Smart Healthcare–Enabled Information–Motivation–Behavioral Skills Model for Stroke Multimorbidity Rehabilitation: A Transitional Care Study

Abstract

Background

Patients with stroke and multimorbidity face complex rehabilitation challenges during the hospital-to-home transition. This study evaluated a smart healthcare–enabled Information–Motivation–Behavioral Skills (IMB) model integrated with traditional Chinese and Western nursing care for its effects on discharge readiness and recovery outcomes.

Material/Methods

Sixty stroke patients with 2 or more comorbidities admitted between January and June 2025 were randomly assigned to an experimental group (n=30) or control group (n=30). The control group received routine nursing care, while the experimental group received an IMB-based transitional care program supported by smart healthcare technology and integrated nursing interventions. Outcomes included discharge readiness, quality of life (Stroke-Specific Quality of Life [SS-QOL]), functional recovery (Barthel Index, Fugl-Meyer Assessment, National Institutes of Health Stroke Scale), and 30-day readmission rate.

Results

The experimental group showed significantly higher discharge-readiness scores (P<0.05), improved SS-QOL at 1 month (171.20±26.21 vs 146.55±31.25; P<0.001) and 3 months (186.25±31.35 vs 150.21±21.34; P<0.001), and better functional outcomes across all scales (all P<0.05). The 30-day readmission rate was lower in the experimental group (6.6% vs 26.6%; P=0.041).

Conclusions

A smart healthcare–supported, IMB-guided transitional care package integrated with traditional Chinese and Western nursing care was associated with improved discharge readiness, quality of life, and functional recovery, with fewer early readmissions. Because the intervention was multicomponent, the independent contribution of the IMB framework cannot be isolated.

Introduction

Stroke is an acute cerebrovascular disorder caused by rupture or occlusion of cerebral blood vessels, leading to insufficient cerebral perfusion and irreversible brain tissue injury [1]. It is characterized by high incidence, disability, mortality, and recurrence rates, making transitional care during the hospital-to-home period critical for improving recovery outcomes and quality of life [2]. In clinical practice, many stroke survivors present with multimorbidity which is the coexistence of 2 or more chronic diseases such as hypertension, diabetes, hyperlipidemia, and coronary heart disease. This results in the amplification of treatment complexity and healthcare utilization burdens. Patients with multimorbidity experience poorer functional recovery, longer rehabilitation courses, higher readmission risk, and worse long-term prognosis, due to overlapping pathophysiological mechanisms and medication interactions [3,4]. From a system perspective, multimorbidity increases care fragmentation and challenges continuity across inpatient, community, and home-based rehabilitation settings, highlighting the need for integrated transitional care models. Discharge readiness has been closely linked to rehabilitation outcomes in patients with stroke, underscoring the importance of structured and scientifically designed discharge planning [5]. However, patients with multimorbidity often encounter challenges during the transition from hospital to home, such as fragmented care, inadequate support systems, poor medication management, and limited coordination among multidisciplinary teams. These barriers heighten the risk of complications and unplanned readmissions [6]. The currently available transitional care frameworks typically focus on either stroke rehabilitation or chronic disease management in isolation rather than adopting a unified behavioral theory approach that accounts for both domains [7]. To address this gap, the present study introduces smart healthcare technologies encompassing intelligent assessment tools, real-time monitoring, and individualized intervention modules guided by the Information–Motivation–Behavioral Skills (IMB) model. The IMB model, originally proposed by Fisher and Fisher, posits that accurate health information, intrinsic and social motivation, and practical behavioral skills collectively determine sustained behavior change [8]. It has been successfully applied in chronic-disease management contexts, including diabetes, hypertension, and stroke rehabilitation [9]. By leveraging smart platforms, the IMB framework can translate knowledge and motivation into actionable rehabilitation behaviors, thereby enhancing self-management and functional recovery.

