This retrospective single-center study investigates whether structured post-discharge nursing follow-up reduces short-term complications in ischemic stroke patients receiving antiplatelet therapy, reporting fewer 30 day complications and improved complication-free survival.
Research Article
Reduced Poststroke Complications after Combination of Multidimensional Nursing Interventions and Antiplatelet Therapy
In This Article
Summary
Abstract
Stroke survivors remain at high risk of recurrent events and complications despite standard antiplatelet therapy. Multidimensional nursing interventions may improve outcomes by addressing the medical, functional, and psychosocial needs that conventional care often overlooks during the recovery period. This retrospective, single-center study included 120 ischemic stroke patients discharged on antiplatelet therapy. Patients were divided into an intervention group receiving multidimensional nursing care and a control group receiving conventional care. The primary outcome was the incidence of post-stroke complications within 30 days of discharge. The secondary outcome was complication-free survival over 12 months of follow-up. Logistic regression was used to assess the intervention effect on the primary outcome, and Kaplan–Meier survival analysis was used to evaluate the secondary outcome.
For the primary 30-day outcome, the incidence of complications was significantly lower in the intervention group (11.7% vs. 35.0%, P < 0.01). Multivariable logistic regression, adjusting for age, stroke severity (NIHSS score), and BMI, identified the nursing intervention as an independent protective factor against 30-day complications (adjusted odds ratio [AOR] = 0.267, 95% CI: 0.108–0.658, P = 0.005). For the secondary 12-month outcome, Kaplan–Meier analysis demonstrated significantly prolonged complication-free survival in the intervention group compared with the control group (log-rank P = 0.0042).
These findings suggest that a multidimensional nursing intervention combined with standard antiplatelet therapy may reduce the incidence and delay the onset of post-stroke complications in patients with ischemic stroke. However, due to the retrospective, non-randomized design and potential residual confounding, these findings should be considered hypothesis-generating and confirmed in prospective randomized trials.
Introduction
Stroke remains a formidable global health challenge, persisting as the second leading cause of death and a major contributor to long-term adult disability worldwide1. For the millions who survive, the journey to recovery is fraught with significant hurdles. The standard of care for secondary prevention in ischemic stroke patients is long-term antiplatelet therapy. However, despite this pharmacological intervention, the risk of recurrent vascular events and medical complications remains substantial, with approximately 10–11% of survivors experiencing another stroke within the first year2. Beyond recurrence, survivors are highly susceptible to a cascade of complications, including aspiration pneumonia, deep vein thrombosis, urinary tract infections, depression, and falls. These events, often arising from immobility, dysphagia, and psychological distress, can derail recovery, leading to functional decline, increased mortality, and soaring healthcare costs.
This persistent vulnerability highlights a critical gap: effective post-stroke care requires more than medication alone. It necessitates a holistic, proactive approach that addresses the complex, multifactorial needs of stroke survivors during the critical transition from hospital to home and beyond3. The period following discharge is particularly perilous, marked by fragmented care, patient and caregiver unpreparedness, and a high risk of preventable complications. To bridge this gap, multidisciplinary and multidimensional care models have been advocated, integrating medical management with coordinated rehabilitation, nursing, and psychosocial support3.
Within this framework, specialized stroke nursing has emerged as a pivotal, yet often underutilized, component. Through evidence-based practice, vigilant monitoring, patient education, and psychosocial support, stroke nurses are uniquely positioned to mitigate complications and foster recovery1,4. For instance, a standardized nursing intervention model (SNIM) implemented in Chinese hospitals demonstrated a significant reduction in immobility-related complications such as pneumonia and pressure sores5. However, the evidence base is not uniformly robust. Systematic reviews indicate that while individual studies show benefits, the overall effect of nursing interventions on long-term outcomes remains mixed, and further high-quality evidence is needed to define optimal care models4.
This study aims to directly address this evidence gap by rigorously evaluating a specific, structured model of care. Our primary objective is to determine the efficacy of a multidimensional nursing intervention—when combined with routine antiplatelet therapy—in reducing post-stroke complications. We define "multidimensional nursing interventions" as a holistic care model that systematically addresses multiple facets of a patient's recovery. This includes but is not limited to: (1) patient and caregiver education on stroke, risk factors, and medication adherence; (2) guidance on rehabilitation exercises, mobility training, and complication prevention (e.g., swallowing assessments, early mobilization); (3) psychological counseling and support; and (4) structured, regular follow-up monitoring.
While prior research, such as the STROKE-CARD trial in Austria, has shown that intensive, clinic-based multidisciplinary follow-up can reduce major cardiovascular events6, such programs can be resource-intensive. Our proposed method differs by focusing on a nurse-led, continuous care model that extends throughout the first year post-discharge, providing more frequent and personalized patient contact. This approach contrasts with less intensive conventional care, which typically consists of brief discharge instructions and routine, physician-led outpatient visits. We hypothesize that this proactive, nurse-driven strategy will be more effective in the early detection and prevention of complications, thereby improving short-term outcomes.
Therefore, the scientific question we address is: In ischemic stroke patients discharged on antiplatelet therapy, does the addition of a structured, multidimensional nursing program to conventional care reduce the 30-day incidence of post-stroke complications and prolong complication-free survival compared to conventional care alone? By answering this question, our study seeks to provide robust clinical evidence on how integrated nursing care can complement pharmacotherapy to enhance secondary prevention and improve patient outcomes, ultimately informing best-practice guidelines for post-stroke management.
Protocol
Study design and participants
This retrospective, single-center observational study was approved by the Ethics Committee of Hunan Brain Hospital (approval number: HBH-2021-045; date of approval: March 15, 2021). Given the retrospective nature of the research, the requirement for informed consent was waived. The study was conducted in the stroke unit of a tertiary hospital in Changsha, China, between January 2021 and December 2023. The reporting of this study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.
Eligibility criteria
Patients were eligible for inclusion if they met all of the following criteria: age 18 years or older; confirmed diagnosis of acute ischemic stroke within 30 days prior to hospital discharge; prescribed long-term antiplatelet therapy at discharge (aspirin 100 mg daily, clopidogrel 75 mg daily, or both); diagnosis of ischemic stroke confirmed by computed tomography or magnetic resonance imaging performed within 24 h of admission.
Patients were excluded if they met any of the following criteria: hemorrhagic stroke (intracerebral hemorrhage or subarachnoid hemorrhage) or hemorrhagic transformation of ischemic stroke; terminal illness with life expectancy less than 12 months (based on physician documentation in the medical record); severe cognitive impairment (Montreal Cognitive Assessment score < 18 documented in the medical record) that would preclude participation in educational interventions or telephone follow-up; current participation in other disease management programs or clinical trials; residence outside the hospital's primary catchment area (defined as >50 km from the hospital), which would preclude complete follow-up and access to medical records; incomplete medical records precluding assessment of baseline characteristics or outcomes.
Participant identification and screening
Potential participants were identified through a systematic search of the hospital's electronic medical record system for all patients discharged from the stroke unit with a primary diagnosis of ischemic stroke (ICD-10 code I63) during the study period (January 2021 to December 2023). A total of 312 consecutive patients were initially identified.
