Impact of an Integrated Medical-Nursing Clinical Pathway on 30-Day Readmission and Long-Term Prognosis in Patients with Acute Exacerbation of COPD: A Retrospective Cohort Study Using Real-World Data

Introduction

Chronic Obstructive Pulmonary Disease (COPD) remains a leading cause of morbidity and mortality worldwide, presenting a formidable challenge to healthcare systems globally.¹ According to the 2024 Global Initiative for Chronic Obstructive Lung Disease (GOLD) report, COPD is currently the third leading cause of death, causing 3.23 million deaths in 2019.² The disease trajectory is punctuated by acute exacerbations (AECOPD), defined as an acute worsening of respiratory symptoms that results in additional therapy.³ These events are not merely transient episodes but are pivotal moments that accelerate lung function decline, severely impair quality of life, and significantly increase the risk of subsequent cardiovascular events and death.⁴

A critical issue in COPD management is the “revolving door” phenomenon of hospital readmissions. A recent comprehensive meta-analysis revealed that the pooled 30-day, 90-day, and 1-year readmission rates following AECOPD hospitalization are as high as 11%, 17%, and 37%, respectively. These frequent readmissions emphasize the urgent need to put forward targeted clinical interventions to adjust or control avoidable risk factors.⁵ Studies indicate that approximately one in five patients hospitalized for AECOPD are readmitted within 30 days,⁶ with rates rising to nearly 50% within a year.⁷ High readmission rates are often attributed to fragmented care delivery, where medical treatment, nursing care, and rehabilitation are siloed. In traditional models, physicians focus on pharmacological interventions, while nurses manage daily care, often lacking a cohesive, synchronized plan for discharge readiness and post-discharge continuity. This disconnect leads to poor medication adherence, inadequate inhaler technique, and failure to recognize early warning signs of deterioration.⁸

To address these gaps, the concept of Integrated Medical-Nursing Management (IMNM) has emerged as a promising strategy. The American Nurses Association defines medical-nursing collaboration as a reliable cooperative process where both parties recognize and accept their respective responsibilities to protect mutual interests and achieve common goals.⁹ Unlike simple multidisciplinary teams, an IMNM model emphasizes a structured clinical pathway where doctors and nurses conduct joint rounds, co-develop dynamic treatment plans, and share responsibility for patient education and pulmonary rehabilitation from admission through to post-discharge follow-up.^9,10 Recent meta-analyses have demonstrated that while integrated disease management and transitional care programs consistently improve disease-specific quality of life, their impact on reducing mortality or long-term hospital readmissions remains mixed and highly dependent on the specific combination of components and overall intervention complexity.^11,12 While previous small-scale studies have suggested that such models can improve self-efficacy and satisfaction,¹³ evidence regarding their impact on “hard” clinical endpoints—specifically long-term readmission and mortality—remains limited and inconsistent. Furthermore, most existing data comes from prospective trials with strict inclusion criteria that may not reflect real-world clinical complexity. Real-world data are particularly valuable in this context as they reflect actual clinical practice patterns, patient compliance, and the true effectiveness of complex interventions outside the controlled trial environment.

Therefore, this study aims to bridge this knowledge gap by conducting a retrospective cohort analysis using real-world data from a tertiary care center in China. By leveraging Propensity Score Matching (PSM) to minimize selection bias, we evaluated whether a standardized IMNM clinical pathway, compared to traditional care, could significantly reduce hospital readmissions and all-cause mortality over a one-year follow-up period in patients with AECOPD. We hypothesized that the synergistic approach of the IMNM model would lead to superior long-term clinical outcomes.

Methods

Study Design and Setting

This was a single-center, retrospective cohort study conducted at the Department of Respiratory and Critical Care Medicine, The First Hospital of Hebei Medical University. The study protocol adhered to the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines.¹⁴ The study was approved by the Ethics Committee of The First Hospital of Hebei Medical University, and the requirement for informed consent was waived due to the retrospective nature of the analysis, although all patients had signed a general authorization for data use upon admission. This study was conducted in accordance with the Declaration of Helsinki.

Participants and Timeline

We screened all patients admitted with a primary diagnosis of AECOPD between January 2022 and January 2025. The implementation of the Integrated Medical-Nursing Management (IMNM) model in our department began comprehensively in August 2023. To evaluate the effectiveness of this model, we utilized a “before-and-after” design:

• Control Group (Pre-implementation): Patients admitted between January 1, 2022, and June 30, 2023, who received traditional care management.
• Intervention Group (Post-implementation): Patients admitted between August 1, 2023, and January 31, 2025, who were managed under the new IMNM pathway.