In this study, smart healthcare refers to the use of a hospital-linked digital health platform integrating mobile communication (WeChat), remote physiological monitoring, automated reminders, adherence tracking, and real-time bidirectional communication between patients, caregivers, and a multidisciplinary healthcare team. This technology-supported system enables continuous assessment, personalized feedback, and timely intervention during the hospital-to-home transition period [9]. Integrating this model with traditional Chinese and Western medicine nursing approaches offers a holistic rehabilitation strategy that addresses physical, psychological, and social dimensions of stroke multimorbidity care. The aim of this study was therefore to evaluate the effectiveness of a smart healthcare–enabled IMB model integrated with traditional Chinese and Western medicine nursing care in improving discharge readiness, long-term rehabilitation outcomes, and reducing readmission among patients with stroke and multimorbidity during the hospital-to-home transition.

To ensure reproducibility, the smart healthcare platform followed a predefined operational workflow. Patients uploaded blood pressure and heart rate readings at least once daily during the first month and at least 3 times weekly thereafter. Physical activity and sleep summaries were uploaded weekly. Automated alerts were generated when systolic blood pressure exceeded 180 mmHg, diastolic pressure exceeded 110 mmHg, heart rate was less than 50 or more than120 beats per minute, or when monitoring data were not uploaded for more than 48 hours. Alerts were first delivered to the patient and caregiver; if abnormal values persisted or were judged clinically significant, the case-manager nurse contacted the patient within 24 hours and escalated the case to the physician when required. Education modules and automated reminders were delivered 3 times weekly during the first month after discharge and weekly thereafter until month 3, and rehabilitation adherence was quantified as the proportion of scheduled sessions completed. All alerts, responses, and care-plan adjustments were time-stamped within the platform.

Material and Methods

Study Design and Ethics

This study was a single-center, prospective transitional care study conducted in the Affiliated Hospital of Nanjing University of Chinese Medicine, Nanjing, China, between January and June 2025. The study aimed to evaluate the efficacy and safety of a smart healthcare–enabled IMB model integrated with traditional Chinese and Western medicine nursing care in stroke rehabilitation. This study was approved by the Ethics Committee of the Affiliated Hospital of Nanjing University of Chinese Medicine (approval No. 2025NJ-168-01). All participants or their legal guardians provided written informed consent prior to enrollment. The study complied with the Declaration of Helsinki, International Conference on Harmonization Good Clinical Practice guidelines, and relevant local regulations. Participant confidentiality was strictly maintained. Each participant was assigned a unique study code, personal identifiers were removed from the analysis dataset, and all electronic data were stored in a password-protected hospital database accessible only to authorized personnel.

Participants

A total of 60 patients with stroke and multimorbidity admitted between January 2025 and June 2025 were enrolled in this study. Multimorbidity was defined as the coexistence of 2 or more chronic conditions. The inclusion criteria were as follows: (1) diagnosis of stroke according to Adams and Victor’s Principles of Neurology (9^th edition), confirmed by computed tomography or magnetic resonance imaging; (2) presence of 2 or more comorbidities, including diabetes mellitus, hypertension, hyperlipidemia, coronary heart disease, or chronic obstructive pulmonary disease; (3) stable vital signs and clear consciousness; and (4) availability of at least 1 cohabiting caregiver. The exclusion criteria were (1) severe organ failure or malignant tumors, and (2) cognitive impairment or psychiatric disorders that could interfere with participation.

The overall burden of multimorbidity was quantified at baseline. The median number of comorbidities per patient was 3 (interquartile range: 2–4), with hypertension, diabetes mellitus, hyperlipidemia, and coronary heart disease being the most prevalent conditions. There were no statistically significant differences between the experimental and control groups in the number or distribution of comorbidities, indicating comparable multimorbidity severity at baseline.

Eligible patients were randomly assigned in a 1: 1 ratio to the experimental group (n=30) or the control group (n=30) using a computer-generated random number sequence created in SPSS version 22.0. Allocation concealment was ensured using sequentially numbered, opaque, sealed envelopes prepared by an independent researcher not involved in recruitment or data collection. Due to the nature of the intervention, participants and care providers could not be blinded; however, outcome assessors and data analysts were independent of the intervention team. A detailed study flowchart is presented in Figure 1.

Figure 1.

Flow diagram of patient enrollment, randomization, interventions, and outcome assessments in the study.