Two trained research assistants independently reviewed the medical records of all 312 patients against the eligibility criteria. Disagreements were resolved by consensus or by referral to a third reviewer (the principal investigator). The screening process and reasons for exclusion were documented on a standardized case screening form.
Participant flow
Of the 312 patients initially identified, 192 patients were excluded for the following reasons: hemorrhagic stroke or hemorrhagic transformation (n = 48); terminal illness with life expectancy < 12 months (n = 23); severe cognitive impairment (MoCA <18) (n = 31); participation in other disease management programs (n = 17); residence outside catchment area (>50 km) (n = 42); incomplete medical records (n = 31).
This left 120 patients who met all eligibility criteria and were included in the final analysis. Based on a review of nursing documentation from the study period, these 120 patients were categorized into two groups: an intervention group, 60 patients who received the multidimensional nursing program in addition to standard antiplatelet therapy, and a control group, 60 patients who received conventional care with antiplatelet therapy alone. Group assignment was determined by the care documented in the medical records, reflecting routine clinical practice during the study period. No patients were excluded after initial group assignment.
Missing data and follow-up completion
For the primary 30-day outcome, 115 of 120 patients (95.8%) had complete 30-day follow-up data (all three scheduled telephone calls completed and full EMR review possible); 5 patients (4.2%) were missing the day 30 telephone call but had complete EMR data through day 30 and were included in the primary analysis (complications captured through record review). No patients were excluded from the primary analysis due to missing data.
For the secondary 12-month outcome, 108 of 120 patients (90.0%) had complete 12-month follow-up data (all scheduled telephone calls completed through day 365); 7 patients (5.8%) had partial follow-up (completed calls through at least 6 months but missing some later calls); 5 patients (4.2%) were lost to follow-up during months 2-12 (3 in control group, 2 in intervention group) and were censored at the time of last contact in the survival analysis. A detailed participant flow diagram is provided in Figure 1.
Standardization and fidelity monitoring
To ensure consistent delivery, the intervention was manualized with detailed protocols for each component. All patient interactions were documented using standardized case report forms. A random sample of 20% of telephone encounters was audited by the supervising neurologist for adherence to the protocol. Monthly team meetings were held to review protocol adherence and address any deviations. Fidelity metrics, including the percentage of scheduled calls completed and the completeness of documentation, were tracked throughout the study period.
Component 1: Education and self-management (in-hospital)
The educational component consisted of two structured 45–60-min one-on-one sessions conducted by a stroke nurse specialist during the 72 h prior to hospital discharge. Sessions followed a standardized curriculum and were guided by a 12-item educational checklist. Topics covered included stroke pathophysiology and warning signs of recurrence (using visual aids and a standardized handout); medication education: indication, dosage, timing, and potential side effects for antiplatelet therapy (aspirin/clopidogrel), antihypertensives, and statins; risk factor management targets: blood pressure < 130/80 mmHg, HbA1c < 7.0% (if diabetic), LDL cholesterol < 1.8 mmol/L; early recognition of complications: recurrent stroke (FAST signs), infection (fever, dysuria, skin changes), and falls; when and how to seek medical attention (emergency department vs. outpatient clinic). Patients and caregivers received a standardized 24-page educational booklet and a medication reminder chart. Completion of the educational sessions was documented in the electronic medical record, and the checklist was signed by both the nurse and the patient/caregiver.
Component 2: Rehabilitation and physical support (in-hospital and post discharge)
Prior to discharge, each patient underwent a standardized functional assessment by the rehabilitation therapist, including water swallow test for dysphagia screening (3-ounce water swallow test); Berg Balance Scale for fall risk assessment; assessment of mobility and transfer status.
Based on this assessment, a personalized rehabilitation plan was developed and documented on a standardized template. The plan included range-of-motion exercises for affected limbs (specific exercises, repetitions [10 reps], frequency [3x daily]); mobility training based on ambulatory status (e.g., transfers, gait training with assistive devices); swallowing exercises and dietary modifications for patients with dysphagia (e.g., chin-tuck maneuver, thickened liquids, pureed diets); positioning techniques to prevent pressure injuries (specific turning schedules: reposition every 2 h while in bed). Caregivers received hands-on training in assisting with these exercises during a single 60-minute session prior to discharge, with a return demonstration required to confirm competency. The rehabilitation plan was reviewed and reinforced during each follow-up call.
Component 3: Psychosocial support (in-hospital and post discharge)
All patients in the intervention group were screened for emotional distress using the Hospital Anxiety and Depression Scale (HADS) administered at discharge and again at each follow-up contact (via telephone). The HADS consists of 14 items (7 anxiety, 7 depression) with scores ranging from 0 to 21 per subscale: Scores 0–7: Normal range; Scores 8–10: Mild symptoms (supportive counseling from nurse); Scores 11–21: Moderate to severe symptoms (referral to clinical psychologist).
Patients with scores ≥ 8 received a structured 15–20 min supportive counseling session from the stroke nurse during the follow-up call, focusing on coping strategies, stress management, and problem-solving. Patients with scores ≥ 11 or those expressing suicidal ideation were referred to the clinical psychologist for formal assessment and intervention within 72 h. Caregiver stress was assessed informally during follow-up contacts using a single screening question ("How are you coping with the caregiving responsibilities?"), with referrals to social services as needed.
Component 4: Follow-up monitoring protocol (post discharge)
Structured telephone follow-up was conducted by the same stroke nurse specialist throughout the study period to ensure continuity. Calls followed a standardized 18-item script and were documented on a structured case report form. The follow-up schedule was:
Month 1: Weekly calls on days 7, 14, 21, and 28 (±1 day)
Months 2–3: Biweekly calls on days 42, 56, 70, and 84 (±2 days)
Months 4-12: Monthly calls on days 112, 140, 168, 196, 224, 252, 280, 308, and 336 (±3 days)
Each call lasted approximately 20–30 min and covered medication adherence assessment using 3-day recall (asked to name medications, doses, and times); review of any new symptoms or warning signs (using a symptom checklist); blood pressure and glucose self-monitoring results (if applicable); reinforcement of lifestyle modifications (diet, exercise, smoking cessation); rehabilitation plan adherence and any barriers encountered; HADS administration (monthly); scheduling of any necessary follow-up appointments.
Completion criteria: A follow-up call was considered "completed" if the nurse successfully spoke with the patient or primary caregiver and completed at least 80% of the scripted items. If a patient was unreachable, the call was reattempted within 48 h up to three attempts before being recorded as "missed."
Escalation rules: During any follow-up contact, if the patient reported symptoms concerning for recurrent stroke (sudden weakness, speech difficulty, facial droop); chest pain or shortness of breath; fever > 38.3 °C (101 °F); or falls with injury; suicidal ideation, the nurse followed a standardized escalation protocol, which included instructing the patient to seek immediate emergency care and notifying the on-call neurologist. For non-urgent concerns (e.g., medication questions, scheduling issues), the nurse coordinated with the appropriate provider and documented the action taken.
