MRC framework, the present quasi‐experimental study was designed as a preliminary evaluation to identify early signals of effectiveness, assess feasibility and inform future definitive trials [12]. Application of the Medical Research Council framework for complex interventions in the present study. MRC, Medical Research Council; RCT, randomised controlled trial; SOP, standard operating procedure. Logic model of the integrated early enteral nutrition management intervention. AGI, acute gastrointestinal injury; ALB, albumin; APACHE II, Acute Physiology and Chronic Health Evaluation II; EICU, emergency intensive care unit; EN, enteral nutrition; GRV, gastric residual volume; Hb, haemoglobin; IAP, intra‐abdominal pressure; ICU, intensive care unit; MDT, multidisciplinary team; MV, mechanical ventilation; NJT, nasojejunal tube; NRS‐2002, Nutritional Risk Screening 2002; PAB, prealbumin; RASS, Richmond Agitation–Sedation Scale; SGA, Subjective Global Assessment; SOP, standard operating procedure; ST36, Zusanli; TCM, traditional Chinese medicine; T0, enrolment; T1, study day 3; T2, study day 7; T3, within 24 h before transfer from the ICU. Study Design This single‐centre quasi‐experimental before‐and‐after study was conducted in the ICU of a tertiary hospital in China between July 2024 and December 2025. The control phase was conducted from July 2024 to March 2025, followed by the intervention phase from April to December 2025. Participants Eligible participants were adults aged ≥ 18 years who had remained in the ICU for at least 24 h and had a Nutritional Risk Screening 2002 (NRS‐2002) score ≥ 3. Additional eligibility criteria were EN support through a gastric or nasojejunal tube, an expected MV duration of at least 72 h, and a mean arterial pressure ≥ 65 mmHg. Exclusion criteria were contraindications to EN (e.g., active upper gastrointestinal bleeding or intestinal obstruction), traumatic brain injury, postoperative abdominal surgery, incomplete data precluding outcome evaluation or refusal to participate. Sample size was calculated using EN target attainment as the primary outcome. The calculation assumed a historical EN target attainment rate of 50% in our ICU and an increase to 80% after the intervention. With a two‐sided α of 0.05 and 80% power, at least 36 patients per group were required after allowing for a 10% dropout or exclusion rate. We therefore planned to enrol 40 patients in each group. Intervention Control Group The control group received routine early EN care, including continuous pump‐assisted delivery of enteral formula as prescribed, routine monitoring of feeding intolerance and standard supportive nursing care such as head‐of‐bed elevation, oral care, feeding tube flushing and lower‐limb circulation support. No protocolised TCM‐based supportive interventions, structured tolerance assessment pathway or study‐specific quality control measures were implemented. Intervention Group In addition to routine care, the intervention group received an integrated early EN management protocol combining conventional care with TCM‐based supportive approaches. The protocol comprised five modules: pre‐implementation assessment, optimisation of feeding strategy, tolerance monitoring with TCM‐based supportive interventions, adjunctive functional support and whole‐process quality control. All procedures were delivered according to the SOP by staff who had completed standardised training and competency assessment. A concise overview of the five protocol modules, their core components and their intended functions and hypothesised pathways is provided in Supporting Information Table S1. Module 1: Pre‐Implementation Assessment Within 4 h of ICU admission, patients underwent NRS‐2002 screening [3]. Those at nutritional risk received further assessment using Subjective Global Assessment and Acute Gastrointestinal Injury (AGI) grading to determine EN feasibility [3]. Target energy requirements were estimated using a weight‐based formula of 25 kcal/kg/day [5, 14]. A multidisciplinary team (MDT) consisting of physicians, nurses and dietitians formulated individualised EN plans. Module 2: Optimisation of Feeding Strategy EN was initiated within 24–48 h after ICU admission when haemodynamic stability was achieved [3]. Special treatment status alone, including prone positioning, neuromuscular blockade or extracorporeal membrane oxygenation, was not considered a routine reason to delay EN; trophic feeding was used when appropriate [15]. Gastric feeding was used initially, while post‐pyloric feeding was considered for patients at high aspiration risk or with impaired gastric emptying [3]. EN was delivered continuously via pump and progressively advanced according to tolerance [16]. Target energy and protein intake were set at 25 kcal/kg/day and 1.2–2.0 g/kg/day, respectively [3, 5]. A standard whole‐protein formula was used as first‐line therapy. The formula was adjusted according to fluid restriction, stress hyperglycaemia and the patient's clinical condition. Immunomodulating formulas containing omega‐3 fatty acids or glutamine were not routinely used in patients with acute respiratory distress syndrome or sepsis [17]. Module 3: Tolerance Monitoring and TCM‐Based Supportive Interventions A protocolised tolerance assessment pathway was used, with reassessment every 6 h for abdominal symptoms, gastric retention, nausea/vomiting, aspiration, gastrointestinal bleeding, bowel sounds and AGI grade [18]. In patients with marked abdominal distension, intra‐abdominal pressure was monitored and EN was reduced or withheld if needed [19]. High‐risk patients received aspiration‐prevention