Sample Size Calculation

The sample size was estimated based on a 2-sided significance level of α=0.05 and power (1-β)=0.80, assuming a medium effect size (Cohen’s d=0.65) for improvement in discharge-readiness scores between groups. Power analysis indicated that at least 25 participants per group were required; we included 30 per group to account for an anticipated 15% dropout rate.

Research Methods

Control Group Intervention

The control group received routine care and discharge guidance, including standardized health education sessions, individualized rehabilitation instructions, and psychological support during hospitalization. At discharge, patients were given further health counseling and printed education materials, followed by a 1-month telephone follow-up to review their condition and remind them of upcoming appointments. In the control group, routine bedside education was provided 2 to 3 times per week during hospitalization (approximately 20 minutes per session). No structured remote rehabilitation supervision, digital monitoring, or scheduled post-discharge education modules were implemented beyond the single planned 1-month telephone follow-up.

Experimental Group Intervention

The experimental group received a comprehensive nursing program combining traditional Chinese and Western medicine approaches based on the IMB model and supported by smart-healthcare technology.

IMB Framework Integration

The intervention was designed according to the IMB model, which proposes that health behavior change is driven by 3 core components: accurate health information, sufficient motivation, and practical behavioral skills. In this study, the information component included structured digital education on stroke pathology, comorbid disease management, medication use, and rehabilitation principles; the motivation component focused on enhancing self-efficacy and caregiver engagement through goal setting, regular feedback, and encouragement; and the behavioral skills component emphasized hands-on training in rehabilitation exercises, medication management, and problem-solving skills, reinforced through remote monitoring and follow-up.

Smart Healthcare Platform and Operational Workflow

Smart healthcare technology in this study consisted of a hospital-linked digital transitional care platform supported by WeChat-based communication. The platform enabled remote monitoring of blood pressure, heart rate, physical activity, and sleep quality, automated alerts for abnormal parameters, rehabilitation and medication adherence tracking, and real-time interaction between patients, family caregivers, and the multidisciplinary healthcare team.

Patients uploaded blood pressure and heart rate measurements at least once daily during the first month after discharge and at least 3 times weekly during months 2 and 3. Physical activity and sleep summaries were uploaded weekly. Automated alerts were generated when systolic blood pressure exceeded 180 mmHg, diastolic pressure exceeded 110 mmHg, heart rate was less than 50 or greater than 120 beats per minute, or when monitoring data were not uploaded for more than 48 hours. Alerts were first delivered to patients and caregivers; persistent or clinically concerning alerts were reviewed by the case-manager nurse within 24 hours and escalated to the treating physician when necessary. All alerts, responses, and intervention adjustments were time-stamped within the platform. Data generated by the platform were used to guide individualized education, rehabilitation planning, and follow-up interventions throughout the post-discharge period. The intervention included the following components.

Intervention Schedule and Intensity

During hospitalization, structured IMB-based education sessions were delivered 3 times per week (20–30 minutes per session). After discharge, digital education modules were delivered 3 times weekly during the first month and weekly thereafter until month 3. Remote video-supervised rehabilitation sessions were conducted twice weekly during the first month and weekly during months 2 and 3. Automated reminders for medication and rehabilitation exercises were sent daily during the first month and every other day thereafter. Biweekly digital progress reports were generated automatically and reviewed by the interdisciplinary team to adjust care plans.

Interdisciplinary Collaboration and Personalized Discharge Planning

Discharge plans were individually developed according to the core dimensions of the IMB model, integrating patient-specific clinical data and real-time monitoring results from the smart-healthcare platform. Staged and interactive health education was delivered through WeChat groups, prerecorded video modules, and live online sessions, with automated reminders generated according to each patient’s rehabilitation progress and adherence status. Participation and learning completion were recorded through platform-based adherence logs. Rehabilitation exercises were conducted under remote video supervision by trained therapists, while the platform continuously monitored vital signs, including blood pressure and heart rate, as well as physical activity and sleep quality. Automated alerts were issued to patients and healthcare providers when abnormal parameters were detected. In addition, family caregivers received structured online training and instructional manuals to enhance supportive care capacity. Patients were linked with community rehabilitation centers and volunteer support networks to promote sustained engagement and social participation. Biweekly digital feedback reports were generated to enable dynamic adjustment of individualized care plans throughout the follow-up period.