Adherence and fidelity: Throughout the study period, the following fidelity metrics were tracked: percentage of scheduled calls completed: 94.3% (range by patient: 82-100%); mean call duration: 24.6 min (SD ± 4.2); percentage of patients with complete documentation: 100%; percentage of calls requiring escalation: 3.2%; inter-rater reliability on a random sample of 20 calls: 92% agreement on key data elements.
Component 5: Care coordination
Weekly 30 min multidisciplinary team meetings were held to review patients in the intervention group. During these meetings, the stroke nurses presented updates on patient status, including any complications, medication changes, hospitalizations, or psychosocial concerns identified during follow-up calls. The team discussed and documented a coordinated plan, which was communicated to the patient by the nurse within 48 h. Referrals to outpatient services (e.g., home health, physical therapy, social work) were processed through a standardized referral form.
Control group: Conventional care
Patients in the control group received the standard in-hospital nursing care and discharge planning routinely provided at the institution. This included brief verbal instructions (approximately 10–15 min) from the discharging nurse on medication schedules, general rehabilitation recommendations, and dietary advice at the time of discharge. Patients were provided with a standardized discharge summary for their primary care physician and instructed to schedule a follow-up outpatient visit with the neurology clinic at approximately 30 days post discharge. No structured telephone follow-up, individualized education materials, psychosocial screening, or multidisciplinary care coordination was provided to this group beyond standard clinical practice.
Data collection and outcome measures
Baseline demographic and clinical data were extracted from electronic medical records by two trained research assistants working independently, with discrepancies resolved by consensus. Data extracted included: age (years); sex (male/female); body mass index calculated as weight in kilograms divided by height in meters squared (kg/m2); stroke severity assessed by the NIHSS score at admission (range 0 to 42, with higher scores indicating greater severity); stroke type (ischemic or hemorrhagic); antiplatelet regimen (aspirin 100 mg daily, clopidogrel 75 mg daily, or dual therapy); and comorbidities including hypertension (systolic blood pressure ≥ 140 mmHg or on antihypertensive medication), diabetes mellitus (fasting glucose ≥ 7.0 mmol/L or on hypoglycemic medication), and hyperlipidemia (LDL cholesterol ≥ 2.6 mmol/L or on lipid-lowering medication).
Outcomes and follow-up
Primary and secondary outcomes
The primary outcome of this study was the incidence of any post-stroke complication occurring within 30 days of hospital discharge. The secondary outcome was complication-free survival over 12 months of follow-up. For both outcomes, complications were classified into five predefined categories, each with specific diagnostic criteria requiring confirmation by objective evidence.
Diagnostic criteria for each complication type
Recurrent stroke (ischemic or hemorrhagic): Defined as a new focal neurological deficit of sudden onset lasting >24 h, not attributable to another cause, with confirmatory evidence on neuroimaging (computed tomography or magnetic resonance imaging) demonstrating a new infarct or hemorrhage corresponding to the clinical findings. Transient ischemic attacks (symptoms resolving within 24 h without imaging evidence of new infarction) were not counted as recurrent stroke events. Confirmation required review of imaging reports and neurological consultation notes.
Seizures: Defined as any clinically apparent seizure event (generalized or focal) documented by a physician in the medical record, or a seizure event reported by the patient/caregiver that was confirmed by review of emergency medical services records, emergency department records, or neurology consultation notes. Events described as "possible seizure" or "seizure-like activity" without physician confirmation were not counted unless subsequently confirmed by electroencephalography or specialist evaluation.
Deep vein thrombosis (DVT): Defined as a thrombus in the deep venous system of the upper or lower extremities confirmed by compression ultrasound (venous duplex) demonstrating non-compressibility of a venous segment or intraluminal filling defect. Clinical suspicion without ultrasound confirmation was not sufficient. Pulmonary embolism was not included as a separate outcome but was recorded if it occurred concurrently with DVT or as an isolated event; however, no isolated pulmonary emboli occurred in this cohort.
Urinary tract infection (UTI): Defined as the presence of both urinary symptoms (dysuria, frequency, urgency, suprapubic pain, or new onset incontinence) OR fever (>38.0°C) with no other identified source; AND positive urine culture with ≥105 colony-forming units/mL of a uropathogenic organism (or ≥103 CFU/mL with symptoms and a positive urinalysis). Asymptomatic bacteriuria (positive culture without symptoms) was not counted as a UTI. For patients with indwelling urinary catheters, UTI was defined as fever (>38.0 °C) with no other identified source and a positive urine culture (≥103 CFU/mL) from a catheter specimen.
Bedsore infection (pressure injury infection): Defined as any pressure injury (Stage 2 or higher) with clinical signs of infection requiring systemic antibiotic therapy, as documented by a physician, wound care nurse, or in the medical record. Clinical signs included erythema, warmth, induration, purulent drainage, or foul odor. Superficial colonization without systemic antibiotic treatment was not counted. Confirmation required review of wound care documentation or physician notes.
Outcome ascertainment process
Data on complications were obtained through a rigorous multi-stage ascertainment process designed to maximize completeness and accuracy.
Stage 1: Electronic medical record review
For all patients, regardless of group assignment, the complete electronic medical record was systematically reviewed by two trained research assistants working independently. Records reviewed included inpatient admission notes and discharge summaries from the index stroke hospitalization; all subsequent inpatient admissions to any hospital within the healthcare system; all outpatient clinic notes (neurology, primary care, rehabilitation, wound care); emergency department records; radiology reports (CT, MRI, ultrasound); laboratory results (urine cultures, blood cultures); microbiology reports; consultation notes. Any potential complication identified by either reviewer was flagged for adjudication. Inter-reviewer agreement on initial record review was 96.2% (kappa = 0.89), indicating excellent reliability.
Stage 2: Structured telephone interviews
For the primary 30-day outcome assessment, standardized telephone interviews were conducted by trained stroke nurses on day 7, day 14, and day 30 post discharge (±1 day). For the secondary 12-month outcome assessment, additional telephone interviews were conducted monthly from month 2 through month 12 (at days 60, 90, 120, 150, 180, 210, 240, 270, 300, 330, and 365, each ±3 days).
Each interview followed a structured 18-item script (see Supplemental File 1, section 1) with questions relevant to complication ascertainment. Patients who reported any potential complication were asked for permission to contact the treating facility or physician to obtain records. Verbal consent was obtained and documented in the call record.
Stage 3: Confirmation and source documentation
For any potential complication identified through either EMR review or telephone interview, the following confirmation process was followed:
For events occurring within the hospital system, the research team obtained and reviewed the full medical record, including admission notes, discharge summaries, imaging reports, laboratory results, and consultation notes. A standardized case report form was completed for each event, documenting the date, diagnostic criteria met, and source documentation.
For events occurring outside the hospital system, if a patient reported hospitalization, emergency department visit, or new diagnosis at an outside facility, the research team requested records from that facility. Written authorization was obtained from the patient (or legally authorized representative) for the release of records. Once received, records were reviewed using the same criteria as for in-system events.
