Nurse-led family participatory support intervention on clinical outcomes in ICU patients: a prospective quasi-experimental study Authors: Yuchen Wu, Guoqiang Wang, Kejing Mu, Han Ruan, Huaping Wei, Weigang Yue, Junfeng Yang, Fangli Ma, Ting Chen, Zhigang Zhang, Xinghua Lv, Xin Wang Published: 2026 ABSTRACT BackgroundDelirium is a common and serious complication in critically ill patients, and family-centered care has emerged as a promising non-pharmacological strategy. However, evidence on structured, nurse-led family participatory support (NFPS) intervention remains limited.AimTo evaluate the effectiveness and safety of NFPS in critically ill adult patients.MethodsAdult patients admitted to intensive care units (ICU) from three tertiary hospitals in Gansu Province were allocated to either the NFPS group or a usual care group without family participation (non-NFPS). In the NFPS group, trained family members participated in care under the guidance of ICU nurses. The primary outcome was the incidence of delirium.ResultsAfter 1:1 propensity score matching, 365 patients were included in each group. The NFPS group demonstrated significant reductions in incidence of delirium [26.85% vs. 34.25%, p = 0.03]. Compared with the control group, the NFPS group also had significantly shorter hospital length of stay (HLOS) (25 [17–34] vs. 30 [22–38] days, P < 0.001), ICU-LOS (13 [7–17] vs. 14 [10–18] days, P < 0.001), duration of mechanical ventilation (10 [6–15] vs. 11 [8–17] days, P < 0.001) and lower hospitalization cost (193,931.99 [117,616.64–294,201.02] vs. 199,330.72 [138,636.05–300,993.88] CNY, P = 0.032). No significant differences were observed in 6-month mortality (6.03% vs. 7.67%, P = 0.380), 1-year mortality (9.32% vs. 11.51%, P = 0.330), ICU-acquired weakness (26.58% vs. 29.59%, P = 0.270), ICU-acquired infection (19.45% vs. 19.73%, P = 0.930), or adverse events (2.19% vs. 3.84%, P = 0.190).ConclusionThe NFPS significantly reduced the incidence of delirium and was associated with shorter durations of mechanical ventilation, ICU-LOS, HLOS, lower hospitalization costs, without increasing the risks of ICU-AW, ICU-acquired infection, or adverse events. No significant differences were observed in long-term mortality between the two groups.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s13054-026-06145-z. Introduction Critically ill patients are at high risk of a range of ICU-related complications, including delirium, ICU-acquired weakness (ICU-AW), long-term psychological and cognitive impairment. These complications are associated with prolonged mechanical ventilation (MV), extended length of stay (LOS), higher healthcare costs, and increased mortality [1–3]. In particular, delirium is a frequent and serious complication in the ICU [2]. Studies have shown that delirium is associated with hospitalization costs 2.5 times higher and a mortality rate 2.9 times higher than patients without delirium [4, 5]. Moreover, delirium is a major risk factor for long-term cognitive impairment among ICU survivors [6]. Early mobilization (EM) has been shown reduce the incidence of delirium, shorten the duration of MV, improve weaning success, and decrease ICU-LOS [2, 7, 8]. However, Patel BK. et al. reported that EM did not improve the quality of life during the 6-month follow-up, and this may be related to delirium and cognitive dysfunction of patients during hospitalization [1]. These findings highlight that EM alone may be insufficient to address the complex neuropsychological needs of ICU patients. Family-centered care has emerged as a promising complementary approach in critical care [9–11]. Family members can provide emotional support, enhance patient orientation, and help alleviate anxiety and stress, which are key contributors to delirium [10, 11]. Increasingly, ICU care models are shifting toward more open visitation policies, encouraging family involvement in daily care and rehabilitation processes, including early mobilization [12–14]. Despite growing interest, high-quality evidence regarding structured, nurse-led family participation interventions remains limited. In particular, it is unclear whether involving family members in a standardized, nurse-led participatory support model can improve clinical outcomes without compromising patient safety [11, 15]. To address this gap, we developed a Nurse-Led Family Participatory Support (NFPS) and conducted a quasi-experimental study to evaluate the effectiveness and safety of NFPS in critically ill patients. Participants and methods Study design This prospective, multicenter, quasi-experimental study following the STROBE Checklists [16] and received approval from the Ethics Committee of the LZU No. 1 Hospital (No. LDYYLL2026-586). All participants provided written informed consent, which was obtained by the study interventionists and overseen by the principal investigator. The study adhered to the principles of the Declaration of Helsinki. Participant privacy was protected through coding of all data. Participants Participants were prospectively enrolled from three adult ICUs in Gansu Province using a time-sequence design. Patients admitted between March 1, 2023, and February 28, 2024 were assigned to the non-NFPS group, while those admitted