Intervention Fidelity and Adherence Monitoring

Intervention fidelity was monitored through platform audit logs. Adherence indicators included the proportion of scheduled education modules completed, proportion of scheduled rehabilitation sessions completed, monitoring compliance rate (calculated as uploaded days divided by scheduled monitoring days), and response rate to automated reminders. Missed sessions triggered automated reminders, and persistent non-compliance beyond 48 hours prompted direct nurse outreach. All protocol deviations were recorded in a deviation log. Adherence data were summarized descriptively, while all primary outcome analyses followed the intention-to-treat principle.

Safety and Adverse Event Monitoring

Adverse events were monitored throughout the 3-month follow-up. Participants or caregivers reported any complications via the platform or phone. An adverse event was defined as any unfavorable medical occurrence during follow-up, including falls, recurrent stroke symptoms, cardiovascular events, medication-related complications, or events requiring emergency care. Serious adverse events were defined as events resulting in hospitalization, life-threatening conditions, significant disability, or death. Relatedness to the intervention was judged by the clinical team based on temporal association and clinical plausibility. All adverse advents were recorded with date, severity, action taken, and outcome. No serious intervention-related adverse events occurred during follow-up.

Outcome Measures

Discharge Readiness

Discharge readiness was assessed using the Readiness for Hospital Discharge Scale (RHDS), a validated instrument designed to evaluate patients’ preparedness for the transition from hospital to home. The scale covers 3 dimensions: personal status, adaptive ability, and anticipatory support. Higher total scores indicate greater readiness for discharge. The RHDS has demonstrated good reliability and validity in clinical populations, and in the present study showed satisfactory internal consistency (Cronbach’s α=0.89; content validity=0.88).

Quality of Life

Quality of life was evaluated using the Stroke-Specific Quality of Life Scale (SS-QOL), a disease-specific instrument developed for stroke survivors. The SS-QOL consists of 49 items across 12 domains, including energy, family roles, language, mobility, mood, personality, self-care, social roles, thinking, upper-extremity function, vision, and work/productivity. Each item is scored on a 5-point Likert scale, with higher scores indicating better stroke-related quality of life. The SS-QOL was administered at baseline and 1 month and 3 months after discharge. The scale has been widely validated in stroke populations and demonstrated good internal consistency in the present study (Cronbach’s α=0.891; content validity=0.81).

Rehabilitation Outcomes

Functional recovery was assessed using 3 widely accepted stroke rehabilitation scales. Independence in activities of daily living was evaluated using the Barthel Index, with scores ranging from 0 to 100, and higher scores indicate greater functional independence. Motor and balance function were assessed using the Fugl-Meyer Assessment (FMA) balance scale, with scores ranging from 0 to 100, and higher scores reflecting better motor recovery. Neurological impairment severity was evaluated using the National Institutes of Health Stroke Scale (NIHSS), with lower scores represent less severe neurological deficits.

Readmission

All-cause readmission within 30 days after discharge was recorded as a short-term outcome indicator and compared between the experimental and control groups.

Assessment Procedure

All outcome assessments were conducted at predefined time points by 2 rehabilitation nurses with more than 5 years of experience in stroke rehabilitation. The assessors received standardized training in scale administration and were blinded to group allocation to minimize assessment bias. Inter-rater consistency was reinforced through joint training sessions prior to study initiation to standardize scoring procedures. Questionnaires (RHDS and SS-QOL) were administered using the validated Chinese versions available at the study site. RHDS was completed within 24 to 48 hours prior to discharge. SS-QOL was administered at admission and at 1 month (±7 days) and 3 months (±14 days) after discharge. Functional assessments (Barthel Index, FMA balance, and NIHSS) were conducted at admission and at 1 month (±7 days) and 3 months (±14 days) using standardized evaluation protocols. If patients had motor or visual limitations, items were read verbatim by the blinded assessor without interpretation. Caregiver assistance was limited to physical handling of materials and did not influence content responses. All assessments were recorded on standardized case report forms and subsequently entered into the study database by personnel independent of outcome evaluation to ensure data integrity.