measures, including cuff pressure management and head‐of‐bed elevation; gastric residual volume was assessed only when clinically indicated [5, 20]. TCM‐based supportive interventions were delivered according to symptom stratification. For abdominal distension or constipation, auricular acupressure, abdominal acupoint massage and acupoint application were used, with soluble‐fibre‐enriched formulas or lactulose when necessary [21, 22]. For nausea or vomiting, acupoint massage targeting Zhongwan (CV12), Neiguan (PC6) and ST36 was combined with auricular acupressure, with prokinetic medication added when indicated [23]. For diarrhoea, auricular acupressure and relevant acupoint massage were applied, with formula adjustment as needed [22]. For gastric retention, acupoint massage, acupoint application and auricular acupressure were provided in addition to standard prokinetic therapy [24, 25]. Detailed standardised procedures for the TCM‐based supportive interventions, including their clinical indications, acupoints, intervention materials, operational steps, session duration, frequency, treatment course, safety monitoring and documentation requirements, are provided in Supporting Information Table S2. Module 4: Adjunctive Functional Support Adjunctive supportive measures were used to facilitate gastrointestinal motility, functional recovery and adherence. Deeply sedated or unconscious patients received intermittent pneumatic compression and passive limb mobilisation [26]. Patients with marked gastric retention additionally received clockwise abdominal massage [22]. Patients with light sedation or preserved consciousness received active in‐bed exercises [26]. Individualised nutrition education was provided to communicative patients and family caregivers [27]. Module 5: Whole‐Process Quality Control This module aimed to enhance intervention fidelity and consistency. It included standardisation of EN procedures, staff training before study initiation, daily supervision by the head nurse or designated quality‐control nurse, real‐time correction of deviations and regular MDT review of nutritional support. Family caregivers also received nutrition‐related education throughout the study. Detailed operational procedures are provided in the SOP. Outcomes The primary outcome was EN target attainment (%) at the prespecified assessment time points of study days 3 and 7. For outcome calculation, the estimated full daily energy target was calculated as enrolment body weight × 25 kcal/kg/day, using the same criterion in both groups. At each assessment time point, EN target attainment was calculated as follows: Secondary laboratory outcomes were haemoglobin (Hb), albumin (ALB) and prealbumin (PAB), measured at enrolment (T0), on study day 3 (T1), on study day 7 (T2) and within 24 h before transfer out of the ICU (T3). Other secondary outcomes were MV duration and ICU length of stay. Process outcomes were time to EN initiation, the proportion of patients receiving post‐pyloric feeding and intervention fidelity. Feeding intolerance was the safety outcome. It was defined as the occurrence of at least one of the following: abdominal distension, constipation, gastric retention (≥ 500 mL/24 h), diarrhoea (≥ 3 episodes/day), nausea or vomiting, aspiration or gastrointestinal bleeding indicated by a positive faecal occult blood test. Component‐level fidelity was evaluated separately for auricular acupressure, acupoint massage and herbal acupoint application. Clinically eligible patients were those who met the prespecified symptom‐stratified indication for the relevant technique and had no corresponding contraindication. Fidelity measures comprised receipt of at least one session among clinically eligible patients and among the entire intervention group, the numbers of completed and scheduled sessions, the number of sessions per recipient, protocol deviations and unplanned discontinuations. Session‐level adherence was calculated as the number of completed sessions divided by the number of scheduled sessions × 100%. Scheduled sessions included only those required by the SOP while the relevant technique remained clinically indicated, tolerated and not contraindicated. Blood samples were collected at T0–T3 as part of routine clinical sampling. Hb was measured in ethylenediaminetetraacetic acid‐anticoagulated whole blood using a Sysmex XN‐9000 automated haematology analyser (Sysmex Corporation, Kobe, Japan) based on the sodium lauryl sulphate haemoglobin method. Serum ALB and PAB were measured using an AU5800 automated biochemistry analyser (Beckman Coulter, Brea, CA, USA). ALB was determined using the bromocresol green method, and PAB was measured using an immunoturbidimetric assay. All measurements were performed by the hospital's central laboratory in accordance with the manufacturers' instructions and established SOPs. The laboratory operated under an ISO 15189‐accredited quality management system, with routine internal quality control procedures applied to all analyses. Statistical Analysis Analyses were performed using IBM SPSS Statistics, version 26.0 (IBM Corp., Armonk, NY, USA) and R version 4.5.1 (R Foundation for Statistical Computing, Vienna, Austria). The distribution of continuous variables was assessed using histograms and Q–Q plots, supplemented by the Shapiro–Wilk test. Normally distributed continuous variables were summarised as mean ± standard deviation (SD) and compared using independent‐samples t‐tests. Categorical variables were summarised as n (%) and compared using the chi‐square test or Fisher's exact test, as appropriate. Follow‐up Hb, ALB and PAB measurements at T1–T3 were