between April 1, 2024, and March 31, 2025 were assigned to the NFPS group. Recruitment was conducted through poster displays and verbal communication. Enrollment continued until the predetermined sample size was reached. All participants were followed up until April 2026. Inclusion criteria: Ⅰ. aged ≥ 18 years; Ⅱ. duration of MV ≥ 48 h; Ⅲ. Glasgow Coma Scale (GCS) score ≥ 11 points; Ⅳ. Provision of voluntary written informed consent by the patient or legal surrogate. Exclusion criteria: Ⅰ. Hemodynamic instability; Ⅱ. Increased intracranial pressure (≥ 20 mmHg); Ⅲ. Active bleeding; Ⅳ. Unstable fracture, limb mutilation or limb movement dysfunction; Ⅴ. Pregnant; Ⅵ. Aortic artery disease with tumor rupture or leakage; Ⅶ. Participation in similar rehabilitation research by the patient. Intervention methods Patients in the non-NFPS group received standard ICU care, including routine nursing and early rehabilitation practices. Standard care consisted of repositioning and back percussion every 2 h, limb positioning in functional alignment, and functional exercises performed solely by ICU rehabilitation therapists according to the ICU early functional exercise protocol (30 min/session, twice daily). Daily living care was provided by nursing staffs. Family members were permitted only 30 min of daily bedside visitation for emotional support. They were not allowed to perform any physical care or rehabilitation activities. Patients in the NFPS group received standard care in addition to the NFPS intervention. The NFPS is a structured, nurse-led care model designed to facilitate safe and systematic family involvement in the care of critically ill patients. The intervention integrates standardized training, nurse-led participation, daily assessment, and dynamic feedback to improve patient outcomes. After the patient is admitted to the ICU, researchers will screen eligible patients and their family members and obtain informed consent. Within 48 h, a standardized training session will be provided to the family members by an advanced practice nurse, covering an introduction to the ICU environment, delirium awareness, participatory care strategies, and safety precautions. Following training, family members were encouraged to participate in patient care under the supervision of ICU nurses in accordance with a predefined protocol. The core components of the intervention included orientation support, emotional support, cognitive stimulation, and early rehabilitation assistance. All activities were non-invasive and conducted in strict compliance with ICU safety regulations. Patients were assessed daily by experienced nursing researchers, who documented clinical status and intervention adherence. The attending physician dynamically adjusted treatment strategies based on the patient’s condition. After discharge, patients were followed up to evaluate long-term outcomes. Infection control and safety measures are strictly enforced throughout the entire intervention process. Criteria for suspending or terminating NFPS were predefined based on established safety recommendations [11, 17], taking into account potential patient instability during activity. The detailed description of the intervention protocol is provided in Fig. 1 and Appendix A. During the entire study period, no changes were made to institutional ICU protocols regarding sedation, mechanical ventilation weaning, routine early mobilization, infection control, or delirium assessment. Staffing and training protocols remained consistent across both phases. Fig. 1Nurse-Led Family Participatory Support Intervention Protocol and mechanisms Nurse-Led Family Participatory Support Intervention Protocol and mechanisms Sample size Sample size calculation used PASS 15 software (NCSS, LLC. Kaysville, Utah, USA). Based on the incidence of delirium which was defined as the primary outcome of this study [2], the expected incidence of delirium was assumed to be 44.02% in the non-NFPS group and 31.38% in the NFPS group. A two-sided hypothesis test was applied to detect the difference in proportions between the two groups. The significance level was set at α = 0.05, and the statistical power was set at 90% (1 − β = 0.90). The Z test for two independent proportions was used, with equal group allocation (N1 = N2). Based on these parameters, the required sample size was calculated to be 304 participants per group. Considering a potential 20% attrition or missing data rate, the final target sample size was adjusted to 380. During the actual recruitment period, 365 participants were enrolled in each group, resulting in a total sample size of 730 participants, which was close to the predetermined target sample size and sufficient to maintain the desired statistical power. Outcomes and assessment tool Demographic information of the patients was extracted from the hospital information system (HIS). Delirium was assessed daily by the patient’s responsible nurse using the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) [18], and the results were documented in the electronic medical record. Research nurses subsequently extracted these documented outcomes from the medical records. Other outcomes, including ICU-acquired weakness (ICU-AW), duration of mechanical