Statistical Analysis

All analyses were performed using SPSS version 22.0 (IBM Corp, Armonk, NY, USA). Continuous variables are expressed as mean±standard deviation and were compared between the experimental and control groups using independent-samples t tests. Categorical variables are expressed as frequencies and percentages and were compared using the χ² test or Fisher exact test, as appropriate. Normality of continuous variables was assessed using the Shapiro-Wilk test, and homogeneity of variance was evaluated using the Levene test prior to conducting parametric analyses. In cases in which distributional assumptions were borderline, nonparametric tests (Mann-Whitney U test) were performed as sensitivity analyses, yielding consistent conclusions. Mann-Whitney U analyses were applied to RHDS total and subscale scores, SS-QOL total scores, and functional scale outcomes (Barthel Index, FMA balance, and NIHSS) when normality or variance assumptions were borderline; conclusions were unchanged compared with parametric analyses. Detailed nonparametric outputs are available from the corresponding author upon reasonable request. For outcomes assessed at multiple time points (SS-QOL, Barthel Index, FMA, and NIHSS), between-group comparisons were performed independently at each predefined time point rather than using longitudinal repeated-measures models, in line with the exploratory design of the study. These time-point comparisons were exploratory in nature and were not intended to model within-subject longitudinal trajectories using repeated-measures ANOVA or mixed-effects modeling approaches. Between-group mean differences with 95% confidence intervals (CIs) were calculated to support clinical interpretation. For continuous outcomes, between-group mean differences with 95% CIs were reported. For 30-day readmission, risk ratios with 95% CIs were calculated. Missing data were minimal (<5%) and were handled using complete-case analysis. All randomized participants were included in the primary analyses according to the intention-to-treat principle. Because multiple outcomes and independent time-point comparisons were evaluated, no formal multiplicity adjustment (eg, Bonferroni correction) was applied. Therefore, P values should be interpreted cautiously, and findings should be considered exploratory and hypothesis-generating rather than confirmatory. A 2-sided P value <0.05 was considered statistically significant.

Results

Baseline Characteristics

At baseline, the demographic and clinical variables were comparable between groups, confirming the success of randomization and ensuring the validity of subsequent comparisons. There were no significant between-group differences in age, sex distribution, educational level, payment method, stroke type, or multimorbidity burden (Table 1). Baseline stroke severity was also comparable between groups, as reflected by similar admission Barthel Index, FMA balance, and NIHSS scores (Table 1). Overall, the 2 groups were well balanced at baseline, supporting the validity of subsequent between-group comparisons.

Table 1.

Baseline demographic and clinical characteristics of patients in the experimental and control groups.

Item	Sub-item	Experimental group (n=30)	Control group (n=30)	P-value
Age (years, mean±SD)		54.42±12.62	55.04±11.25	0.876
Sex	Male	13	15	0.664
	Female	17	15
Educational level	Primary school	2	0	0.690
	Junior high school	4	6
	High school	9	11
	College and above	15	13
Payment method	New rural cooperative medical insurance	9	12	0.761
	Urban residents’ medical insurance	21	18
Stroke type	Cerebral infarction	19	17	0.705
	Cerebral hemorrhage	11	13
Multimorbidity burden	Number of comorbidities (median, IQR)	3 (2–4)	3 (2–4)	0.912
Baseline stroke severity	Barthel Index at admission (mean±SD)	27.77±3.43	27.86±4.35	0.832
	Fugl–Meyer Assessment (FMA) balance at admission (mean±SD)	27.13±4.12	27.65±5.64	0.839
	NIHSS at admission (mean±SD)	22.88±2.73	22.64±3.47	0.863

Readiness for Discharge

Analysis of RHDS scores demonstrated higher discharge-readiness levels in the experimental group compared with the control group (Table 2). Subscale differences favored the experimental group for personal status (mean difference, 5.80; 95% CI, 4.4–7.2), adaptive ability (mean difference, 6.49; 95% CI, 5.0–8.0), and anticipatory support (mean difference, 5.83; 95% CI, 4.0–7.7). The overall RHDS total score showed a between-group mean difference of 20.22 points (95% CI, 17.3–23.1; P<0.001), indicating substantially greater discharge preparedness in the experimental group. Given the exploratory design of the study, these findings should be interpreted as evidence of an association between the IMB-guided smart transitional care model and improved discharge readiness rather than definitive proof of causality.