analysed using linear mixed‐effects models. Patient was included as a random effect, whereas group, time and the group × time interaction were included as fixed effects. The corresponding T0 value was included as a covariate. Exploratory associations between process outcomes and EN target attainment were examined separately on Days 3 and 7 using Spearman's rank correlation for time to EN initiation and point‐biserial correlation for post‐pyloric feeding. Group‐adjusted sensitivity analyses were performed using separate patient‐level linear regression models for each process measure and assessment time point; each model included study group and either time to EN initiation or post‐pyloric feeding. Heteroscedasticity‐consistent (HC3) robust standard errors were used. The Benjamini–Hochberg procedure was applied separately to the four correlation tests and the four group‐adjusted process‐measure coefficients to control the false discovery rate. Because the process outcomes were not independently assigned, these analyses were considered exploratory and were not intended to establish causal, component‐specific or mediating effects. To assess potential temporal confounding arising from the quasi‐experimental before‐and‐after design, sensitivity analyses were performed separately for Day 3 and Day 7 EN target attainment using patient‐level multiple linear regression models. Calendar time was calculated in months from the beginning of the study based on each patient's admission date and was entered as a continuous covariate together with study group. The adjusted group coefficient represented the between‐group difference in EN target attainment after accounting for a linear calendar‐time trend. Adjusted mean differences and 95% confidence intervals (CIs) were reported. All tests were two‐sided, with p < 0.05 indicating statistical significance. Cohen's d was calculated for continuous outcomes and classified by absolute magnitude as negligible (< 0.20), small (0.20–0.49), medium (0.50–0.79) or large (≥ 0.80) [28, 29]. Ethical Considerations, Institutional Approval and Trial Registration Ethical approval was granted by the Medical Ethics Committee of Jiangsu Province Hospital of Chinese Medicine (Affiliated Hospital of Nanjing University of Chinese Medicine) on 17 October 2023 (Approval No. 2023NL‐234‐02). The study was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants or their legally authorised representatives before enrolment. The study was registered in the Chinese Clinical Trial Registry (ChiCTR2400086184). Results Participant Flow and Intervention Fidelity Of the 106 potentially eligible patients screened, 20 were excluded and 86 were enrolled. Four patients were subsequently excluded, leaving 82 patients in the final analysis (intervention group, n = 40; control group, n = 42). The participant flow is shown in Figure 3. Flow diagram of participant screening, allocation, follow‐up and analysis. EICU, emergency intensive care unit; EN, enteral nutrition; FAS, full analysis set; MV, mechanical ventilation. Protocol‐delivery indicators showed high implementation fidelity. All patients in the intervention group initiated EN within 48 h after ICU admission. The post‐training competency pass rate for the TCM‐based supportive techniques was 100%, and more than 95% of scheduled 6‐hourly tolerance assessments were completed. Auricular acupressure and acupoint massage were each delivered to all 13 clinically eligible patients (100.0%; 32.5% of the intervention group), whereas herbal acupoint application was delivered to all nine eligible patients (100.0%; 22.5% of the intervention group). Session‐level adherence was 97.2% for auricular acupressure (140/144 scheduled sessions) and 100.0% for both acupoint massage (48/48) and herbal acupoint application (34/34). The median (interquartile range) number of sessions per recipient was 10 (9–12) for auricular acupressure, 4 (3–4) for acupoint massage and 4 (3–4) for herbal acupoint application. Four scheduled auricular acupressure sessions involving three patients were missed during nursing implementation. No unplanned discontinuations occurred (Supporting Information Table S3). Baseline Characteristics No statistically significant between‐group differences were observed in the measured baseline characteristics (all p > 0.05; Table 1). TABLE 1Baseline characteristics of participants (intervention group n = 40, control group n = 42).VariableIntervention (n = 40)Control (n = 42)Statistic p Age (years)75.0 ± 11.273.1 ± 15.7 t = 0.6460.520Sex, male, n (%)30 (75.0)26 (61.9) χ 2 = 1.0740.300Height (cm)167.9 ± 6.7167.9 ± 6.8 t = −0.0030.997Weight (kg)65.6 ± 11.565.6 ± 11.4 t = −0.0070.994BMI (kg/m2)23.2 ± 3.623.3 ± 4.0 t = −0.0990.921APACHE II score22.8 ± 6.720.8 ± 6.7 t = 1.3440.183SOFA score8.5 ± 2.98.0 ± 3.3 t = 0.6630.509Primary diagnosis, n (%)Sepsis/septicaemia20 (50.0)23 (54.8) χ 2 = 0.0440.833Severe pneumonia/AECOPD8 (20.0)12 (28.6) χ 2 = 0.4180.518Cardiac arrest5 (12.5)2 (4.8) χ 2 = 0.7360.391Cardiac disease3 (7.5)2 (4.8) χ 2 = 0.0030.955Other4 (10.0)3 (7.1) χ 2 = 0.0050.946Comorbidities, n (%), multiple allowedCardiovascular disease25 (62.5)28 (66.7) χ 2 = 0.0270.870Endocrine/metabolic13 (32.5)17 (40.5) χ 2 = 0.2710.603Neurological disease14 (35.0)16 (38.1) χ 2 = 0.0040.951Pulmonary disease8 (20.0)12 (28.6) χ 2 = 0.4180.518Other9 (22.5)16 (38.1) χ 2 = 1.6730.196NRS‐2002 score4.7 ± 1.24.5 ± 1.2 t = 0.7500.456mNUTRIC score4.4 ± 1.44.2 ± 1.5 t = 0.5730.568Blood glucose (mmol/L)8.5 ± 2.39.0 ± 3.0 t = −0.9000.371ALB (g/L)31.3 ± 5.629.6 ± 5.8 t = 1.4120.162PAB (mg/L)105.8 ± 54.697.0 ± 51.0 t = 0.7540.453Hb (g/L)104.2 ± 30.8109.9 ± 