Inclusion Criteria

(1) Diagnosed with AECOPD according to GOLD criteria; (2) Age > 40 years; (3) Absence of other severe uncontrolled systemic diseases (eg., unstable angina, malignancy); (4) Complete electronic medical records including follow-up data.

Exclusion Criteria

(1) Severe cognitive or sensory impairment preventing communication; (2) Require invasive mechanical ventilation upon admission (indicating critical instability confounding the management model effect); (3) Concurrent diagnosis of other severe lung diseases (eg., active tuberculosis, lung cancer, interstitial lung disease); (4) Participation in other interventional trials within 3 months.

Intervention: The Integrated Medical-Nursing Management (IMNM) Model

The core of the Intervention Group was the restructuring of the care team into “Triads” consisting of one attending physician, one primary nurse, and the patient. The intervention was standardized using a daily clinical checklist to ensure reproducibility. This checklist mandated the completion of joint assessments, daily collaborative bedside rounds, and step-by-step rehabilitation targets. Specific components included:

1. Joint Assessment & Planning: Upon admission, the physician and nurse performed a dual assessment. The physician focused on pathology and medication, while the nurse assessed functional status, inhalation technique, and social support. A unified care plan was then generated within 24 hours.
2. Collaborative Rounds: Unlike separate rounds in the Control Group, the IMNM team conducted daily joint bedside rounds. This ensured that medical decisions (eg., changing antibiotics) were immediately integrated into nursing care plans (eg., monitoring specific side effects), and nursing observations (eg., sputum changes) were directly communicated to physicians.
3. Standardized Pulmonary Rehabilitation: The nurse initiated early rehabilitation (eg., pursed-lip breathing, effective coughing, early mobilization) under physician guidance starting from day 2 of admission, tailored to the patient’s tolerance.
4. Structured Discharge & Follow-up: Discharge criteria were mutually agreed upon. A “Discharge Passport” including an action plan for worsening symptoms was provided. Post-discharge, the same nursing team conducted follow-ups via WeChat or telephone at 2 weeks, 1 month, and 3 months to reinforce adherence and triage early deterioration.

The Control Group received standard care where physicians managed medical orders and nurses executed them. Rounds were separate, rehabilitation education was ad-hoc rather than systematic, and post-discharge follow-up was passive (patient-initiated) or limited to standard clinic appointments.

Data Collection and Outcome Measures

Data were extracted from the Hospital Information System (HIS) and follow-up databases. The IMNM measurements and protocols were aligned with the analytic guidelines of the 2024 GOLD report and standard collaborative care pathways, ensuring that metrics such as CAT scores, inhaler technique assessments, and follow-up intervals adhered to internationally recognized standards.¹⁵ Baseline characteristics included age, gender, BMI, smoking history, GOLD classification, FEV1% predicted, Charlson Comorbidity Index (CCI), and history of exacerbations in the prior year.

Primary Outcomes

The rate of hospital readmission due to AECOPD or all-cause readmission at 30, 90, 180, and 365 days post-discharge.

Secondary Outcomes

(1) All-cause mortality at 1 year; (2) COPD Assessment Test (CAT) scores evaluated at discharge and 3 months; (3) Pulmonary function (FEV1% predicted) at 3 months; (4) Medication adherence (assessed by the COPD Self-Management Scale, defining adherence as >80% compliance); (5) Length of stay and hospitalization costs.

Statistical Analysis

To mitigate selection bias inherent in the observational design, we employed Propensity Score Matching (PSM). A logistic regression model was constructed to calculate propensity scores based on age, sex, smoking history, GOLD stage, CCI, and prior exacerbations. Patients were matched 1:1 using the nearest neighbor method with a caliper of 0.1.

Continuous variables were expressed as mean ± standard deviation (SD) or median (IQR) and compared using Student’s t-test or Mann–Whitney U-test. Categorical variables were presented as frequencies (percentages) and compared using the Chi-square test or Fisher’s exact test. Survival analysis for readmission and mortality was performed using Kaplan-Meier curves and the Log rank test. Cox proportional hazards models were used to estimate Hazard Ratios (HR) and 95% Confidence Intervals (CI). Interaction effects for subgroup analyses were determined by adding interaction terms (treatment group × subgroup variable) to the logistic regression models to assess if the treatment effect varied across strata. A sensitivity analysis was performed by excluding patients with a high comorbidity burden (Charlson Comorbidity Index > 3) to test the robustness of the primary outcomes. All analyses were performed using SPSS version 26.0 (IBM Corp., Armonk, NY, USA). A two-sided P-value < 0.05 was considered statistically significant.