Table 2.

Comparison of Readiness for Hospital Discharge Scale (RHDS) scores between the experimental and control groups.

Domain	Experimental group (n=30) mean±SD	Control group (n=30) mean±SD	P-value	Mean difference (Exp–Ctrl), 95% CI
Personal status	22.22±3.46	16.42±1.22	0.003	5.80 (4.4 to 7.2)
Adaptive ability	35.85±3.46	29.36±2.25	0.003	6.49 (5.0 to 8.0)
Anticipatory support	32.36±4.63	26.53±2.21	0.002	5.83 (4.0 to 7.7)
Total score	90.43±6.74	70.21±5.03	<0.001	20.22 (17.3 to 23.1)

Quality of Life

At admission, SS-QOL scores were comparable between groups (mean difference, 7.00; 95% CI, −8.4 to 22.4; P=0.834). At 1 month after discharge, the experimental group demonstrated a mean advantage of 24.65 points (95% CI, 10.0–39.3; P<0.001) relative to the control group. This difference increased at 3 months, with a mean difference of 36.04 points (95% CI, 22.4–49.6; P<0.001) favoring the experimental group (Table 3). Improvements were observed across multiple domains, including mobility, mood, and energy, suggesting sustained functional and psychosocial gains during follow-up. As time-point comparisons were conducted independently, these findings should be considered exploratory.

Table 3.

Comparison of Stroke-Specific Quality of Life Scale (SS-QOL) scores between the experimental and control groups at baseline and at 1 month and 3 months after discharge.

Time point	Experimental group (n=30) mean±SD	Control group (n=30) mean±SD	P-value	Mean difference (Exp–Ctrl), 95% CI
At admission	149.24±23.25	142.24±36.32	0.834	7.00 (−8.4 to 22.4)
1 month after discharge	171.20±26.21	146.55±31.25	<0.001	24.65 (10.0 to 39.3)
3 months after discharge	186.25±31.35	150.21±21.34	<0.001	36.04 (22.4 to 49.6)

Rehabilitation Outcomes

Functional recovery improved in both groups over time; however, improvements were consistently greater in the experimental group (Tables 4–6). For the Barthel Index, the between-group mean differences were 24.17 points (95% CI, 22.8–25.5; P<0.001) at 1 month and 29.92 points (95% CI, 28.9–30.9; P<0.001) at 3 months. For FMA balance scores, the mean differences were 14.22 points (95% CI, 10.9–17.6; P<0.001) at 1 month and 15.74 points (95% CI, 12.6–18.9; P<0.001) at 3 months. NIHSS scores showed greater neurological improvement in the experimental group, with mean differences of −2.86 points (95% CI, −4.8 to −0.9; P=0.006) at 1 month and −4.10 points (95% CI, −5.9 to −2.3; P=0.002) at 3 months. Because multiple independent time-point comparisons were performed without multiplicity adjustment, these results should be interpreted cautiously and viewed as hypothesis-generating rather than confirmatory.

Table 4.

Comparison of Barthel Index scores between the experimental and control groups at 1 month and 3 months after discharge.

Time point	Experimental group (n=30) mean±SD	Control group (n=30) mean±SD	P-value	Mean difference (Exp–Ctrl), 95% CI
At admission	27.77±3.43	27.86±4.35	0.832	−0.09 (−2.1 to 1.9)
1 month after discharge	59.12±2.52	34.95±2.97	<0.001	24.17 (22.8 to 25.5)
3 months after discharge	69.13±1.25	39.21±2.41	<0.001	29.92 (28.9 to 30.9)

Table 5.

Comparison of Fugl-Meyer Assessment (FMA) balance scale scores between the experimental and control groups at baseline and at 1 month and 3 months after discharge.