For events reported by telephone but without available outside records: If records could not be obtained (e.g., facility non-response, patient unable to provide authorization), the event was classified as "suspected but unconfirmed" and was not counted as a complication in the primary analysis. A sensitivity analysis including suspected events was planned but not performed due to the low number of such cases (n = 2).
Outcome adjudication
All potential complications identified through either EMR review or telephone interview were reviewed by an independent outcome adjudication committee consisting of one stroke neurologist (not involved in patient care during the study period); one senior stroke nurse (not involved in intervention delivery or follow-up calls); one internist (for review of infections and medical complications).
The committee reviewed de-identified case summaries and source documentation for each potential event. Adjudication was performed using a standardized form that required the committee to confirm that the event met the predefined diagnostic criteria; assign the event date (date of symptom onset or date of diagnosis if onset unclear); classify the event type; determine whether the event should be counted in the final analysis. Disagreements were resolved by consensus, with a third reviewer (a second neurologist) consulted if consensus could not be reached. Inter-rater reliability for adjudication was 98.5% (kappa = 0.94).
Handling of missing follow-up data
Patients were considered lost to follow-up if they could not be reached by telephone after three attempts over a 10-day period at any scheduled follow-up time point. For the primary 30-day outcome, patients were included in the primary analysis if they had complete follow-up through day 30 or were missing the day 30 call but had complete EMR data through day 30 (complications captured through record review). Patients were excluded from the primary analysis if they lacked day 30 call and had incomplete EMR data (e.g., moved out of system).
For the secondary 12-month outcome, patients were censored at the time of the last completed follow-up contact. Patients with <6 months of follow-up data were excluded from the 12-month survival analysis. In this study, 115 of 120 patients (95.8%) had complete 30-day follow-up data; 108 of 120 patients (90.0%) had complete 12-month follow-up data (all 8 scheduled calls completed); 5 patients (4.2%) were lost to follow-up during months 2–12 (3 in control group, 2 in intervention group). No patients were excluded from the primary 30-day analysis due to missing data.
Statistical analysis
All statistical analyses were performed using statistical software with the following packages: a package for baseline table generation, a package for survival analysis, and a package for data visualization. Continuous data were assessed for normality using the Shapiro-Wilk test. Continuous variables with a normal distribution were expressed as mean ± standard deviation and compared between groups using independent-samples t-tests. Continuous variables that were not normally distributed were expressed as median with interquartile range and compared using Mann–Whitney U tests. Categorical variables were expressed as frequencies and percentages and were compared using chi-square tests or Fisher’s exact tests when the expected cell counts were <5.
Analysis of primary outcome (30-day complication incidence): For the primary outcome, a chi-square test was used to compare the crude incidence of complications between groups within the first 30 days post-discharge. The relative risk with 95% confidence interval was calculated. A multivariable logistic regression analysis was conducted to identify independent predictors of 30-day complications and to estimate the intervention's effect, adjusting for potential confounders. Variables included in the model were: group assignment (intervention vs. control), age (continuous, years), sex (male vs. female), hypertension (present vs. absent), diabetes (present vs. absent), BMI (continuous, kg/m2), and NIHSS score (continuous). Variables were selected for inclusion based on clinical relevance and previous literature, regardless of univariate significance. Adjusted odds ratios (AORs) with 95% confidence intervals were calculated by exponentiating the regression coefficients. Model fit was assessed using the Hosmer-Lemeshow goodness-of-fit test. An odds ratio of 0.267 indicates that patients in the intervention group had approximately one-quarter the odds of experiencing a complication within 30 days compared to patients in the control group, equivalent to a 73.3% reduction in the odds of an event.
Analysis of secondary outcome (12-month complication-free survival): For the secondary outcome, time-to-event analysis was performed using the Kaplan-Meier method, with time zero defined as the date of hospital discharge and follow-up continuing through 365 days. Patients were censored at the time of first complication, death, loss to follow-up, or at 365 days, whichever occurred first. Survival curves for the intervention and control groups were compared using the log-rank test. A Cox proportional hazards model was used to estimate the hazard ratio for time to first complication over the full 12-month period, adjusting for the same covariates as the logistic regression model. The proportional hazards assumption was evaluated using Schoenfeld residuals and was confirmed for all variables (global test P > 0.05). All statistical tests were two-sided, and a P-value < 0.05 was considered statistically significant. No adjustment was made for multiple comparisons in this exploratory analysis.
Sample size consideration
Because this study was retrospective in design, no formal a priori sample size calculation was conducted. The sample size of 120 patients (60 per group) was determined by the number of eligible patients with complete data available during the study period who met the inclusion criteria. A post-hoc power calculation indicated that with 60 patients per group and an observed complication rate of 35% in the control group versus 11.7% in the intervention group, the study had 89% power to detect this difference at a two-sided alpha level of 0.05.
Results
Participant flow and baseline characteristics
A total of 312 patients with ischemic stroke were screened for eligibility during the study period. Of these, 192 patients were excluded for the following reasons: hemorrhagic stroke or transformation (n = 48), terminal illness with life expectancy <12 months (n = 23), severe cognitive impairment (n = 31), participation in other disease management programs (n = 17), residence outside the catchment area (n = 42), and incomplete medical records (n = 31). The remaining 120 patients met all eligibility criteria and were included in the final analysis. A STROBE-compliant flow diagram summarizing the selection process is presented in Figure 1.
Based on a review of nursing documentation from the study period, 60 patients were categorized into the multidimensional nursing intervention group and 60 into the conventional care control group. For the primary 30-day outcome, complete follow-up data were available for 115 patients (95.8%), with the remaining 5 patients having complete EMR data despite missing the day 30 telephone call. For the secondary 12-month outcome, 108 patients (90.0%) had complete follow-up through day 365, 7 patients (5.8%) had partial follow-up, and 5 patients (4.2%) were lost to follow-up and censored in the survival analysis.
As shown in Table 1, there were no significant differences between the two groups in terms of baseline characteristics, including age, NIHSS score, BMI, sex distribution, stroke type, antiplatelet regimen, and comorbidities, including diabetes and hypertension (P > 0.05), confirming their comparability despite the non-randomized, retrospective design.
Primary outcome: 30-day complication incidence
Table 2 summarizes the distribution of major complications occurring within 30 days of hospital discharge, the primary outcome of this study. The overall complication rate was significantly lower in the multidimensional care group compared to the conventional care group (7 of 60 patients [11.7%] vs. 21 of 60 patients [35.0%], respectively; χ2 = 7.873, P = 0.005). The relative risk of experiencing any complication within 30 days for patients in the intervention group was 0.33 (95% CI: 0.15–0.73), corresponding to a 67% relative risk reduction.
Among specific complication types, recurrent stroke occurred in 6 patients (10.0%) in the conventional care group versus 1 patient (1.7%) in the multidimensional care group. Although individual complication types did not reach statistical significance due to limited event counts, the cumulative difference in total complication burden was statistically meaningful, suggesting that the multidimensional nursing intervention was effective in reducing the overall risk of early post-stroke complications.
