Dynamic Modified Early Warning Score for predicting postoperative mortality in acute type A aortic dissection: a retrospective study

Introduction

Acute type A aortic dissection (AAAD) has a rapid onset and a mortality rate of up to 23.7% within 48 hours (1). Surgical repair remains the mainstay of treatment. Due to the complexity of surgery and the specificity of the condition, preoperative antiplatelet and antithrombotic therapy, intraoperative high-dose systemic heparinization therapy, extensive blood dilution, and fibrinolysis all increase the risk of postoperative organ bleeding and death in patients (2,3). According to reports, the postoperative mortality of AAAD is 16.51–25% (4-7). Therefore, early identification and intervention for high-risk patients after surgery are critical for improving outcomes. Nevertheless, there remains a lack of validated and objective tools to predict postoperative mortality risk in this population.

Vital signs are often the earliest indicators of abnormal physiological changes (8). Based on this principle, Early Warning Score (EWS) systems—most notably the Modified Early Warning Score (MEWS) and the National Early Warning Score (NEWS)—have been developed to detect patient deterioration and have proven useful in predicting inpatient mortality (9-11). Although EWS has been studied in emergency settings for acute aortic dissection (AD) (12), its postoperative use in AAAD is unclear, and the prognostic value of dynamic postoperative MEWS changes has not been examined. This study aimed to evaluate the predictive performance of MEWS at multiple postoperative time points within the first 24 hours following surgery for AAAD, and to explore the association between MEWS trajectories and postoperative mortality. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-898/rc).

Methods

Research design

This study is a single-center retrospective analysis conducted on 279 AAAD patients who underwent surgery at The First Affiliated Hospital of Shantou University Medical College between January 2020 and December 2022. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of The First Affiliated Hospital of Shantou University Medical College (No. B-2022-286) and individual consent for this retrospective analysis was waived.

Patient inclusion criteria

• Diagnosed as having AAAD based on cardiac ultrasound and aortic computed tomography (CT) angiography, and underwent aortic surgical treatment;
• Age ≥18 years old;
• Postoperative survival time ≥24 hours;
• Complete case data including the patient’s general information and vital signs during the time period required for data collection, was available;
• AD patients with an acute onset time within 2 weeks.

Patient exclusion criteria

• Pregnant;
• Type B aortic dissection (TBAD) patients;
• Recurrent AD patients;
• AD caused by Marfan syndrome or iatrogenic injury.

Data collection

Demographic and clinical data were collected from the hospital’s electronic and paper systems, including patient characteristics such as gender, age, body mass index (BMI), hypertension, diabetes, smoking status, alcohol consumption, pre-hospital time, waiting time for surgery, left ventricular diameter (LVD), blood sugar levels, left ventricular ejection fraction (LVEF), and cardiopulmonary bypass time. Additionally, the MEWS was recorded at admission and at seven postoperative time points (0, 4, 8, 12, 16, 20, and 24 hours). A total of 1,953 sets of comprehensive vital sign data were collected, with all measurements being consistently recorded using standardized methods across all groups. The primary outcome of the study was in-hospital mortality among patients with AAAD. The MEWS was assessed using standardized clinical parameters: systolic blood pressure, heart rate, respiratory rate, body temperature, and level of consciousness. Blood sugar and other laboratory values were measured using standardized laboratory assays within the hospital’s certified clinical laboratory. Echocardiography was performed by certified sonographers following uniform protocols to determine LVEF and LVD. Smoking and alcohol consumption statuses were assessed based on patient self-reports documented during admission. Due to the retrospective design, some variables, such as lactate levels and malperfusion signs, were not consistently available, and thus were not included in the analysis.

Definition

The MEWS was selected for its simplicity, availability at the bedside, and suitability for high-frequency postoperative monitoring in intensive care unit (ICU) settings. The MEWS consists of five physiological indicators, including vital signs and consciousness. Four of the parameters are scored on a scale from 0 to 3, while the temperature parameter is scored from 0 to 2, with a total score ranging from 0 to 14. The higher the score, the greater the risk level. In this study, the MEWS was assessed at 0, 4, 8, 12, 16, 20, and 24 hours post-AAAD surgery. If a deterioration in any vital sign was observed within two hours before and after a given time point, that deteriorated vital sign was recorded. In cases where multiple vital signs showed deterioration, the worst (highest risk) vital sign was selected.

Statistical analysis

All statistical analyses were performed using SPSS 26.0 and R software. Descriptive statistics were used to summarize the data. Normally-distributed continuous variables were expressed as mean ± standard deviation, and comparisons between the surviving and deceased groups were conducted using independent-sample t-tests. Non-normally distributed continuous variables were presented as median and interquartile range, and differences between groups were assessed using the Mann-Whitney U test. Categorical variables were analyzed using the chi-square (χ²) test. The predictive accuracy of the MEWS at various time points was assessed using receiver operating characteristic (ROC) curves. Latent class growth modeling (LCGM) was used to classify patients into three trajectory models, with Akaike information criterion (AIC), Bayesian information criteria (BIC), and entropy used to select the optimal model. Analysis of variance (ANOVA) and Tukey post hoc tests were conducted for group comparisons, and logistic regression was used to identify independent predictors of mortality.

