Survival Prediction Model Construction and Validation for Patients with Major Burns

Background

Burns are the sixth leading cause of death from injuries globally, particularly for patients with major burns (≥70% total body surface area, TBSA), who face severe complications and poor prognosis. Despite advancements in burn treatment, there is still a lack of an effective predictive model to assess the overall survival (OS) of these patients. Therefore, this study aims to identify key prognostic factors affecting the survival of patients with major burns and construct a feasible predictive model to help clinicians quickly evaluate survival probabilities and make clinical decisions early in treatment.

Research Source

This study was conducted by a research team from multiple institutions, including the First Affiliated Hospital of Naval Medical University and Shanghai Jiao Tong University School of Medicine. The team includes researchers such as Runzhi Huang, Yuntao Yao, and Linhui Li. The study was published online on July 4, 2024, in the International Journal of Surgery.

Research Process and Results

1. Study Subjects and Data Collection

The study included 144 patients with burns ≥70% TBSA admitted to the First Affiliated Hospital of Naval Medical University between 2010 and 2020. Additionally, 185 patients from the 2014 Kunshan aluminum dust explosion incident were used as an external validation cohort. The primary endpoint was overall survival (OS), with patients categorized into survival and death groups.

2. Statistical Analysis and Predictive Model Construction

The study initially screened potential predictors using the χ² test and Kaplan-Meier (K-M) survival analysis. Subsequently, multivariate Cox regression analysis identified independent prognostic factors. A nomogram based on individual patient factors was constructed to predict 1-week, 4-week, 8-week, and 12-week survival probabilities.

3. Key Findings

The study found that sex, the percentage of third- and fourth-degree burns, and organ dysfunction were significant independent factors affecting patient survival. Specifically: - Sex: Male patients had a significantly higher risk of death than females (HR=2.22, 95% CI=1.09–4.50, p=0.028). - Burn Depth: A higher percentage of third- and fourth-degree burns was associated with lower survival rates (HR=8.10, 95% CI=2.11–31.10, p=0.002). - Organ Dysfunction: Organ dysfunction significantly increased the risk of death (HR=8.65, 95% CI=3.11–24.10, p<0.001).

4. Nomogram Construction and Validation

Based on the independent factors, a nomogram was constructed to quickly assess patient survival probabilities. Internal and external validation (Kunshan cohort) demonstrated the model’s good predictive accuracy, consistency, and discrimination. Calibration curves and ROC curves further validated the model’s reliability.

5. External Validation

Validation using the Kunshan cohort showed that the model performed well in predicting survival rates for patients with major burns. Particularly for 8-week and 12-week survival predictions, the model’s AUC (area under the curve) reached 0.881 and 0.920, respectively, indicating high predictive accuracy.

Conclusion and Significance

This study successfully identified key factors affecting the survival of patients with major burns and constructed an effective predictive model. The model not only has significant scientific value but also provides clinicians with a practical tool to quickly assess survival probabilities early in treatment, thereby optimizing treatment plans. Additionally, the study highlights the importance of burn depth and organ dysfunction in predicting patient outcomes, offering new directions for future burn treatment research.

Research Highlights

1. Innovation: The first nomogram based on individual patient factors to predict survival rates for patients with major burns.
2. Practicality: The model demonstrated good predictive accuracy through internal and external validation, making it highly valuable for clinical application.
3. Multicenter Validation: External validation using the Kunshan cohort further enhanced the model’s reliability and generalizability.

Additional Valuable Information

The study also explored the impact of early fluid resuscitation on the prognosis of patients with major burns, finding that most patients received adequate fluid resuscitation within 24 hours. This further supports the importance of early fluid resuscitation in burn treatment.

This study provides new tools and methods for prognostic assessment in patients with major burns, holding significant scientific and clinical value.