ventilation, ICU-LOS, adverse events, ICU-acquired infections, and hospitalization costs, were also extracted from medical records. ICU-AW was diagnosed using the Medical Research Council (MRC) score scale, with a score < 48 indicating ICU-AW [3, 11]. Mortality was defined as the proportion of participants who died within 6‑month or 1 year after enrollment, calculated as the number of deaths divided by the total number of enrolled participants [19]. For patients discharged from the ICU or hospital before the 6‑month or 1‑year time point, survival status was determined by structured telephone follow-up conducted by trained research nurses at 6 months (± 2 weeks) and 12 months (± 2 weeks). If the patient was unavailable, information was obtained from a pre‑identified family contact. Death was confirmed by the family member or by hospital readmission records when applicable. ICU-acquired infection defined as an infection occurring > 48 h after ICU admission, attributable to ICU-specific procedures or pathogens. This included ventilator-associated pneumonia, catheter-associated bloodstream infections, and catheter-related infections, urinary tract infection, etc. Diagnosis required confirmatory laboratory tests of sputum, blood cultures, catheter tip cultures, and urine [12]. Statistical analysis All data were double-entered and cross-checked using EpiData version 3.1 software. Statistical analyses were conducted using IBM SPSS Statistics version 22.0 (IBM Corp., Armonk, NY, USA). Descriptive statistics were used to summarize baseline characteristics and outcome variables. Categorical variables are presented as frequencies and percentages, and comparisons between the two groups were performed using the Chi-square test or Fisher’s exact test, as appropriate. For continuous variables, the distribution of the data was assessed using the Shapiro-Wilk test for normality. Normally distributed variables are reported as mean ± standard deviation (SD) and were compared between groups using the independent samples t-test. Variables that were not normally distributed are presented as median with interquartile range (IQR) and were compared using the Mann-Whitney U test. To reduce potential selection bias and control for baseline differences, propensity score matching (PSM) was performed. Propensity scores were estimated using a logistic regression model. Variables included in the model were selected based on clinical relevance and previous literature, including age, sex, smoking history, Charlson Comorbidity Index (CCI), and SOFA score. Patients in the two groups were matched using 1:1 nearest-neighbor matching without replacement, with a caliper width of 0.02 of the standard deviation of the logit of the propensity score. After matching, the balance of baseline variables between the two groups was assessed. After PSM, clinical outcomes were compared between the two groups. Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated for categorical outcomes. Cases with missing key outcome variables or incomplete observation indicators were excluded from the final statistical analysis to ensure the reliability of the results. Missing data were assessed prior to analysis. If the proportion of missing values was less than 5%, a complete-case analysis was performed. Results Participant characteristics and propensity score matching A total of 1309 patients were recruited in this trial, of which 436 patients were included in the non-NFPS group and 460 patients in the NFPS group. After 1:1 PSM, 365 patients were included in each group. Before matching, significant differences were observed between the two groups in several baseline characteristics, including alcohol use, education level, occupation, body mass index (BMI), and SOFA score (all \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:P<\:$$\end{document}0.05). After matching, 365 pairs of patients were successfully generated, and the baseline demographic and clinical characteristics were well balanced between the two groups (all P > 0.05). The flow Diagram of participant enrollment, allocation, follow-up, and analysis is shown in Fig. 2. The baseline demographic and clinical characteristics of the study population before and after propensity score matching are presented in Tables 1 and 2. Fig. 2Flow Diagram of participant enrollment, allocation, follow-up, and analysis Flow Diagram of participant enrollment, allocation, follow-up, and analysis (NFPS: nurse-led family participatory support; non-NFPS: without NFPS). Table 1Baseline characteristics before and after PSM (Continuous variables)Demographic characteristicsBefore PSMAfter PSMnon-NFPS (n = 436)NFPS (n = 460)Z P non-NFPS (n = 365)NFPS (n = 365)Z P Age 50.24 ± 13.9951.49 ± 15.1991.5980.1149.87 ± 14.0950.20 ± 15.210.380.70 CCI 3.12 ± 1.612.97 ± 1.58−1.2580.2083.10 ± 1.623.12 ± 1.610.210.83 APACHE Ⅱ 14.33 ± 5.3914.35 ± 5.210.1660.86814.60 ± 5.2614.29 ± 5.160.300.77 SOFA 3.82 ± 1.843.53 ± 1.752.4060.0163.68 ± 1.833.75 ± 1.710.940.35 BMI 23.66 ± 3.1822.88 ± 3.13−3.881<0.000123.395 ± 3.2323.28 ± 3.120.570.57PSM: propensity score matching. NFPS: nurse-led family participatory support. non-NFPS: without NFPS. APACHE Ⅱ: Acute Physiology and Chronic Health Evaluation. SOFA: Sequential Organ Failure