Results

A total of 279 patients were included in this study. A retrospective analysis was conducted on 326 observed subjects from January 2020 to December 2022. Exclusions were made for the following: one case of intraoperative dissection, six cases of chronic dissection, two cases of Marfan syndrome, one case of pregnancy, two cases of intraoperative death, five cases of postoperative death within 24 hours, 16 cases of ascending aortic aneurysm/dilation, two cases of secondary surgery, and 12 cases of B-type AD. The subjects were divided into a surviving group and a deceased group based on in-hospital mortality after AAAD surgery. The surviving group consisted of 242 cases, while the deceased group included 37 cases, resulting in a 13.3% mortality.

Baseline characteristics are summarized in Table 1. The majority of patients were male (71.3%), with 46.2% ≥60 years of age. Smoking status and cardiopulmonary bypass time differed significantly between groups (P<0.05), while the MEWS at admission showed no statistical difference.

Among the 279 AAAD patients, the highest MEWS 24 hours after surgery was 12, and the lowest was 0. Mean MEWS was higher in the deceased group than in the survival group at all postoperative time points (0–24 hours, P<0.05; Table 2 and Figure 1), but both groups showed a gradual decline in the MEWS over time. Figure 2 shows the ROC curve for MEWS on admission. The area under the curve (AUC) for the MEWS on admission was 0.517, which was not statistically significant.

Figure 1 MEWS at each time point. MEWS, Modified Early Warning Score.

Figure 2 ROC for MEWS on admission for AAAD. AAAD, acute type A aortic dissection; MEWS, Modified Early Warning Score; ROC, receiver operating characteristic.

Table 3 and Figure 3 present a comparison of the basic parameters for the MEWS ROC curves at different time points. The maximum MEWS was 0.758 [95% confidence interval (CI): 0.673–0.843] at 12 hours, with an optimal cutoff point of 4, sensitivity of 0.622, and specificity of 0.781. The ROC curves at each time point post-AAAD surgery were all greater than 0.7, and the optimal cutoff point throughout the 24-hour period ranged from 3 to 5.

Figure 3 Line plot of AUC values for MEWS at 0th, 4th, 8th, 12th, 16th, 20th, and 24th hour after AAAD surgery. AAAD, acute type A aortic dissection; AUC, area under the curve; MEWS, Modified Early Warning Score.

We used the MEWS optimal critical point 4 at the 12th hour as the basis for dividing patients into two groups 24 hours after AAAD surgery. AAAD patients with at least one time point above a MEWS of 4 points after surgery were classified as the high MEWS group, Otherwise, they were classified as the low MEWS group. As shown in Table S1 and Figure 4, there was a statistically significant difference between the surviving group and the deceased group.

Figure 4 Number of survivors and deaths in the high MEWS group and low MEWS group. MEWS, Modified Early Warning Score.

Logistic regression identified 12-hour MEWS, cardiopulmonary bypass time, and smoking as independent risk factors for in-hospital mortality (Table 4). In multivariate analysis adjusted for age and sex, 12-hour MEWS remained associated with mortality [odds ratio (OR) =1.613; 95% CI: 1.287–2.021; P<0.001].

Figure 5 presents three distinct trajectories of MEWS levels. LCGM analysis and time scale were employed to explore the longitudinal changes in MEWS (additional details provided in Table S2). In model selection, although the AIC and BIC values of the four-class model were similar to those of the three-class model, the three-class solution was ultimately chosen after considering model fit indices and clinical interpretability. Over time, all three identified MEWS trajectories showed a downward trend. The Chi-squared test yielded χ²=41.283, degrees of freedom (DF) =2, P<0.05, indicating statistically significant differences in mortality outcomes across the different trajectories. The mortality for class 1 (45.9%), was higher than that for class 2 (6.0%) and class 3 (11.8%), highlighting a higher mortality risk in class 1 patients (Table S3). ANOVA demonstrated significant differences in the MEWS between categories (F =59.14, P<0.05) (Table S4). Further Tukey post hoc testing confirmed significant differences in MEWS among the three groups. These findings underscore the significant relationship between MEWS trajectories and mortality outcomes, supporting the potential of MEWS trajectories in predicting patient prognosis. A highly stable trajectory is significantly associated with an increased risk of mortality after AAAD surgery.

Figure 5 MEWS trajectory. Class 1: patients with persistently high MEWS and a relatively stable plateau. Class 2: patients with low MEWS. Class 3: patients with moderate MEWS showing a mild downward trend. These classifications were data-driven based on MEWS trajectory patterns rather than predefined cut-off values. MEWS, Modified Early Warning Score.