Results

Baseline Characteristics and Matching

A total of 287 patients admitted with AECOPD were initially screened. After applying exclusion criteria, 220 eligible patients remained (110 in the Control Group and 110 in the Intervention Group). Before matching, patients in the Intervention Group were significantly older (mean age 71.2 vs. 65.8 years, P=0.03) and had a higher burden of comorbidities (Charlson Comorbidity Index 3.8 vs. 2.6, P<0.01) compared to the Control Group, reflecting the real-world tendency to admit more complex cases in recent years.

After 1:1 Propensity Score Matching, 60 pairs of patients (n=120 total) were selected. The standardized mean differences for all covariates were less than 0.1 after matching, indicating excellent balance between the two groups. The mean age of the matched cohort was approximately 68.5 years, and the majority were male (72.5%). Detailed baseline characteristics are presented in Table 1. The patient selection and matching process is illustrated in Figure 1.

Table 1 Baseline Characteristics Before and After Propensity Score Matching

Figure 1 Flowchart of patient selection and propensity score matching process. A total of 287 patients were screened, and 120 patients were finally included in the matched analysis.

Primary Outcomes: Hospital Readmission

The implementation of the IMNM model resulted in a substantial reduction in hospital readmissions. As shown in Table 2, the 30-day readmission rate was significantly lower in the Intervention Group compared to the Control Group (15.0% vs. 28.3%; Odds Ratio [OR] 0.50, 95% CI 0.25–0.98; P=0.046). This benefit persisted and widened over time. At one year, the readmission rate was 26.7% in the Intervention Group versus 51.7% in the Control Group (P<0.001).

Table 2 Primary and Secondary Outcomes in the Propensity Score Matched Cohort

Kaplan-Meier survival analysis (Figure 2) demonstrated a significant difference in readmission-free survival between the groups (Log-rank χ²=9.43, P=0.002). The Intervention Group had a significantly longer median time to first readmission. The Cox proportional hazards model confirmed that the IMNM model was an independent protective factor against readmission (Hazard Ratio [HR] 0.38, 95% CI 0.21–0.70). A sensitivity analysis excluding 38 patients with a CCI > 3 showed consistent results, with the Intervention Group maintaining a significantly lower 1-year readmission rate compared to the Control Group (22.5% vs. 48.6%, P=0.012), confirming the robustness of the findings regardless of severe comorbidity burden.

Figure 2 Kaplan-Meier survival estimates for freedom from hospital readmission within 365 days. The Intervention Group (Integrated Medical-Nursing Management) showed significantly superior readmission-free survival compared to the Control Group (Log-rank P=0.002).

Secondary Outcomes: Mortality and Clinical Status

All-cause mortality at 365 days was significantly lower in the Intervention Group (3 patients, 5.0%) compared to the Control Group (8 patients, 13.3%; P=0.041). The reduction in AECOPD-specific readmissions was particularly robust (20.0% vs. 43.3%, P<0.001). Furthermore, healthcare utilization was reduced, with the Intervention Group showing fewer emergency department visits (0.8 ± 1.2 vs. 1.8 ± 1.6 visits/year, P=0.002) and fewer total hospital days over 12 months (8.3 ± 5.1 vs. 14.7 ± 8.2 days, P<0.001).

Regarding patient-reported outcomes, although baseline CAT scores were similar, the Intervention Group achieved significantly lower (better) CAT scores at the 3-month follow-up (15.8 ± 3.9 vs. 21.2 ± 5.7, P<0.001). Medication adherence at 12 months was markedly higher in the IMNM group (80.0% vs. 53.3%, P=0.004). Patient satisfaction scores were also superior in the Intervention Group (90.0% satisfied vs. 63.3%, P=0.011).