Assessment. CCI: Comorbidity Index. BMI: Body mass index Table 1Baseline characteristics before and after PSM (Continuous variables)Demographic characteristicsBefore PSMAfter PSMnon-NFPS (n = 436)NFPS (n = 460)Z P non-NFPS (n = 365)NFPS (n = 365)Z P Age 50.24 ± 13.9951.49 ± 15.1991.5980.1149.87 ± 14.0950.20 ± 15.210.380.70 CCI 3.12 ± 1.612.97 ± 1.58−1.2580.2083.10 ± 1.623.12 ± 1.610.210.83 APACHE Ⅱ 14.33 ± 5.3914.35 ± 5.210.1660.86814.60 ± 5.2614.29 ± 5.160.300.77 SOFA 3.82 ± 1.843.53 ± 1.752.4060.0163.68 ± 1.833.75 ± 1.710.940.35 BMI 23.66 ± 3.1822.88 ± 3.13−3.881<0.000123.395 ± 3.2323.28 ± 3.120.570.57PSM: propensity score matching. NFPS: nurse-led family participatory support. non-NFPS: without NFPS. APACHE Ⅱ: Acute Physiology and Chronic Health Evaluation. SOFA: Sequential Organ Failure Assessment. CCI: Comorbidity Index. BMI: Body mass index Baseline characteristics before and after PSM (Continuous variables) PSM: propensity score matching. NFPS: nurse-led family participatory support. non-NFPS: without NFPS. APACHE Ⅱ: Acute Physiology and Chronic Health Evaluation. SOFA: Sequential Organ Failure Assessment. CCI: Comorbidity Index. BMI: Body mass index Table 2Baseline characteristics before and after PSM (Categorical variables)Demographic characteristicsBefore PSMAfter PSM non-NFPS (n = 436) NFPS (n = 460) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{x}^{2}$$\end{document} P non-NFPS (n = 365) NFPS (n = 365) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{x}}^{2}$$\end{document} P Hospitals ICU270(61.93%)268(58.26%)1.6380.441226(61.92%)214(58.63%)3.070.22N/RICU57(13.07%)72(15.65%)47(12.88%)64(17.53%)EICU109(25%)120(26.09%)92(25.21%)87(23.84%) Sex Male274(62.84%)297(64.57%)0.2870.592233(63.84%)229(62.74%)0.090.76Female162(37.16%)163(35.43%)132(36.16%)136(37.26%) Smoking history Yes137(31.42%)293(63.7%)2.380.123128(35.07%)125(34.25%)0.050.82No299(68.58%)167(36.3%)237(64.93%)240(65.75%) Alcohol use Yes83(19.04%)113(24.57%)4.0030.045179(49.04%)194(53.15%)1.230.27No353(80.96%)347(75.43%)186(50.96%)171(46.85%) Marital status Married347(79.59%)364(79.13%)0.2870.866294(80.55%)297(81.37%)0.100.95Single57(13.07%)58(12.61%)43(11.78%)42(11.51%)Divorced/Widowed32(7.34%)38(8.26%)28(7.67%)26(7.12%) Education Primary school211(48.39%)219(47.61%)8.8310.032175(47.95%)181(49.59%)4.480.21Junior/senior high school142(32.57%)131(28.48%)114(31.23%)101(27.67%)Associate/bachelor degree44(10.09%)76(16.52%)42(11.51%)58(15.89%)Master or above39(8.94%)34(7.39%)31(8.49%)25(6.85%) Occupation Worker34(7.8%)34(7.39%)15.8560.00733(9.04%)24(6.58%)14.170.02Farmer203(46.56%)248(53.91%)182(49.86%)185(50.68%)Office worker43(9.86%)42(9.13%)37(10.14%)29(7.95%)Unemployed31(7.11%)50(10.87%)20(5.48%)46(12.6%)Retired87(19.95%)63(13.7%)66(18.08%)62(16.99%)Others38(8.72%)23(5%)27(7.4%)19(5.21%)PSM: propensity score matching. NFPS: nurse-led family participatory support. non-NFPS: without NFPS Table 2Baseline characteristics before and after PSM (Categorical variables)Demographic characteristicsBefore PSMAfter PSM non-NFPS (n = 436) NFPS (n = 460) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{x}^{2}$$\end{document} P non-NFPS (n = 365) NFPS (n = 365) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{x}}^{2}$$\end{document} P Hospitals ICU270(61.93%)268(58.26%)1.6380.441226(61.92%)214(58.63%)3.070.22N/RICU57(13.07%)72(15.65%)47(12.88%)64(17.53%)EICU109(25%)120(26.09%)92(25.21%)87(23.84%) Sex Male274(62.84%)297(64.57%)0.2870.592233(63.84%)229(62.74%)0.090.76Female162(37.16%)163(35.43%)132(36.16%)136(37.26%) Smoking history Yes137(31.42%)293(63.7%)2.380.123128(35.07%)125(34.25%)0.050.82No299(68.58%)167(36.3%)237(64.93%)240(65.75%) Alcohol use Yes83(19.04%)113(24.57%)4.0030.045179(49.04%)194(53.15%)1.230.27No353(80.96%)347(75.43%)186(50.96%)171(46.85%) Marital status Married347(79.59%)364(79.13%)0.2870.866294(80.55%)297(81.37%)0.100.95Single57(13.07%)58(12.61%)43(11.78%)42(11.51%)Divorced/Widowed32(7.34%)38(8.26%)28(7.67%)26(7.12%) Education Primary school211(48.39%)219(47.61%)8.8310.032175(47.95%)181(49.59%)4.480.21Junior/senior high school142(32.57%)131(28.48%)114(31.23%)101(27.67%)Associate/bachelor degree44(10.09%)76(16.52%)42(11.51%)58(15.89%)Master or above39(8.94%)34(7.39%)31(8.49%)25(6.85%) Occupation Worker34(7.8%)34(7.39%)15.8560.00733(9.04%)24(6.58%)14.170.02Farmer203(46.56%)248(53.91%)182(49.86%)185(50.68%)Office worker43(9.86%)42(9.13%)37(10.14%)29(7.95%)Unemployed31(7.11%)50(10.87%)20(5.48%)46(12.6%)Retired87(19.95%)63(13.7%)66(18.08%)62(16.99%)Others38(8.72%)23(5%)27(7.4%)19(5.21%)PSM: propensity score matching. NFPS: nurse-led family participatory support. non-NFPS: without NFPS Baseline characteristics before and after PSM (Categorical variables) PSM: propensity score matching. NFPS: nurse-led family participatory support. non-NFPS: without NFPS Comparison of clinical outcomes between the two groups The incidence of delirium was significantly lower in the NFPS group than in the non-NFPS group (26.85% vs. 34.25%, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{x}^{2}$$\end{document}= 4.707, P = 0.030; OR = 1.419, 95% CI 1.034–1.948). No significant differences were observed between the two groups in 6-month mortality (6.03% vs. 7.67%, P = 0.380), 1-year mortality (9.32% vs. 11.51%, P = 