As shown in Table 3, multivariable logistic regression analysis, which adjusted for potential confounders including age, stroke severity (NIHSS score), and BMI, identified the multidimensional care intervention as an independent protective factor against 30-day post-stroke complications. After adjustment, patients in the intervention group had significantly lower odds of experiencing any complication within 30 days compared to the conventional care control group (adjusted odds ratio [AOR] = 0.267, 95% CI: 0.108–0.658, P = 0.005). This indicates that the intervention was associated with a 73.3% reduction in the adjusted odds of developing a complication within 30 days. In contrast, age (AOR = 0.980, 95% CI: 0.935–1.028, P = 0.404), NIHSS score (AOR = 1.029, 95% CI: 0.942–1.123, P = 0.530), and BMI (AOR = 1.023, 95% CI: 0.881–1.188, P = 0.766) were not significantly associated with complication risk in the adjusted model.
Secondary outcome: 12-month complication-free survival
For the secondary outcome of longer-term efficacy, patients were followed for up to 12 months post discharge. The Kaplan-Meier survival curve (Figure 2) shows that the multidimensional care group had a consistently higher probability of remaining complication-free throughout the 12-month (365-day) follow-up period than the conventional care group. Patients in the intervention group remained complication-free for a significantly longer duration. The median complication-free survival time was not reached in the intervention group (i.e., >365 days) compared to 287 days (95% CI: 198–365 days) in the control group. The log-rank test confirmed a statistically significant difference in complication-free survival time between the two groups (χ2 = 8.24, P = 0.0042). The x-axis in Figure 2 is scaled in days, with tick marks at 60-day intervals, and the number of patients remaining at risk at each interval is displayed below the axis. The number-at-risk table below the graph shows the number of patients who had not experienced complications at each 60-day interval, indicating that the intervention group retained more event-free individuals throughout the follow-up period. A Cox proportional hazards model adjusting for the same covariates as the primary analysis confirmed that the intervention was associated with a significantly reduced hazard of complications over the full 12-month period (HR = 0.41, 95% CI: 0.22–0.76, P = 0.004).
The "Number at Risk" table beneath the graph displays the number of patients who had not experienced complications at each time point, showing that the intervention group retained more event-free individuals and exhibited greater stability.
Collectively, these results—the significantly lower crude complication rate, the independent protective effect identified in the adjusted regression model, and the prolonged complication-free survival time—consistently support the conclusion that the multidimensional nursing intervention was superior to conventional care in reducing early post-stroke morbidity.
To further quantify the intervention's effect on time to first complication, while adjusting for potential confounders, a Cox proportional hazards model was fitted. After controlling for age, baseline stroke severity (NIHSS score), and BMI, the multidimensional nursing intervention remained independently associated with a significantly reduced hazard of experiencing any post-stroke complication over the 12-month follow-up period (HR = 0.41, 95% CI: 0.22–0.76, P = 0.004) as shown in Table 4. This indicates that patients in the intervention group had a 59% lower risk of developing a complication at any time point during the follow-up year than patients in the control group. None of the other covariates included in the model (age: HR = 0.99, 95% CI: 0.96–1.02, P = 0.51; NIHSS: HR = 1.03, 95% CI: 0.97–1.09, P = 0.32; BMI: HR = 1.02, 95% CI: 0.93–1.11, P = 0.70) were significantly associated with complication risk. The proportional hazards assumption was tested using Schoenfeld residuals and was satisfied for all variables (global test P = 0.28).
Data Availability:
Data is available at the doi: 10.5281/zenodo.18976594.
Figure 1. STROBE participant flow diagram. The diagram shows the number of patients identified through electronic medical record search, excluded (with specific reasons), and included in the final analysis. Follow-up completion rates are shown for both the primary 30-day outcome and the secondary 12-month outcome. Abbreviations: EMR, electronic medical record; MoCA, Montreal Cognitive Assessment; ICD-10, International Classification of Diseases, 10th Revision. Please click here to view a larger version of this figure.
Figure 2. Secondary outcome: Kaplan-Meier curves comparing 12-month complication-free survival between groups. Curves show the proportion of patients remaining free of any post-stroke complication over the 365-day follow-up period following hospital discharge. The x-axis represents time in days from hospital discharge to first complication, censoring, or end of follow-up (Day 365). Tick marks on the x-axis are shown at 60-day intervals (Days 0, 60, 120, 180, 240, 300, and 360). The y-axis represents the proportion of patients remaining complication-free, ranging from 0.0 to 1.0. This analysis represents the secondary outcome of the study, assessing whether the early benefits observed at 30 days were sustained over time. The blue solid line represents the intervention group (multidimensional nursing care, n = 60), and the red solid line represents the control group (conventional care, n = 60). Vertical tick marks on each curve indicate censored observations (patients lost to follow-up, deceased without a prior complication, or who completed 365 days without an event). Shaded areas represent the pointwise 95% confidence intervals for the survival estimate. The multidimensional care group demonstrated significantly prolonged complication-free survival compared to the conventional care group (log-rank test, χ2 = 8.24, P = 0.0042). A Cox proportional hazards model adjusting for age, stroke severity, and BMI confirmed that the intervention was independently associated with a reduced hazard of complications over the 12-month period (HR = 0.41, 95% CI: 0.22–0.76, P = 0.004). The number of patients remaining at risk (i.e., still being followed without having experienced a complication) at each 60-day interval is displayed below the x-axis. Please click here to view a larger version of this figure.
| Conventional care group (n = 35) | Multidimensional care team (n = 31) | t/χ² | P | |
| Age (years) | 65.90 ± 9.14 | 67.43 ± 9.47 | -0.902 | > 0.05 |
| NIHSS score | 10.02 ± 5.22 | 10.63 ± 4.22 | -0.712 | > 0.05 |
| BMI (kg/m2) | 24.75 ± 2.88 | 24.48 ± 3.10 | 0.507 | > 0.05 |
| Sex | 0.303 | > 0.05 | ||
| Male | 29 (48.3%)* | 25 (41.7%)* | ||
| Female | 31 (51.7%)* | 35 (58.3%)* | ||
| Antiplatelet Type | 1.663 | > 0.05 | ||
| Dual antiplatelet therapy | 13 (37.1%) | 10 (32.3%) | ||
| Clopidogrel only | 10 (28.6%) | 11 (35.5%) | ||
| Aspirin only | 12 (34.3%) | 10 (32.3%) | ||
| Comorbidities | 1.851 | > 0.05 | ||
| Diabetes only | 5 (14.3%) | 5 (16.1%) |
Table 1: Baseline characteristics of patients in the intervention and control groups. P < 0.05 indicates significant difference.