27.9 t = −0.8730.385Vasoactive agents, n (%)10 (25.0)16 (38.1) χ 2 = 1.0740.300RASS ≥ −2, n (%)21 (52.5)20 (47.6) χ 2 = 0.0490.825CRRT, n (%)9 (22.5)8 (19.0) χ 2 = 0.0130.910Prone positioning, n (%)4 (10.0)4 (9.5) χ 2 = 0.0001.000SIMV mode, n (%)22 (55.0)21 (50.0) χ 2 = 0.0540.817Prokinetic agents, n (%)17 (42.5)15 (35.7) χ 2 = 0.1630.687AGI grade I–II, n (%)34 (85.0)28 (66.7) χ 2 = 2.8060.094 Note: Data are presented as mean ± SD or n (%). Continuous variables were compared using independent‐samples t‐tests, and categorical variables were compared using the chi‐square test.Abbreviations: AECOPD, acute exacerbation of chronic obstructive pulmonary disease; AGI, acute gastrointestinal injury; ALB, albumin; APACHE II, Acute Physiology and Chronic Health Evaluation II; BMI, body mass index; CRRT, continuous renal replacement therapy; Hb, haemoglobin; mNUTRIC, modified Nutrition Risk in the Critically Ill; NRS‐2002, Nutritional Risk Screening 2002; PAB, prealbumin; RASS, Richmond Agitation–Sedation Scale; SD, standard deviation; SIMV, synchronised intermittent mandatory ventilation; SOFA, Sequential Organ Failure Assessment. Baseline characteristics of participants (intervention group n = 40, control group n = 42). Note: Data are presented as mean ± SD or n (%). Continuous variables were compared using independent‐samples t‐tests, and categorical variables were compared using the chi‐square test. Abbreviations: AECOPD, acute exacerbation of chronic obstructive pulmonary disease; AGI, acute gastrointestinal injury; ALB, albumin; APACHE II, Acute Physiology and Chronic Health Evaluation II; BMI, body mass index; CRRT, continuous renal replacement therapy; Hb, haemoglobin; mNUTRIC, modified Nutrition Risk in the Critically Ill; NRS‐2002, Nutritional Risk Screening 2002; PAB, prealbumin; RASS, Richmond Agitation–Sedation Scale; SD, standard deviation; SIMV, synchronised intermittent mandatory ventilation; SOFA, Sequential Organ Failure Assessment. Primary Outcome The estimated daily energy target was comparable between the intervention and control groups (1639.6 ± 287.5 vs. 1640.4 ± 285.0 kcal/day; p = 0.991). Delivered enteral energy was significantly higher in the intervention group on Day 3 (1156.4 ± 396.5 vs. 869.0 ± 315.7 kcal/day; p < 0.001) and Day 7 (1369.5 ± 274.5 vs. 1039.4 ± 267.3 kcal/day; p < 0.001) (Table 2). TABLE 2Comparison of outcome indicators between groups (intervention n = 40, control n = 42).VariableIntervention (n = 40)Control (n = 42) t/χ 2 value p Cohen's d Energy target, delivery and attainmentEstimated full daily energy target, kcal/day1639.6 ± 287.51640.4 ± 285.0 t = −0.0120.9910.00Delivered enteral energy on Day 3, kcal/day1156.4 ± 396.5869.0 ± 315.7 t = 3.641 < 0.001 0.80Day 3 EN target attainment (%)70.0 ± 17.553.5 ± 18.7 t = 4.112 < 0.001 0.91Delivered enteral energy on Day 7, kcal/day1369.5 ± 274.51039.4 ± 267.3 t = 5.515 < 0.001 1.22Day 7 EN target attainment (%)83.7 ± 9.464.1 ± 15.4 t = 6.941 < 0.001 1.53Hb (g/L)Baseline (T0)104.2 ± 30.8109.9 ± 27.9 t = −0.8730.3850.19Day 3 (T1)96.1 ± 32.2106.8 ± 27.4 t = −1.6280.1070.36Day 7 (T2)96.4 ± 28.5107.1 ± 27.6 t = −1.7300.0880.38Pre‐transfer (T3)95.1 ± 27.5101.3 ± 29.4 t = −0.9750.3320.22ALB (g/L)Baseline (T0)31.3 ± 5.629.6 ± 5.8 t = 1.4120.1620.31Day 3 (T1)30.5 ± 4.829.6 ± 6.0 t = 0.7740.4410.17Day 7 (T2)33.9 ± 7.129.9 ± 6.6 t = 2.626 0.010 0.58Pre‐transfer (T3)34.3 ± 5.529.4 ± 6.6 t = 3.612 < 0.001 0.80PAB (mg/L)Baseline (T0)105.8 ± 54.697.0 ± 51.0 t = 0.7540.4530.17Day 3 (T1)114.5 ± 54.3109.2 ± 58.1 t = 0.4230.6730.09Day 7 (T2)131.7 ± 67.6114.2 ± 57.8 t = 1.2610.2110.28Pre‐transfer (T3)150.8 ± 70.9122.7 ± 63.0 t = 1.9040.0610.42MV duration (h)245.0 ± 191.2273.8 ± 190.8 t = −0.6820.4970.15ICU length of stay (d)19.1 ± 11.120.8 ± 12.3 t = −0.6440.5220.14Feeding intolerance, n (%)13/40 (32.5)13/42 (31.0) χ 2 = 0.0230.880— Note: Data are presented as mean ± SD or n/N (%). Independent‐samples t‐tests were used for the estimated energy target, delivered enteral energy, EN target attainment, duration of MV and ICU length of stay. The overall incidence of feeding intolerance was compared using the chi‐square test. Hb, ALB and PAB were compared between groups at each time point using independent‐samples t‐tests for descriptive purposes, while longitudinal changes were further analysed using linear mixed‐effects models (see Table 3). Cohen's d represents the absolute standardised mean difference; values of < 0.20, 0.20–0.49, 0.50–0.79 and ≥ 0.80 were interpreted as negligible, small, medium and large effect sizes, respectively. Bold p values indicate statistical significance (p < 0.05).Abbreviations: ALB, albumin; EN, enteral nutrition; Hb, haemoglobin; ICU, intensive care unit; MV, mechanical ventilation; PAB, prealbumin; SD, standard deviation. Comparison of outcome indicators between groups (intervention n = 40, control n = 42). Note: Data are presented as mean ± SD or n/N (%). Independent‐samples t‐tests were used for the estimated energy target, delivered enteral energy, EN target attainment, duration of MV and ICU length of stay. The overall incidence of feeding intolerance was compared using the chi‐square test. Hb, ALB and PAB were compared between groups at each time point using independent‐samples t‐tests for descriptive purposes, while longitudinal changes were further analysed using linear mixed‐effects models (see Table 3). Cohen's d represents the absolute standardised mean difference; values of < 0.20, 0.20–0.49, 0.50–0.79 and ≥ 0.80 were interpreted as negligible, small, medium and large effect sizes, respectively. Bold p values indicate statistical significance (p < 0.05). Abbreviations: ALB, albumin; EN, enteral nutrition; Hb, haemoglobin; ICU, intensive care unit; MV, mechanical ventilation; PAB, prealbumin; SD, standard deviation. EN target attainment was significantly higher in the intervention group on Day 3 (70.0% ± 17.5% vs. 53.5% ± 18.7%; p < 0.001; Cohen's d = 0.91) and Day 7 (83.7% ± 9.4% vs. 64.1% ± 15.4%; p < 0.001; Cohen's d = 1.53). The effect sizes were large at both assessment points (Table 2). In the calendar‐time‐adjusted sensitivity analyses, EN target attainment remained significantly higher in the intervention group on both Day 3 (adjusted mean difference, 19.52 percentage points; 95% CI, 1.04–38.00; p = 0.039) and Day 7 (adjusted mean difference, 19.43 percentage points; 95% CI, 6.40–32.45; p = 0.004). No significant linear calendar‐time trend was observed for Day 3 (−0.24 percentage points per month; 95% CI, −1.57 to 1.08; p = 0.718) or Day 7 (0.01 percentage points per month; 95% CI, −0.92 to 0.95; p = 0.975) (Supporting Information Table S4). Secondary Outcomes Nutrition‐Related Laboratory Indicators At the individual assessment points, ALB was significantly higher in the intervention group than in the control group on Day 7 (33.9 ± 7.1 vs. 29.9 ± 6.6 g/L; p = 0.010) and at the pre‐transfer assessment (34.3 ± 5.5 vs. 29.4 ± 6.6 g/L; p < 0.001). No statistically significant between‐group differences in Hb or PAB were observed at any individual time point (Table 2). In the longitudinal analyses, Hb showed significant main effects of group (p = 0.048) and time (p < 0.001), but no significant group‐by‐time interaction (p = 0.172). For ALB, the group‐by‐time interaction was statistically significant (p = 0.002), whereas the group and time main effects were not significant (p = 0.176 and p = 0.922, respectively). PAB showed a significant time effect (p = 0.004), but neither the group effect nor the group‐by‐time interaction was statistically significant (p = 0.501 and p = 0.123, respectively) (Table 3). TABLE 3Linear mixed‐effects model (repeated measures, Hb/ALB/PAB; baseline T0 as covariate).VariableGroup effect p‐valueTime effect p‐valueGroup × time interaction p‐valueHb 0.048 < 0.001 0.172ALB0.1760.922 0.002 PAB0.501 0.004 0.123 Note: Hb, ALB and PAB were analysed using linear mixed‐effects models with group, time and the group × time interaction as fixed effects and the baseline value (T0) as a covariate. Bold p values indicate statistical significance (p < 0.05).Abbreviations: ALB, albumin; Hb, haemoglobin; PAB, prealbumin. Linear mixed‐effects model (repeated measures, Hb/ALB/PAB; baseline T0 as covariate). Note: Hb, ALB and PAB were analysed using linear mixed‐effects models with group, time and the group × time interaction as fixed effects and the baseline value (T0) as a covariate. Bold p values indicate statistical significance (p < 0.05). Abbreviations: ALB, albumin; Hb, haemoglobin; PAB, prealbumin. Clinical Outcomes No statistically significant between‐group differences were observed in the duration of MV (245.0 ± 191.2 vs. 273.8 ± 190.8 h; p = 0.497) or ICU length of stay (19.1 ± 11.1 vs. 20.8 ± 12.3 days; p = 0.522) (Table 2). Safety Outcome The overall incidence of feeding intolerance did not differ significantly between groups (Table 2), and no significant between‐group differences were observed in any individual feeding intolerance symptom (Table 4). TABLE 4Incidence of feeding intolerance symptoms (intervention n = 40, control n = 42).SymptomIntervention n (%)Control n (%) χ 2 value p Abdominal distension/constipation9 (22.5)13 (31.0) χ 2 = 0.3770.539Gastric residual volume ≥ 500 mL/24 h5 (12.5)8 (19.0) χ 2 = 0.2590.611Diarrhoea (≥ 3 stools/day)4 (10.0)6 (14.3) χ 2 = 0.0650.799Nausea/vomiting3 (7.5)5 (11.9) χ 2 = 0.0900.764Aspiration1 (2.5)2 (4.8)—1.000Gastrointestinal bleeding (occult blood positive)1 (2.5)2 (4.8)—1.000 Note: Data are presented as n (%). Patients could experience more than one symptom; therefore, the symptom categories were not mutually exclusive. The chi‐square test was used for between‐group comparisons, except for aspiration and gastrointestinal bleeding, which were analysed using Fisher's exact test. An em dash (—) indicates not applicable. Incidence of feeding intolerance symptoms (intervention n = 40, control n = 42). Note: Data are presented as n (%). Patients could experience more than one symptom; therefore, the symptom categories were not mutually exclusive. The chi‐square test was used for between‐group comparisons, except for aspiration and gastrointestinal bleeding, which were analysed using Fisher's exact test. An em dash (—) indicates not applicable. Process Outcomes The intervention group initiated EN significantly earlier than the control group (24.1 ± 9.2 vs. 34.5 ± 8.8 h; p < 0.001) and had a significantly higher proportion of patients receiving post‐pyloric feeding (15/40 [37.5%] vs. 3/42 [7.1%]; p = 0.002) (Table 5). In the exploratory correlation analyses, a shorter time to EN initiation was associated with higher EN target attainment on both Day 3 (Spearman's ρ = −0.436, p < 0.001) and Day 7 (ρ = −0.554, p < 0.001). Post‐pyloric feeding was also positively associated with EN target attainment on Day 3 (point‐biserial r = 0.342, p = 0.002) and Day 7 (r = 0.409, p < 0.001). All four correlations remained statistically significant after Benjamini–Hochberg correction (adjusted p ≤ 0.002). In supplementary sensitivity analyses adjusted for study group, the associations of earlier EN initiation and post‐pyloric feeding with higher EN target attainment remained statistically significant at both assessment points, including after Benjamini–Hochberg correction (Supporting