0.330), ICU-AW (26.58% vs. 29.59%P = 0.270), ICU-acquired infections (19.45% vs. 19.73%, P = 0.930), or adverse events (2.19% vs. 3.84% P = 0.190). In the NFPS group, a total of 8 patients (2.19%) experienced adverse events, including 4 cases of unplanned extubation (UEX), 2 cases of reintubation within 48 h, and 2 cases of ICU readmission within 48 h. In the non-NFPS group, 14 patients (3.84%) experienced adverse events, including 5 cases of UEX, 3 cases of reintubation within 48 h, 4 cases of ICU readmission within 48 h, and 2 cases of falls during rehabilitation (Table 3). Table 3Comparison of clinical outcomes between the two groups after PSM (n = 365)Outcomesnon-NFPS/n(%)NFPS/n(%) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{x}}^{2}$$\end{document} P OR (95%CI) Delirium Yes125(34.25%)98(26.85%)4.7070.031.419 (1.034, 1.948)No240(65.75%)267(73.15%) 6-month mortality Death28(7.67%)22(6.03%)0.7730.381.295 (0.727, 2.31)Survival337(92.33%)343(93.97%) 1-year mortality Death42(11.51%)34(9.32%)0.940.331.266 (0.785, 2.041)Survival323(88.49%)331(90.68%) ICU-AW Yes108(29.59%)97(26.58%)0.8210.271.161 (0.84, 1.604)No257(70.41%)268(73.42%) ICU-acquired infection Yes72(19.73%)71(19.45%)0.0090.931.018 (0.706, 1.467)No293(80.27%)294(80.55%) Adverse events Yes14(3.84%)8(2.19%)1.6870.191.78 (0.737, 4.296)No351(96.16%)357(97.81%)The non‑NFPS group served as the reference category for all odds ratios Table 3Comparison of clinical outcomes between the two groups after PSM (n = 365)Outcomesnon-NFPS/n(%)NFPS/n(%) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{x}}^{2}$$\end{document} P OR (95%CI) Delirium Yes125(34.25%)98(26.85%)4.7070.031.419 (1.034, 1.948)No240(65.75%)267(73.15%) 6-month mortality Death28(7.67%)22(6.03%)0.7730.381.295 (0.727, 2.31)Survival337(92.33%)343(93.97%) 1-year mortality Death42(11.51%)34(9.32%)0.940.331.266 (0.785, 2.041)Survival323(88.49%)331(90.68%) ICU-AW Yes108(29.59%)97(26.58%)0.8210.271.161 (0.84, 1.604)No257(70.41%)268(73.42%) ICU-acquired infection Yes72(19.73%)71(19.45%)0.0090.931.018 (0.706, 1.467)No293(80.27%)294(80.55%) Adverse events Yes14(3.84%)8(2.19%)1.6870.191.78 (0.737, 4.296)No351(96.16%)357(97.81%)The non‑NFPS group served as the reference category for all odds ratios Comparison of clinical outcomes between the two groups after PSM (n = 365) The non‑NFPS group served as the reference category for all odds ratios Comparison of hospitalization outcomes Hospitalization costs were significantly lower in the NFPS group than in the non-NFPS group (193,931.99 [117,616.64–294,201.02] vs. 199,330.72 [138,636.05–300,993.88] CNY, Z = − 2.14, P = 0.032). Patients in the NFPS group had a significantly shorter HLOS (25 [17–34] vs. 30 [22–38] days, Z = − 6.662, P < 0.001), ICULOS (13 [7–17] vs. 14 [10–18] days, Z = − 4.711, P < 0.001) and duration of mechanical ventilation (10 [6–15] vs. 11 [8–17] days, Z = − 4.663, P < 0.001) than those in the non-NFPS group, as shown in Table 4. Table 4Comparison of hospitalization outcomesOutcomesnon-NFPS group/Median (IQR)NFPS group/Median (IQR)Z P Hospitalization costs (CNY)199330.72(138636.05, 300993.88)193931.99(117616.64, 294201.02)−2.140.032Hospital LOS(days)30 (22, 38)25 (17, 34)−6.662<0.001ICU LOS(days)14 (10, 18)13 (7, 17)−4.711<0.001Duration of MV(days)11(8, 17)10(6, 15)−4.663<0.001 Table 4Comparison of hospitalization outcomesOutcomesnon-NFPS group/Median (IQR)NFPS group/Median (IQR)Z P Hospitalization costs (CNY)199330.72(138636.05, 300993.88)193931.99(117616.64, 294201.02)−2.140.032Hospital LOS(days)30 (22, 38)25 (17, 34)−6.662<0.001ICU LOS(days)14 (10, 18)13 (7, 17)−4.711<0.001Duration of MV(days)11(8, 17)10(6, 15)−4.663<0.001 Comparison of hospitalization outcomes 199330.72 (138636.05, 300993.88) 193931.99 (117616.64, 294201.02) Hospital LOS (days) ICU LOS (days) Duration of MV (days) Discussion In this prospective quasi-experimental study with propensity score matching, we evaluated the clinical effectiveness and safety of the NFPS in critically ill patients. Our primary findings were that the NFPS significantly reduced the incidence of delirium, shortened the duration of mechanical ventilation, decreased ICU and hospital length of stay, and lowered hospitalization costs, without increasing the risks of ICU‑AW, ICU-acquired infection, or adverse events. No significant differences were observed in 6‑month or 1‑year mortality between the two groups. These findings suggest that the NFPS is a safe and effective strategy for improving short-term outcomes in critically ill patients. The incidence (26.85% vs. 34.25%) of delirium in the NFPS group was lower than that in the non-NFPS group, and this result is consistent with accumulating evidence supporting family involvement as an effective non-pharmacological strategy for delirium prevention [10, 20]. Mitchell et al. found that ICU patients without family participation in critical care had a threefold higher incidence of delirium and significantly higher anxiety and depression scores compared to those with family involvement [21].Meta-analysis indicate that family involvement is an effective measure for preventing delirium in critically ill patients, and the effectiveness of delirium prevention varies depending on the level of family involvement, with direct