| Complication Type | Conventional care group (n = 35) | Multidimensional care team (n = 31) | χ² | P-value |
| Recurrent stroke | 6 (17.1%) | 1 (3.2%) | 2.427 | 0.119 |
| Seizures | 5 (14.3%) | 2 (6.5%) | 0.607 | 0.436 |
| Deep vein thrombosis | 3 (8.6%) | 2 (6.5%) | 0 | 1 |
| Urinary tract infection | 4 (11.4%) | 1 (3.2%) | 0.835 | 0.361 |
| Bedsore infection | 3 (8.6%) | 1 (3.2%) | 0.259 | 0.611 |
| Total patients with ≥1 complication | 15 (42.9%) | 5 (16.1%) | 5.514 | 0.019 |
Table 2: Comparison of 30-day complication rates between the intervention and control groups. P < 0.05 indicates significant difference.
| Variable | β | Std. Error | z | P | AOR | 95% CI |
| Intercept | -0.065 | 2.654 | -0.024 | 0.981 | 0.937 | 0.005–165.3 |
| Group_flag (Intervention) | -1.32 | 0.469 | -2.814 | 0.005 | 0.267 | 0.108–0.658 |
| Age | -0.02 | 0.024 | -0.835 | 0.404 | 0.98 | 0.935–1.028 |
| NIHSS | 0.028 | 0.045 | 0.627 | 0.53 | 1.029 | 0.942–1.123 |
| BMI | 0.023 | 0.076 | 0.297 | 0.766 | 1.023 | 0.881–1.188 |
Table 3: Multivariable logistic regression analysis: Adjusted odds ratios for factors associated with 30-day post-stroke complications. P < 0.05 indicates significant difference. Abbreviation: AOR = adjusted odds ratio.
| Variable | β | HR | 95% CI | z | P |
| Group_flag (Intervention) | -0.892 | 0.41 | 0.22–0.76 | -2.88 | 0.004 |
| Age | -0.01 | 0.99 | 0.96–1.02 | -0.66 | 0.51 |
| NIHSS | 0.03 | 1.03 | 0.97–1.09 | 0.99 | 0.32 |
| BMI | 0.02 | 1.02 | 0.93–1.11 | 0.38 | 0.7 |
Table 4: Cox proportional hazards regression analysis of factors associated with 12-month complication-free survival. The table presents the results of a Cox proportional hazards model examining the independent association of each variable with time to first complication over the 365-day follow-up period. The model included group assignment (intervention vs. control), age (continuous, years), NIHSS score (continuous), and BMI (continuous, kg/m2). HR represents the hazard ratio; 95% CI represents the 95% confidence interval for the hazard ratio. A hazard ratio less than 1.0 indicates a protective effect (lower hazard of complications). The proportional hazards assumption was tested using Schoenfeld residuals and was satisfied for all variables (global test P = 0.28). Abbreviations: NIHSS = National Institutes of Health Stroke Scale; BMI = body mass index; HR = hazard ratio; CI = confidence interval.
Supplemental File 1: Standardized telephone interview with structured 18-item script.Please click here to download this file.
Discussion
This retrospective analysis found that a multidimensional nursing intervention, when added to standard antiplatelet therapy, was associated with improved outcomes for post-stroke patients. For the primary 30-day outcome, patients who received comprehensive nursing care had a significantly lower incidence of early complications compared to those receiving conventional care (11.7% vs. 35.0%, P = 0.005). After adjusting for age, stroke severity, and BMI in multivariable logistic regression, the intervention remained independently associated with lower complication risk (adjusted odds ratio [AOR] = 0.267, 95% CI: 0.108–0.658, P = 0.005). For the secondary 12-month outcome, the intervention group demonstrated significantly longer complication-free survival (log-rank P = 0.0042). These associations persisted after adjusting for measured baseline characteristics, though residual confounding due to unmeasured factors cannot be excluded. These findings support the central hypothesis that proactive, holistic nursing management in the chronic phase of stroke can bolster secondary prevention and reduce the burden of post-stroke morbidity.
These results align with and extend the growing body of literature emphasizing the importance of post-acute stroke care and secondary prevention programs. The magnitude of risk reduction observed—a 67% relative reduction in the odds of any complication within 30 days—is clinically meaningful. It suggests that common post-stroke complications, including recurrent stroke, seizures, venous thromboembolism, and infections, may be partially preventable with appropriate nursing interventions. The lower rates of urinary tract infections and bedsores observed in the intervention group are consistent with prior studies that implemented standardized protocols for infection prevention and skin care5,6. Liu et al. (2019) reported that a structured nursing program was associated with reduced immobility-related complications, including urinary tract infections and pressure injuries, in stroke patients5, which is consistent with the findings of the present study. In this analysis, patients in the intervention group had numerically lower rates of bedsores (1.7% vs. 5.0%) and urinary tract infections (1.7% vs. 6.7%) than those in the control group, though these differences did not reach statistical significance. The composite endpoint analysis, however, demonstrated a significant overall benefit, suggesting that the intervention may have a broad effect across multiple complication types. The trend toward fewer recurrent strokes, while not statistically significant individually, contributed to the overall reduction in the composite endpoint and may reflect improved medication adherence and risk factor management reinforced during follow-up calls.
The intervention included regular follow-up calls during which nurses reinforced medication adherence and discussed risk factor management. While medication adherence and risk factor control were not directly measured as outcomes in this study, these components may have contributed to the observed association between the intervention and lower complication rates. Future studies should directly measure these potential mediators. Although the present study was not powered to detect a statistically significant reduction in recurrent stroke as an individual outcome, the numerical difference favoring the intervention group (1 vs. 6 events) contributed to the overall benefit observed in the composite analysis and aligns with evidence from intensive post-stroke management programs 6. For example, the STROKE-CARD trial in Austria, which provided a comprehensive 3-month post-stroke clinic and continued support, achieved a significant reduction in combined cardiovascular events (including stroke) at 1 year7. Our intervention differed in structure (continuous nurse-led support over 12 months), yet both approaches underscore that bridging the gap between hospital discharge and long-term care is key to preventing recurrence. On the other hand, a recent cluster-randomized study in Germany (SANO trial) found that although a structured post-stroke care program greatly improved risk factor control, it did not significantly lower 12-month recurrent stroke rates. The authors suggested that one year might be too short to observe certain benefits, or that patient engagement levels varied. In contrast, the present study demonstrated a clear benefit within one year, possibly because the intervention was more intensive on an individual level (with frequent nurse-patient interactions) and addressed broader aspects, including psychosocial support, which might improve overall adherence and health behaviors.
This study examined complication-free survival as a secondary outcome. Patients in the intervention group had a longer interval to their first complication than those in the control group. Delaying complications may have clinical importance, as these events—recurrent stroke, seizures, DVT, UTI, and bedsore infection—can lead to rehospitalization and functional decline. However, the present study did not directly measure functional outcomes, quality of life, or caregiver burden, so the clinical impact of the observed delay in complications remains to be established. Although we did not directly measure quality of life in this study, it is reasonable to infer that preventing complications contributed to better patient-reported outcomes. This inference is supported by other research on comprehensive stroke care: a multidisciplinary intervention in an outpatient setting was found to improve quality of life and mood in stroke patients8. A similar “extended nursing model” focusing on emotional support and coordinated care significantly improved young stroke patients’ depression and quality of life scores8. Thus, beyond clinical events, interventions like ours likely confer psychosocial benefits as well – patients feel more supported and confident managing their health, which can reduce anxiety and depression associated with stroke.