Information Table S5). TABLE 5Protocol‐delivery indicators (intervention n = 40, control n = 42).VariableIntervention (n = 40)Control (n = 42)Statistic p EN initiation time (h), mean ± SD24.1 ± 9.234.5 ± 8.8 t = −5.230 < 0.001 Post‐pyloric feeding via NJT, n (%)15 (37.5)3 (7.1) χ 2 = 9.320 0.002 Note: Data are presented as mean ± SD or n (%). Time to EN initiation was compared using the independent‐samples t‐test, and post‐pyloric feeding was compared using the chi‐square test. Both indicators were predefined process outcomes. Bold p values indicate statistical significance (p < 0.05).Abbreviations: EN, enteral nutrition; NJT, nasojejunal tube; SD, standard deviation. Protocol‐delivery indicators (intervention n = 40, control n = 42). Note: Data are presented as mean ± SD or n (%). Time to EN initiation was compared using the independent‐samples t‐test, and post‐pyloric feeding was compared using the chi‐square test. Both indicators were predefined process outcomes. Bold p values indicate statistical significance (p < 0.05). Abbreviations: EN, enteral nutrition; NJT, nasojejunal tube; SD, standard deviation. Discussion This study was guided by the MRC framework for complex interventions. Through evidence synthesis, logic model development and expert consultation, we developed an integrated early EN management protocol combining conventional care with TCM‐based supportive approaches for mechanically ventilated ICU patients. We then conducted a preliminary quasi‐experimental evaluation of the protocol. EN target attainment was significantly higher in the intervention group than in the control group on both Day 3 and Day 7, with large effect sizes (Cohen's d = 0.91 and 1.53, respectively). These between‐group differences remained statistically significant after adjustment for a linear calendar‐time trend, indicating that the primary findings were robust to this adjustment. Serum ALB levels were also higher in the intervention group on Day 7 and before transfer from the ICU, whereas the incidence of feeding intolerance did not differ significantly between groups. Taken together, these findings provide preliminary evidence that implementation of the integrated protocol was associated with higher early EN delivery, without a detected increase in feeding intolerance. Interpretation of Intervention Pathways From the Perspective of the MRC Framework From the perspective of the MRC framework for complex interventions, the logic model developed in this study prespecified three potential pathways: earlier initiation of EN, optimisation of post‐pyloric feeding and adjunctive TCM‐based supportive interventions. The intervention group initiated EN significantly earlier than the control group, and all patients in the intervention group achieved EN initiation within 48 h after ICU admission. This finding indicates a high level of adherence to the protocol's early‐initiation target. The exploratory analyses further showed that a shorter time to EN initiation was associated with higher EN target attainment at both assessment points, and these associations remained statistically significant after adjustment for study group. These findings are consistent with the prespecified logic model and suggest that the observed associations were not explained solely by pooled between‐group differences. Nevertheless, they do not establish a causal or mediating pathway. The present results are also broadly consistent with previous ICU feeding protocol studies and nurse‐led feeding pathway interventions, which have shown that protocolised management can facilitate earlier EN initiation and improve nutritional delivery. For example, a before‐and‐after study conducted in Estonia by Padar et al. [30] showed that a nurse‐driven feeding protocol improved EN delivery in ICU patients without increasing gastrointestinal complications. Similarly, a study conducted in South Korea by Kim et al. [31] reported earlier EN initiation and improved feeding practices following implementation of an EN protocol. With respect to optimisation of the feeding route, the proportion of post‐pyloric feeding was significantly higher in the intervention group than in the control group. Previous studies have shown that post‐pyloric feeding may reduce gastric retention and feeding interruptions and thereby improve actual EN delivery [32, 33]. Exploratory analyses in the present study showed that post‐pyloric feeding was positively associated with EN target attainment on both Days 3 and 7, and these associations remained statistically significant after adjustment for study group. The observed pattern accords with the prespecified intervention pathway and previous reports that optimisation of the feeding route may improve nutritional delivery and reduce feeding interruptions [30, 32, 33]. However, because the feeding route was selected according to clinical indications rather than independently assigned, the observed associations should not be interpreted as evidence of a causal pathway or an isolated component effect. Regarding the TCM‐related pathway, previous studies have suggested that ST36 stimulation, auricular acupressure and related gastrointestinal regulatory techniques may influence gastrointestinal motility through autonomic regulation and modulation of gastrointestinal hormone secretion [10, 21, 34, 35, 36, 37]. This evidence provided the rationale for incorporating TCM‐based techniques into the protocol as supportive components intended to address gastrointestinal symptoms and potential barriers to EN delivery. These techniques were delivered according to patients' gastrointestinal symptoms and clinical indications, alongside