participation of family members in caregiving demonstrating greater effectiveness in preventing delirium [2, 10, 20]. Similarly, studies on flexible visitation policies have reported lower delirium incidence and shorter delirium duration compared with restricted visitation models [12–14]. Notably, evidence suggests that in-person visitation is more effective than electronic video visitation in alleviating feelings of loneliness and abandonment among ICU patients [22], though telecommunication still offers mental health benefits [23]. Our findings extend prior work by demonstrating that a structured, nurse-led participatory model, rather than unstructured or purely visitation-based approaches, may further enhance the effectiveness of family involvement. This is consistent with recent high-quality trials of nurse-led family support interventions, such as the FICUS trial, which highlighted the importance of structured, protocol-driven engagement in improving outcomes for patients and families [15]. In addition, qualitative evidence suggests that family members perceive such structured interventions as beneficial and value the support provided by ICU nurses [24]. The beneficial effects of NFPS are likely mediated through several interconnected mechanisms. Firstly, family presence provides emotional reassurance, reduce anxiety and fear, and facilitate patient reorientation, all of which are key modifiable risk factors for preventing delirium [8, 14, 20]. Secondly, family-assisted mobilization may increase the frequency and continuity of physical activity compared with therapist-only rehabilitation, thereby reducing immobility-related neuroinflammation and neurotransmitter dysregulation, which are considered key mechanisms in the development of delirium [1, 2, 7]. Thirdly, the integration of structured family support may improve patient engagement, adherence to care, and overall psychological well-being, creating a more favorable recovery environment [15, 25, 26]. Notably, the magnitude of effect in our study (absolute risk reduction of 7.4%) is clinically meaningful and comparable to that reported in recent meta‑analyses of family‑centered delirium prevention programs [10, 20]. However, unlike some previous trials that employed flexible or unrestricted visitation without active mobilization training [13, 26], our intervention explicitly combined family presence with protocol‑guided physical activity, which may have produced synergistic benefits. In addition to reducing delirium, NFPS was associated with shorter duration of mechanical ventilation, reduced ICU-LOS and hospital LOS, along with a significant reduction in hospitalization costs. These findings highlight the potential of NFPS not only to improve clinical outcomes but also to reduce healthcare resource utilization. The observed economic benefits are likely driven by the combined effects of delirium reduction, earlier weaning from mechanical ventilation, and accelerated recovery [4, 5, 27]. Given the substantial burden of ICU resource utilization, NFPS may represent a cost-effective strategy, particularly in settings with limited rehabilitation staffing. Firstly, delirium is a well-established independent predictor of post-ICU mortality and long-term cognitive impairment [4–6]. The incidence of long-term cognitive dysfunction in patients with delirium is 3.9 times that of patients without delirium [4]. Therefore, we hypothesize that lower incidence of delirium may indirectly contribute to reduce 1-year mortality, shorter ICU-LOS and duration of MV, and consequently, reduced hospitalization costs for ICU patients. Secondly, Family presence provides comfort, reduces anxiety and fear [10, 25, 26], and may improve coping and treatment adherence [28, 29], fostering a more conducive environment for recovery. As well as, family assistance may allow for more frequent or longer duration of mobilization sessions within safe limits, potentially accelerating functional recovery and weaning [12, 14, 30]. Thirdly, family members participation in the care of ICU patients under the guidance of medical personnel, acquiring skills and confidence, potentially facilitating smoother transitions out of the ICU and reducing readmission risks [29–31]. The combination of reduced delirium, enhanced support, potentially greater mobility, and better discharge planning likely explains the observed improvements in survival, LOS, and costs. Although we observed a trend toward lower 6‑month (6.03% vs. 7.67%) and 1‑year mortality (9.32% vs. 11.51%) in the NFPS group, neither difference reached statistical significance. This lack of significance may be attributed to the relatively low overall mortality in our cohort, which limited power to detect modest differences. Alternatively, the benefits of family participation on survival may require longer follow‑up or may be mediated through improvements in post‑ICU functional status rather than directly affecting mortality. Previous studies have similarly reported inconsistent effects of early mobilization on long-term survival [1, 7], suggesting that mortality reduction may be more dependent on patient case mix and the intensity of rehabilitation than on the specific involvement of family members. Safety and