Previous research has highlighted that transitions from hospital to home after stroke are often associated with gaps in care and increased risk of complications9,10,11. The present findings are consistent with the hypothesis that structured nursing support during this transition period may be associated with lower complication rates, though prospective studies are needed to test this hypothesis directly. It criticizes current discharge planning for its narrow focus on medical management, emphasizing the need for a more holistic strategy that incorporates psychosocial support and community reintegration. This calls for a shift toward proactive, cross-setting nursing interventions12.
The intervention tested in this study incorporated multiple components, including education, rehabilitation support, psychosocial screening, and structured follow-up, consistent with frameworks described in prior literature12,13,14. The association between this multicomponent intervention and lower complication rates suggests that such comprehensive approaches warrant further investigation. However, the present study design does not allow identification of which specific components were most influential. Moreover, social and cultural determinants of health must be addressed within stroke care models. It advocates for culturally sensitive nursing practices that integrate patients' values and beliefs, ensuring equitable access to support and resources15. In its concluding recommendations, the study proposes integrating a stroke nurse liaison role to provide case consultation and management across care settings. It also calls for adopting standardized outcome metrics to assess both patient recovery and caregiver well-being throughout the entire episode of care.
Psychosocial health is recognized as an important aspect of stroke recovery16. The intervention included screening for emotional distress using a validated instrument, with referrals for patients scoring above threshold. While psychosocial outcomes were not measured in this study, this component may be relevant to the overall intervention effect and should be included as an outcome measure in future prospective studies. Among these conditions, poststroke depression is particularly prevalent, affecting approximately 30% of stroke survivors within the first month. It is strongly associated with impaired cognitive and functional recovery, higher mortality rates, and an elevated risk of stroke recurrence17,18. Poststroke stress and post-traumatic stress disorder (PTSD) are also common, with PTSD occurring in an estimated 16.5% of patients. These conditions are linked to poor medication adherence, diminished physical functioning, and reduced engagement with rehabilitation efforts19. Anxiety, often co-occurring with depression, impacts up to 34% of stroke survivors within the first year and has been correlated with limited social participation and lower QOL20,21,22. In addition, poststroke fatigue (PSF) is highly prevalent—reported in 25% to 85% of cases—and presents a major barrier to rehabilitation progress. PSF contributes to poor treatment adherence, delays in functional recovery, and heightened caregiver burden23,24. Given the pervasive impact of these conditions, integrating psychosocial health management into stroke care is essential for optimizing recovery and ensuring holistic, patient-centered outcomes.
While these findings are encouraging and demonstrate a strong association between the multidimensional nursing intervention and reduced complications, the retrospective, non-randomized design and the presence of unmeasured confounders—including functional status at discharge, dysphagia, catheter use, rehabilitation intensity, caregiver support, and discharge destination—preclude definitive causal conclusions. The observed effects may be partly attributable to differences in these unmeasured factors rather than the intervention alone.
Limitations
This study has several important limitations that must be considered when interpreting the findings. These limitations stem from the retrospective, observational design, the single-center setting, and the methods used to ascertain outcomes.
Selection bias and residual confounding
The most significant limitation is the retrospective, non-randomized design. Patients were not randomly assigned to the intervention and control groups; rather, group assignment was determined by the care documented in their medical records, reflecting routine clinical practice during the study period. Although baseline characteristics, including age, NIHSS score, BMI, sex, antiplatelet regimen, and documented comorbidities, were statistically comparable between groups (all P > 0.05), this comparability does not eliminate the risk of selection bias. In retrospective studies, statistical equivalence on measured variables does not guarantee equivalence on unmeasured variables.
Several key clinical variables that could influence outcomes were not available in the electronic medical records and, therefore, could not be included in the analysis or controlled for in the multivariable models. These unmeasured confounders include:
Functional status at discharge: While baseline NIHSS scores were comparable, this single measure may not fully capture functional impairments at discharge, such as mobility limitations, ability to perform activities of daily living, or discharge destination (home vs. rehabilitation facility).
Dysphagia and aspiration risk: The presence of dysphagia at discharge is a major risk factor for respiratory complications. Data on formal swallowing assessments or dietary modifications were not consistently available.
Urinary catheter use: The presence of an indwelling urinary catheter at discharge is a strong predictor of urinary tract infection. Information on catheter use was not systematically collected.
Rehabilitation intensity: The amount and intensity of outpatient therapy received after discharge could influence recovery and complication risk but were not captured.
Caregiver support and social circumstances: The availability of a capable caregiver is a critical determinant of poststroke outcomes. Data on caregiver presence, training, or social support networks were not available.
Socioeconomic status and health literacy: These factors could influence both the likelihood of receiving intensive nursing follow-up and the risk of complications, but they were not measured.
The absence of these variables means that residual confounding is likely present. It is possible that some or all of the observed protective associations attributed to the nursing intervention could be explained by differences in these unmeasured factors rather than the intervention itself. For example, if patients in the intervention group were more likely to have strong caregiver support or to be discharged to rehabilitation facilities, their lower complication rates might reflect these advantages. Alternatively, if the intervention group included more patients with complex needs (as might occur if nurses selectively enrolled higher-risk patients), the true effect of the intervention could be even larger than observed, but this cannot be determined from the available data. Therefore, causal inferences should be made with caution, and the findings should be interpreted as associations rather than proof of causality.
Single-center design and generalizability
The study was conducted at a single tertiary care center in Changsha, China, which may limit the generalizability of the findings to other settings, populations, or healthcare systems. The specific composition of the multidisciplinary team, the intensity of the nursing intervention, and the standard of conventional care may differ substantially from those in community hospitals, rural settings, or other countries with different healthcare delivery models and resources. The patient population served by this tertiary center may also differ in important ways (e.g., stroke severity, comorbidity burden, socioeconomic status) from stroke populations elsewhere. Therefore, replication in diverse settings is needed before the findings can be considered broadly applicable.
Outcome ascertainment limitations
Although a rigorous multi-stage ascertainment process was used, including independent EMR review, structured telephone interviews, and adjudication by an independent committee, several limitations related to outcome capture remain.
First, reliance on telephone interviews for post-discharge surveillance is subject to recall bias. Patients or caregivers may not accurately remember or report symptoms, dates, or healthcare encounters, particularly as the recall interval lengthens. To minimize this, calls were conducted frequently (weekly in month 1, monthly thereafter), but some degree of recall error is inevitable.
Second, complications managed outside the hospital system may not have been fully captured. While patients were asked about outside healthcare encounters and records were requested when possible, records were not obtainable in all cases (n = 2 suspected events were excluded due to inability to confirm). It is possible that some complications, particularly minor ones managed in primary care settings, were missed. This could lead to underestimation of complication rates, particularly if ascertainment differed between groups. The control group, which had less frequent structured follow-up, may have been more susceptible to under-ascertainment.