conventional measures such as adjustments to the feeding rate or formula and the use of prokinetic agents when required. However, no statistically significant between‐group differences were observed in the overall incidence of feeding intolerance or in any individual feeding intolerance outcome. Therefore, the higher EN target attainment observed in the intervention group cannot be attributed to improved gastrointestinal tolerance. Because the TCM‐based techniques were not evaluated independently, their specific contribution to EN delivery and their potential mechanisms could not be established. The findings should therefore be interpreted as reflecting the overall effect of the integrated multicomponent protocol rather than the effect of any individual component. International implementation of the TCM‐based components would require context‐specific adaptation. Although these techniques are relatively low‐technology bedside procedures that may be standardised through competency‐based training, their feasibility and acceptability outside China cannot be assumed. Transferability may depend on local scope‐of‐practice and regulatory requirements, staff familiarity and training needs, the availability and standardisation of intervention materials, and the cultural preferences of patients and healthcare professionals. Future implementation in other settings should therefore involve consultation with local stakeholders, culturally appropriate communication and pilot evaluation of acceptability, intervention fidelity and safety. Consistent with the MRC framework, local adaptations should preserve the intended supportive functions of the components while allowing their mode of delivery to reflect the local clinical and cultural context [12]. Interpretation of Nutrition‐Related Laboratory Indicators Alongside the higher EN delivery observed in the intervention group, ALB levels were significantly higher on Day 7 and before transfer from the ICU. The linear mixed‐effects model also showed a significant group‐by‐time interaction for ALB, indicating that the pattern of change in ALB differed between groups during follow‐up. Although these findings were temporally aligned with higher EN delivery, the present study design does not establish that improved nutritional delivery caused the observed differences in ALB. These laboratory indicators should be interpreted cautiously because ALB and PAB are influenced by inflammation, fluid distribution and overall disease recovery, in addition to nutritional intake [38, 39]. PAB showed a significant time effect but no significant group effect or group‐by‐time interaction, indicating change over time without evidence of different temporal patterns between groups. For Hb, significant group and time main effects were observed, whereas the group‐by‐time interaction was not significant. The group effect indicated a lower adjusted mean Hb level across follow‐up assessments in the intervention group. However, there was no evidence that the temporal pattern of Hb differed between groups, and none of the between‐group comparisons at individual assessment points reached statistical significance. The Hb findings should therefore not be interpreted as evidence of an intervention‐related difference in Hb recovery. Hb is also influenced by haemodilution, phlebotomy‐related blood loss and transfusion practices in critically ill patients [40, 41, 42]. Among the three laboratory indicators, ALB showed the clearest longitudinal between‐group pattern. Nevertheless, all three indicators should be regarded as supportive and nonspecific rather than definitive markers of nutritional efficacy. Interpretation of Clinical Outcomes In the present study, neither the duration of MV nor ICU length of stay differed significantly between groups. These findings are generally consistent with evidence from previous randomised trials of acute‐phase nutritional support conducted in Saudi Arabia, Canada, Australia and New Zealand, and the United States [43, 44, 45]. For example, the US multicentre EDEN trial showed that differences in short‐term nutritional delivery did not necessarily translate into improvements in ventilator‐related outcomes, mortality or infectious complications [45]. Thus, although nutritional delivery may be improved during the acute‐phase, corresponding changes in complex clinical outcomes such as MV duration and ICU length of stay may not be detectable. This interpretation is also compatible with the design of the present study. Our sample size calculation was based on EN target attainment as the primary outcome. In contrast, MV duration and ICU stay vary substantially and are influenced by disease severity, infectious complications, trajectories of organ dysfunction and weaning strategies. Therefore, the current sample size may have been insufficient to detect significant differences in these more complex clinical outcomes. Safety Evaluation The overall incidence of feeding intolerance did not differ significantly between the intervention and control groups, and no significant between‐group differences were observed in any individual intolerance symptom. Thus, no increase in feeding intolerance was detected during the study. This finding is also broadly consistent with previous studies on protocolised EN management. Padar et al. showed that a nurse‐driven feeding protocol improved nutritional delivery without increasing gastrointestinal symptoms or intra‐abdominal pressure problems [30]. Likewise, the EDEN trial suggested that although tolerance profiles may differ between feeding strategies, such differences do not necessarily translate into