non‑inferiority for other ICU complications Importantly, we found no statistically difference in the incidence of ICU-AW (26.58% vs. 29.59%, P = 0.270), ICU-acquired infection (19.45% vs. 19.73%, P = 0.930), or adverse events (2.19% vs. 3.84%, P = 0.190) between the two groups. These findings addressing common concerns regarding the safety of family participation in ICU care [12, 13, 32]. These findings indicate that, when implemented within a structured and nurse-supervised framework, family involvement does not compromise patient safety or infection control. The absence of increased adverse events further supports the feasibility of integrating family members into routine ICU care. Nevertheless, this study did not directly assess ICU nurses’ perceptions of workload, stress, or disruption to clinical care. Previous studies have raised concerns that family presence may increase staff burden or interfere with procedures [27, 32, 33], while others have highlighted possible psychological stress for family members in the ICU environment. Our data cannot confirm or refute such concerns regarding staff experience. Future research should include direct measurement of nurse-reported outcomes, time‐motion analyses, and implementation barriers to fully evaluate the impact of NFPS on healthcare providers. Taken together, these results provide evidence that structured family participation in early mobilization is feasible and safe in the ICU setting from the perspective of patient outcomes, while acknowledging that staff perceptions warrant dedicated investigation. Potential mechanisms and contextual factors Several interconnected mechanisms may explain the observed benefits. Firstly, delirium reduction likely serves as a central driver of downstream improvements. Delirium is associated with prolonged mechanical ventilation, longer LOS, higher costs, and increased risk of long‑term cognitive impairment [4–6]. By lowering delirium incidence, NFPS may attenuate this cascade. Secondly, family involvement enhances patient psychological well‑being, reducing anxiety and fear that can trigger sympathetic overactivity and sleep disruption—both precipitant to delirium [26, 29]. Thirdly, family members may provide additional encouragement and practical assistance, enabling patients to achieve higher levels of mobilization than would be feasible with therapist‑only sessions. This is particularly relevant in resource‑constrained settings where rehabilitation staff are limited. Fourth, the training provided to family members may improve their confidence and competence, facilitating smoother transitions from ICU to ward and home, potentially reducing readmission rates [24, 30]. However, the impact of family presence is complex and influenced by visitation type, reason for ICU admission, patient’s neurological status, and family engagement level [10, 20]. There are significant challenges in involving family members in the care of ICU patients. The level of family involvement may be influenced by patient-specific factors (such as health literacy, beliefs, education level, and experiences), organizational factors (such as policies, practices, and culture), and societal factors (such as norms, regulations, and policies) [11, 31, 34]. Patients experiencing confusion often perceive and deeply value their family’s presence, deriving significant comfort and security. Therefore, integrating structured family participation into early mobilization programs may represent a promising strategy to improve both clinical outcomes and patient experience in the ICU. While the NFPS intervention demonstrated clinical benefits, it is important to recognize that requiring family members to visit twice daily and actively participate in care may impose significant burdens, including time commitment, travel costs, work absence, and psychological stress. Nevertheless, the low attrition rate in the NFPS group (1.74%) suggests that the intervention was generally acceptable to participating families, and no family member formally withdrew due to perceived burden. Future studies should prospectively assess family‑reported outcomes, such as caregiver strain, anxiety, depression, and satisfaction, to better understand the trade‑offs between patient benefit and family well‑being. Strengths and limitations This study has several notable strengths. Firstly, it evaluated a structured, nurse-led family participatory support intervention (NFPS) integrated with early mobilization in ICU patients, addressing an important gap in family-centered critical care. Unlike previous studies that primarily focused on flexible visitation or passive family presence, our intervention emphasized standardized, nurse-supervised active participation of family members in patient care, including orientation support, emotional support, cognitive stimulation, and assistance with early mobilization. This represents a more comprehensive and operationalized model of family-centered ICU care. Secondly, the multicenter design within three tertiary hospital and the extended follow-up period enhance the robustness and clinical relevance of the findings. Thirdly, the intervention demonstrated consistent benefits across multiple clinically meaningful outcomes, including delirium, duration