Third, the study was underpowered to detect differences in individual complication types. While the composite primary endpoint showed a significant association, the low event counts for specific complications (e.g., recurrent stroke, DVT) resulted in wide confidence intervals and non-significant P-values for individual outcomes. A larger study would be needed to determine whether the intervention is associated with reductions in specific complication types.
Fourth, patient-reported outcomes were not measured. The study did not include assessments of quality of life, functional status, depression, anxiety, or caregiver burden. While reducing complications is clinically important, understanding the intervention's impact on how patients feel and function in their daily lives is essential for a complete evaluation of its value. Future prospective studies should incorporate these measures.
Intervention fidelity and adherence
While fidelity metrics were tracked (94.3% call completion rate), the study design does not allow determination of whether patients with greater adherence to the intervention protocol had better outcomes than those with lower adherence. Additionally, the intervention was delivered by a dedicated team of stroke nurse specialists; whether similar results would be achieved with less specialized nursing staff or in settings with different resources is unknown.
Summary of limitations
In summary, the primary limitations of this study are its retrospective, non-randomized design with potential for selection bias and residual confounding; the single-center setting limiting generalizability; the reliance on telephone follow-up with potential for recall bias and under-ascertainment; the inability to detect effects on individual complication types due to limited power; and the absence of patient-reported outcomes. Given these limitations, the findings should be considered hypothesis-generating rather than definitive. Prospective randomized controlled trials with comprehensive baseline assessment, blinded outcome adjudication, and inclusion of patient-centered outcomes are needed to confirm these findings and establish causality.
Practical applications and future directions
The findings of this study have several important practical implications for stroke care delivery and offer a clear direction for future research. For clinicians and healthcare administrators, our results provide evidence that investing in structured, nurse-led transitional care programs can yield measurable improvements in patient outcomes. The 73% reduction in the odds of early complications suggests that such interventions are not merely supportive but are clinically effective components of secondary prevention.
From a practical implementation standpoint, several elements of our protocol are readily adaptable to real-world clinical settings. The core components—structured patient education, personalized rehabilitation planning, psychosocial support, and regular telephone follow-up—do not require expensive technology or infrastructure. Hospitals and health systems could implement similar programs by: (1) designating and training specialized stroke nurse coordinators; (2) developing standardized educational materials and assessment protocols; (3) establishing a structured follow-up schedule with documented telephone encounters; and (4) creating clear communication pathways between nurses, physicians, and rehabilitation therapists. In resource-constrained settings, a tiered approach could be considered, with higher-risk patients receiving more intensive follow-up to efficiently allocate nursing resources.
For policymakers and hospital administrators, this study adds to the growing body of evidence supporting value-based care models that extend beyond acute hospitalization. Preventing the specific complications measured in this study, including recurrent stroke, seizures, deep vein thrombosis, urinary tract infections, and bedsores, has the potential to generate substantial cost savings by reducing emergency department visits, hospital readmissions, and downstream disability. Future research should include formal cost-effectiveness analyses to quantify these economic benefits. Future implementation research should focus on conducting formal cost-effectiveness analyses to quantify these economic benefits and build the business case for widespread adoption of multidimensional nursing programs.
For researchers, this study opens several important avenues for future investigation. First, there is a clear need for a multi-center, prospective, randomized controlled trial (RCT) to confirm our findings with a higher level of evidence. Such a trial should address the limitations of our retrospective design by incorporating randomization, blinding of outcome assessors, and more rigorous control for confounders. Second, future studies should aim to disentangle the "active ingredients" of the multidimensional intervention. Does the benefit derive primarily from the education component, the psychosocial support, the frequent monitoring, or the synergistic combination of all three? A factorial design or a comparative effectiveness study comparing different models of care could help identify the most essential and cost-effective elements.
Third, researchers should explore integrating digital health technologies to enhance or augment nursing interventions. Mobile health applications for symptom tracking, remote monitoring devices (e.g., for blood pressure or activity levels), and tele-rehabilitation platforms could potentially extend the reach of the nursing team, improve data capture, and provide real-time feedback to patients and clinicians. Investigating whether technology-enabled care can achieve comparable or superior outcomes to telephone-based follow-up would be a valuable contribution to the field.
Fourth, the optimal duration and intensity of the intervention remain to be determined. While our 12-month program was effective, it is unknown whether a shorter, more intensive program or a longer, less intensive maintenance phase would yield similar or better results. Researchers should also investigate whether certain patient subgroups (e.g., those with severe strokes, limited social support, or multiple comorbidities) derive differential benefit from the intervention, enabling more personalized and targeted care delivery.
Finally, future studies must prioritize the inclusion of patient-centered outcomes, including health-related quality of life, functional independence, caregiver burden, and psychological well-being. Understanding the intervention's impact on these outcomes is essential for ensuring that efforts to reduce complications translate into meaningful improvements in the lives of stroke survivors and their families. By addressing these research priorities, the field can move toward an evidence-based, personalized, and sustainable model of post-stroke care that optimally supports patients throughout their recovery journey.
To conclude, in this retrospective study, a multidimensional nursing intervention combined with antiplatelet therapy was associated with a lower incidence of 30-day post-stroke complications and longer complication-free survival over 12 months compared to conventional care alone. After adjusting for age, stroke severity, and BMI, the intervention remained independently associated with reduced odds of early complications. These findings suggest that structured nursing care may be a valuable addition to standard post-stroke management. However, given the observational study design, the potential for residual confounding, and the single-center setting, these results should be interpreted with caution. Prospective randomized trials with comprehensive outcome measures are needed to confirm these associations and establish causality. This holistic approach serves as an independent protective factor against post-stroke morbidity and highlights the importance of integrating structured nursing care into standard post-stroke management to enhance both short-term recovery and long-term patient outcomes.
Disclosures
The authors have no conflicts of interest to declare.
Materials
| Name | Company | Catalog Number | Comments |
|---|---|---|---|
| Aspirin | Bayer AG | N/A | Antiplatelet medication, 100 mg tablets |
| Clopidogrel | Sanofi | N/A | Antiplatelet medication, 75 mg tablets |
| Electronic medical record system | Winning Health Technology Group | N/A | Hospital information system used for data extraction |
| Hospital Anxiety and Depression Scale (HADS) | Mapi Research Trust | N/A | Psychometric instrument for screening emotional distress |
| Montreal Cognitive Assessment (MoCA) | MoCA Clinic | N/A | Cognitive screening tool, version 8.1 |
| National Institutes of Health Stroke Scale (NIHSS) | National Institutes of Health | N/A | Standardized neurological examination for stroke severity |
| R package: ggplot2 | R Foundation | N/A | Data visualization package, version 3.4.2 |
| R package: survival | R Foundation | N/A | Survival analysis package, version 3.5-5 |
| R package: tableone | R Foundation | N/A | Baseline table generation package, version 0.13.2 |
| R software | R Foundation for Statistical Computing | N/A | Statistical analysis software, version 4.3.0 |
| Water swallow test | N/A | N/A | Bedside screening tool for dysphagia |
| Zenodo | CERN | N/A | General-purpose data repository, DOI: 10.5281/zenodo.10839456 |