Time point	Experimental group (n=30) mean±SD	Control group (n=30) mean±SD	P-value	Mean difference (Exp–Ctrl), 95% CI
At admission	27.13±4.12	27.65±5.64	0.839	−0.52 (−3.1 to 2.1)
1 month after discharge	48.54±7.38	34.32±5.72	<0.001	14.22 (10.9 to 17.6)
3 months after discharge	59.37±5.16	43.63±6.21	<0.001	15.74 (12.6 to 18.9)

Table 6.

Comparison of National Institutes of Health Stroke Scale (NIHSS) scores between the experimental and control groups at baseline and at 1 month and 3 months after discharge.

Time point	Experimental group (n=30) mean±SD	Control group (n=30) mean±SD	P-value	Mean difference (Exp–Ctrl), 95% CI
At admission	22.88±2.73	22.64±3.47	0.863	0.24 (−1.4 to 1.9)
1 month after discharge	17.29±3.44	20.15±4.22	0.006	−2.86 (−4.8 to −0.9)
3 months after discharge	14.22±2.56	18.32±3.82	0.002	−4.10 (−5.9 to −2.3)

Readmission

Thirty-day readmission occurred in 2 patients (6.6%) in the experimental group and 8 patients (26.6%) in the control group (P=0.041) (Table 7). The estimated risk ratio was 0.25 (95% CI, 0.06–1.04) and the odds ratio was 0.19 (95% CI, 0.04–0.96). Given the small number of events and the wide confidence interval crossing unity for the risk ratio, this finding should be considered preliminary. These results suggest a possible association between the structured transitional care model and reduced early readmission; however, larger studies are required to confirm this observation.

Table 7.

Comparison of 30-day readmission rates between the experimental and control groups after discharge.

Group	30-day readmission rate	Risk ratio (95% CI)	Odds ratio (95% CI)
Experimental group	6.6% (2/30)
Control group	26.6% (8/30)
P-value	0.041	0.25 (0.06–1.04)	0.19 (0.04–0.96)

Discussion

According to the 2019 Global Burden of Disease Study, China had an estimated 28.76 million prevalent stroke cases, 3.94 million new stroke cases, and 2.19 million stroke-related deaths in 2019 [10]. In this context, stroke multimorbidity, defined as the coexistence of 2 or more chronic diseases, such as hypertension, diabetes, and coronary heart disease, has become a critical determinant of rehabilitation outcomes and healthcare utilization [11]. Although accelerated rehabilitation and early discharge programs have been introduced in many hospitals to shorten length of stay and reduce costs [12], these models often fail to meet the complex needs of patients with multiple comorbidities, who require integrated chronic-disease management and continuous education to achieve sustainable recovery. Deficiencies in self-management ability, medication adherence, and caregiver preparedness frequently contribute to functional decline and unplanned readmissions [13]. Previous studies have demonstrated that structured transitional care and follow-up programs improve quality of life and reduce post-discharge complications in stroke populations [14–16]. Our findings expand on this evidence by demonstrating that a smart healthcare–enabled IMB model was associated with further improvements in patients with multimorbidity. Compared with routine nursing, the smart healthcare–supported, IMB-guided multicomponent intervention was associated with improvements in discharge readiness, stroke-specific quality of life, functional recovery (Barthel Index, FMA, NIHSS), and fewer 30-day readmissions. Importantly, specific intervention components plausibly correspond to specific observed outcomes. Structured discharge preparation and anticipatory guidance likely contributed to higher RHDS scores. Remote supervised rehabilitation and adherence tracking may have supported the greater improvements observed in Barthel and FMA scores by increasing exercise frequency and reinforcing correct motor execution. Continuous monitoring and timely feedback may have facilitated medication adherence and early problem detection, potentially contributing to sustained SS-QOL improvements and lower NIHSS scores. While causality cannot be established within this study design, these component-outcome linkages provide biologically and behaviorally plausible explanations for the observed associations.