of mechanical ventilation, ICULOS, HLOS, and hospitalization costs, highlighting its potential impact on both patient recovery and healthcare resource utilization. Fourthly, the study included a relatively large matched sample and used propensity score matching to reduce baseline imbalance between groups, thereby strengthening the robustness of the comparative analyses. In addition, all participating ICUs maintained stable clinical protocols throughout the study period, which helped reduce potential confounding from major practice changes over time. Finally, this study provides novel evidence on integrating family participation with early rehabilitation and delirium prevention strategies. Our findings thus broaden the current understanding of how family members can contribute to multidimensional ICU recovery support beyond emotional visitation alone. Several limitations should be acknowledged. Firstly, the study uses a time‑sequence design, which may introduce temporal bias. Although we confirmed that no major changes occurred in ICU protocols or staffing during the study period, unmeasured temporal factors cannot be fully excluded. Propensity score matching and standardized protocols were applied to mitigate confounding, but residual temporal bias remains a possibility. Secondly, delirium assessments were performed as part of routine clinical care by bedside nurses who were aware of group allocation because family participation was visibly apparent during the intervention phase. Therefore, assessor blinding was not feasible, which may have introduced observer bias. Thirdly, although the NFPS intervention followed a standardized protocol, the actual intensity and fidelity of family participation were not quantitatively measured. We did not systematically record the duration, frequency, completion rate, or specific types of family‑assisted activities performed during each session. Therefore, the predefined adherence target could not be formally evaluated, which limits the interpretability of intervention reproducibility and fidelity. Future studies should incorporate structured fidelity checklists, electronic logs, or digital monitoring systems to better quantify intervention delivery. Fourthly, family-centered outcomes were not formally assessed. Variables such as caregiver psychological burden, satisfaction, perceived barriers to participation, and overall caregiving experience were not evaluated, limiting understanding of the broader impact and sustainability of NFPS in routine clinical practice. In addition, ICU nurses’ perceptions regarding workload, workflow disruption, implementation burden, and acceptability of the intervention were not explored. Future implementation studies should therefore include both caregiver-reported and staff-reported outcomes to better evaluate the feasibility, acceptability, and long-term sustainability of structured family participatory interventions in ICU settings. Additionally, this study was not prospectively registered in a public clinical trial database. Although registration was not mandatory for quasi‑experimental designs at the study outset, we recognize that current standards increasingly require registration of all interventional studies. This limitation should be considered when interpreting the findings, and we will ensure prospective registration for future studies. Despite these limitations, the NFPS intervention demonstrated consistent improvements in short‑term clinical outcomes without increasing safety risks, supporting its feasibility and effectiveness in the ICU setting. Future multicenter randomized controlled trials with blinded outcome assessment, quantitative fidelity tracking, and inclusion of family‑centered and staff‑centered outcomes are warranted to confirm and extend these findings. Conclusion The NFPS strategy significantly reduced the incidence of delirium in critically ill patients and was associated with shorter durations of mechanical ventilation, reduced ICULOS and HLOS, and lower hospitalization costs. No significant differences were observed in ICU-AW, ICU-acquired infection, adverse events or long-term mortality between the two groups. These findings suggest that NFPS is a safe and effective approach to improving short-term clinical outcomes a while optimizing healthcare resource utilization in the ICU. Further multicenter prospective studies are warranted to confirm these findings and to evaluate the long-term impact of NFPS. Quality control Before the start of the study, the study protocol was approved by the ethics committee of the research unit. all participants were given a unified professional training to clearly master the research purpose, research content, research significance and other considerations. During the study, the subjects were included in strict accordance with the inclusion criteria, the data were accurately measure and the researchers conducted follow-up instructions. In addition, research members give regular feedback about the research situation and timely discuss and solve the existing problems. The data was recorded using EpiData 3.1 software and managed with SPSS22.0 software by two researchers. Supplementary Information Supplementary Material 1 Supplementary Material 1