worse infection or mortality outcomes [45]. The incidence of aspiration was low in both groups, which may be related to the widespread implementation of head‐of‐bed elevation and continuous pump feeding in routine practice. One serious adverse event, multiple organ dysfunction syndrome, occurred during the study period; however, it was judged not to be causally related to the intervention. Overall, no short‐term safety signal related to the integrated management protocol was identified during the study. Study Limitations Several limitations should be acknowledged. First, this was a quasi‐experimental before‐and‐after study. The two groups were broadly similar with respect to the measured baseline characteristics, and the calendar‐time‐adjusted sensitivity analyses yielded results consistent with the primary analyses. However, the control and intervention phases did not overlap in calendar time. Adjustment for a linear calendar‐time trend therefore could not fully account for nonlinear temporal changes or unmeasured concurrent changes in staff competency, the implementation environment and clinical practice. Residual temporal confounding cannot be excluded. Second, this was a single‐centre study with a relatively limited sample size. The sample size calculation assumed an absolute increase in EN target attainment from 50% to 80%. However, the observed between‐group differences were smaller, at 16.5 percentage points on Day 3 and 19.6 percentage points on Day 7, indicating that the effect assumed for sample size planning was optimistic. Although the primary outcome differed significantly between groups at both assessment points, the study was not designed to provide adequate statistical power for secondary outcomes, and the limited sample size precluded reliable subgroup analyses. Accordingly, non‐significant secondary findings should be interpreted cautiously and should not be regarded as evidence of equivalence or the absence of a between‐group difference. Third, the intervention consisted of five modules and multiple components, representing a typical complex intervention; therefore, the current design does not allow the independent contribution of each individual component to be disentangled. Future studies may consider factorial designs or phased implementation strategies to better identify the key active components. Fourth, some laboratory indicators may have been influenced by inflammatory status, fluid shifts and concurrent clinical management. In addition, the study did not include a more comprehensive panel of inflammatory markers or long‐term follow‐up outcomes, which limits the interpretation of mechanisms and longer‐term effects. Conclusion Implementation of an integrated early EN management protocol developed within the MRC framework was associated with higher EN target attainment on study Days 3 and 7 in mechanically ventilated ICU patients. ALB showed a significant group‐by‐time interaction, whereas MV duration, ICU length of stay and feeding intolerance did not differ significantly between groups. As a preliminary evaluation of a multicomponent complex intervention, this study suggests that the protocol was feasible to implement within the study setting. The findings provide a basis for context‐sensitive adaptation and further evaluation in multicentre randomised controlled trials. Funding This study was supported by a research grant from Jiangsu Province Hospital of Chinese Medicine (Y23024); Project title: Construction and Application of an Integrated Traditional Chinese and Western Medicine Management Protocol for Early Enteral Nutrition in Mechanically Ventilated ICU Patients. Ethics Statement This study was approved by the Medical Ethics Committee of Jiangsu Province Hospital of Chinese Medicine (Affiliated Hospital of Nanjing University of Chinese Medicine) on 17 October 2023 (Approval No. 2023NL‐234‐02) and was conducted in accordance with the Declaration of Helsinki. Consent Written informed consent was obtained from all participants or their legally authorised representatives prior to enrolment. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Table S1: Summary of the modules, core components and intended functions of the integrated early enteral nutrition management protocol. Table S2: Standardised procedures for TCM‐based supportive interventions. Table S3: Component‐level delivery and fidelity of TCM‐based supportive interventions in the intervention group (n = 40). Table S4: Sensitivity analyses of EN target attainment adjusted for calendar time. Table S5: Group‐adjusted associations of process outcomes with EN target attainment. Acknowledgements During the preparation of this manuscript, the authors used OpenAI GPT‐5 to assist with grammar correction, sentence‐structure refinement and the translation of selected phrases. All AI‐assisted text was critically reviewed, verified and revised by the authors. All authors approved the final manuscript and take full responsibility for its content. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. References Associated Data Supplementary Materials Table S1: Summary of the modules, core components and intended functions of the integrated early enteral nutrition management protocol. Table S2: Standardised procedures for TCM‐based supportive interventions. Table S3: Component‐level delivery and fidelity of TCM‐based supportive interventions in the intervention group (n = 40). Table S4: Sensitivity analyses of EN target attainment adjusted for calendar time. Table S5: Group‐adjusted associations of process outcomes with EN target attainment. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.