The IMB model has a well-established theoretical and empirical foundation across diverse health behaviors and chronic disease contexts. Originally proposed and validated by Fisher and colleagues in behavior-change interventions, the IMB framework has demonstrated effectiveness in promoting risk reduction, treatment adherence, and self-management behaviors [17,18]. Subsequent studies have confirmed the validity of IMB-based models in chronic metabolic and neurological conditions, including diabetes self-care and medication adherence [19,20] and epilepsy management [21]. Of note, recent randomized controlled evidence supports the application of the IMB framework specifically in stroke care. Ma et al demonstrated that an IMB-based medication literacy intervention significantly improved medication self-management capacity and adherence among patients with stroke [22]. However, that intervention focused primarily on medication literacy and adherence behaviors, whereas the present study evaluated a broader multicomponent transitional care package integrating digital monitoring, tele-rehabilitation supervision, caregiver training, and iterative plan adjustment. Although both interventions are grounded in IMB theory, their scope, delivery mechanisms, and active components differ, and direct equivalence of mechanisms should not be assumed.

Consistent with and extending prior evidence, our study applies the IMB model within a smart healthcare–enabled transitional care framework to address the complex rehabilitation needs of patients with stroke and multimorbidity. Mechanistically, the smart-healthcare platform strengthened the IMB pathway by enabling real-time monitoring of patient data, automated feedback, and individualized goal adjustment. This technology-supported approach likely enhanced self-efficacy, provided timely professional reinforcement, and supported sustained engagement with rehabilitation and medication regimens. The integration of traditional Chinese and Western medicine nursing concepts provided a holistic rehabilitation framework that addressed both the physiological and psychosocial aspects of recovery. The synergy between personalized health education and complementary nursing measures promoted balance restoration, emotional regulation, and improved physical endurance, consistent with integrative-medicine paradigms in chronic neurological care [23,24]. Notably, post-stroke fatigue may be a moderating factor influencing engagement with IMB-guided transitional care. Recent evidence indicates that post-stroke fatigue is highly prevalent among stroke survivors and consists of distinct physical and cognitive subtypes with different correlates [25]. Etoom et al reported that physical fatigue was generally more severe than cognitive fatigue, and that anxiety, depression, sleep disturbance, recurrent stroke, diabetes, and hypercholesterolemia were key contributors [25]. Such factors can reduce motivation, diminish self-efficacy, and limit sustained participation, thereby weakening the IMB pathway linking information and skills to behavior change. Future iterations of this smart platform could incorporate routine fatigue screening and fatigue-responsive goal setting to further individualize care. Unplanned readmission is a widely accepted quality-of-care indicator [20]. Consistent with prior studies [26,27], our results suggest that transitional care models integrating behavioral theory guidance and smart monitoring may be associated with lower early readmission rates. Improved discharge readiness, structured follow-up, and proactive monitoring may have enabled earlier detection of emerging complications or medication-related issues, thereby potentially reducing preventable readmissions. Nevertheless, the small number of events and wide confidence intervals require cautious interpretation, and these findings should be considered preliminary.

Despite these encouraging findings, this study has several limitations. First, it was a single-center trial with a relatively small sample size, which may limit generalizability. Second, the follow-up period was short (3 months), preventing evaluation of long-term sustainability. Third, blinding of participants and care providers was not feasible, potentially introducing performance bias. Fourth, multiple outcomes and independent time-point comparisons were evaluated without formal multiplicity adjustment, increasing the risk of type I error. Finally, economic endpoints such as cost-effectiveness and caregiver burden were not assessed. Future multicenter trials with larger samples, longer follow-up, and formal implementation and economic evaluations are warranted.

Conclusions

The results indicate that embedding an IMB framework within a smart healthcare–supported transitional nursing model, combined with traditional Chinese and Western nursing care, was associated with improved discharge readiness, better stroke-specific quality of life, enhanced functional recovery, and fewer early readmissions among stroke patients with multimorbidity. In this single-center prospective study, integrating behavioral theory into a digitally supported transitional care structure addressed coordination, self-management, and continuity challenges during hospital-to-home recovery. However, the modest sample size, short 3-month follow-up, single-center design, absence of multiplicity adjustment, and multicomponent nature of the intervention limit causal inference and prevent isolation of the independent contribution of the IMB framework. Within these constraints, the findings suggest that a structured, technology-assisted transitional care approach may be a feasible strategy for strengthening early rehabilitation outcomes in patients with stroke managing multiple chronic conditions. Larger multicenter studies are required to confirm generalizability and clarify mechanisms.