Subgroup Analyses

To explore the consistency of the intervention effect, we performed subgroup analyses stratified by age, GOLD stage, comorbidity burden, and smoking status (Table 3 and Figure 3). The protective effect of the IMNM model against readmission was consistent across all subgroups (P for interaction > 0.05 for all). Notably, patients with a high comorbidity burden (Charlson Index > 2) and those with frequent prior exacerbations (≥2/year) showed a strong trend towards greater benefit, suggesting that the integrated model is particularly effective for complex, high-risk patients.

Table 3 Subgroup Analyses for the Risk of 1-Year Hospital Readmission

Figure 3 Forest plot of subgroup analyses for the risk of 1-year hospital readmission. The Odds Ratios (OR) favor the Intervention Group across all pre-specified subgroups including age, disease severity (GOLD stage), and comorbidity burden.

Discussion

This retrospective cohort study provides compelling real-world evidence that an Integrated Medical-Nursing Management (IMNM) model significantly improves long-term outcomes for patients with AECOPD. Our findings demonstrate that compared to traditional care, the IMNM model reduced the 30-day readmission rate by approximately 50% and the 1-year readmission rate by nearly half. Crucially, we observed a significant survival benefit, with 1-year mortality dropping from 13.3% to 5.0%. These results underscore the vital importance of breaking down clinical silos and establishing a seamless, collaborative care pathway.

Our results align with and extend recent literature emphasizing the value of integrated care. A 2019 study by Prieto-Centurion et al found that comprehensive care transition interventions could reduce readmissions, but their study focused mainly on discharge planning.¹⁶ Our model goes further by integrating care from the moment of admission. The significant reduction in 30-day readmission (15.0% vs. 28.3%) validates the efficacy of the “joint rounds” component of our model. In traditional settings, medication discrepancies or lack of patient understanding regarding inhaler techniques often go unnoticed until discharge.^17,18 By having nurses and physicians round together, these issues are identified and rectified in real-time, ensuring that the patient is truly “medically ready” and “functionally ready” for discharge.

The improvement in medication adherence (80% in the intervention group) is a likely driver of the long-term benefits observed. Non-adherence is a known predictor of mortality in COPD, as evidenced by studies linking poor adherence to increased mortality risk.^19,20 The IMNM model’s structured follow-up, conducted by the same nursing team familiar with the patient, created a continuum of trust. This echoes the findings of Willard-Grace et al²¹ who reported that lay health coaching improved self-management and appropriate inhaler use in COPD patients. However, our study suggests that this does not require external health coaches but can be effectively delivered by ward nurses empowered through an integrated pathway.

Interestingly, while FEV1 improved numerically in the intervention group, the difference was not statistically significant at 3 months. This is consistent with the understanding that lung function decline in COPD is largely irreversible. However, the significant improvement in CAT scores and 6-minute walk distance indicates that while we may not change the physiology, we can significantly improve the patient’s functional status and symptom control through early rehabilitation and better self-management.^22,23 This supports the GOLD 2024 recommendation that management goals should prioritize symptom reduction and risk reduction over spirometry alone.²⁴

Our study has limitations. First, as a retrospective single-center study, causal inference is inherently limited. Despite using PSM to rigorously control for observed confounders, unmeasured variables such as socio-economic status or health literacy could still influence outcomes. Second, the sample size (n=120) is relatively small, though adequate to detect differences in primary outcomes. Third, data on readmissions to other hospitals might be underestimated, although our region’s insurance system allows for reasonably complete tracking. Furthermore, the retrospective design and lack of a complex, multistage sampling strategy limit the extrapolation of these findings. As such, the associations observed, including those in the Cox proportional hazards regression, may have been influenced by random center-specific variations or unmeasured confounders inherent to retrospective studies. Additionally, the evaluation of covariates such as medication adherence may be subject to sparse events in our retrospective dataset. As methodological studies have indicated, sparse data can lead to a phenomenon known as “monotone likelihood”, which causes the inflation of odds ratios or hazard ratios in multivariate regression models. Therefore, a monotone likelihood limitation exists in our clinical data, and the large effect estimates (and wide confidence intervals) observed for medication adherence should be interpreted with caution.²⁵ Multistage or multicenter prospective designs are needed to confirm these results and ensure they are not due to random center-specific effects.

Conclusion

In conclusion, the Integrated Medical-Nursing Management model is associated with reduced hospital readmissions and mortality, as well as improved patient quality of life and treatment adherence in AECOPD patients. While these real-world findings are promising, the single-center retrospective nature of this study warrants confirmation through larger, multicenter prospective trials to establish causality and broad generalizability.