Discussion

This study demonstrates that the MEWS, a simple physiological scoring tool, has prognostic value for in-hospital mortality following AAAD surgery. The MEWS measured at 12 hours postoperatively had the highest predictive accuracy. Additionally, distinct MEWS trajectories within the first 24 hours were significantly associated with mortality risk, underscoring the value of dynamic monitoring.

Unlike the NEWS, the MEWS does not include blood oxygen saturation or the administration of oxygen. MEWS is simpler. In a multicenter prospective survey of 410 critically ill patients, compared the NEWS, MEWS was considered a valid predictive score for morbidity and mortality in critically ill patients (13). Therefore, we selected the MEWS for this study due to its simplicity and feasibility in high-frequency postoperative monitoring. Unlike more complex scores [e.g., Sequential Organ Failure Assessment (SOFA), Acute Physiology and Chronic Health Evaluation II (APACHE II)], the MEWS can be calculated at bedside without laboratory data, making it practical for ICU settings. Although it does not include oxygen saturation or biochemical parameters, it effectively captures vital sign deterioration.

Our results further showed that the MEWS in the deceased group was consistently higher than those in the surviving group across all postoperative time points. MEWS values remained above 0.7, supporting its applicability for early postoperative assessment in AAAD patients. Our results are similar to other studies, and the MEWS is a useful tool for predicting in-hospital mortality (9,10,14,15).

Few studies have explored the application of EWSs in AAAD patients. Liu et al. reported that preoperative NEWS could predict clinical deterioration during emergency observation in patients with acute AD (12). However, their study focused exclusively on preoperative patients in emergency department settings. In contrast, our research fills a crucial gap in the literature by focusing on the postoperative period, a phase where dynamic monitoring of patient status is critical. Our use of LCGM to analyze MEWS trajectories is a novel approach that provides deeper insights into the physiological changes that occur after surgery. By identifying high-risk MEWS trajectory patterns, our study offers a new dimension for early warning systems in the postoperative management of AAAD patients.

Although previous studies have reported that EWS at admission could predict in-hospital mortality, our data showed otherwise (16-18). The better predictive performance of postoperative MEWS can be explained by several factors. This difference may stem from the specific pathophysiological and management characteristics of AAAD. Patients who are able to undergo surgery typically receive urgent preoperative stabilization, including sedation, blood pressure management, and fluid resuscitation, which can temporarily normalize vital signs. This stabilization may help them survive until surgery, but it does not necessarily eliminate the risk of postoperative complications or death. Additionally, the MEWS is not specifically designed to capture the unique pathophysiological features of AAAD preoperatively, which may limit its utility before surgical intervention. As a result, the MEWS at admission may not accurately reflect the severity of the disease, reducing its discriminatory power.

In contrast, the complexity of surgery and the specific conditions of AAAD, including high-dose systemic heparinization, extensive blood dilution, and fibrinolysis, increase the risk of postoperative bleeding and death (2). The postoperative MEWS, assessed in the immediate post-surgery period, may better reflect the patient’s physiological reserve and response to surgical stress, making it more relevant for predicting post-surgical outcomes in AAAD.

Vital signs are an important part of monitoring adverse events during a patient’s hospitalization (19). Currently, most vital sign-based EWSs utilize the data obtained when patients were at the peak of disease at a point in time or over a period of time for predicting patient prognosis (20-22). However, dynamic trend trajectories of EWSs have been shown to help predict patient prognosis more accurately (12,23). Our study demonstrates that high-risk MEWS trajectories are strongly associated with a significantly increased risk of mortality. This dynamic analysis captured early physiological changes that static scores might miss.

Identifying the optimal monitoring time is crucial in resource-limited ICU settings. The MEWS at 12 hours demonstrated the best discrimination, indicating that targeted assessment at this time may improve early risk stratification. Both surviving and deceased groups showed a general downward trend in MEWS over time, reflecting responses to therapeutic interventions. However, persistently high MEWS trajectories remained a strong warning signal for poor outcomes.

The optimal cutoff value for MEWS is still debated (13,24,25), which may be due to the fact that it is associated with different diseases and outcome events, and is also influenced by the different time points at which MEWS is calculated and the size of the study sample. In this study, the optimal cutoff points at different time points within 24 hours after AAAD surgery were distributed between 3–5 points. Notably, when the MEWS was ≥4, patients typically exhibited severe physiological deterioration, including hypotension, tachycardia, or impaired consciousness. Rapid activation of critical care interventions is warranted when the MEWS exceeds this threshold to prevent further clinical deterioration.

While the MEWS is not specific to aortic pathology, it effectively reflects systemic physiological deterioration. Compared to more complex scoring systems such as German Registry for Acute Aortic Dissection Type A (GERAADA), SOFA, or APACHE II. The MEWS is easier to implement for continuous bedside monitoring. Future studies should directly compare these systems in AAAD patients to clarify their relative utility.

This study implemented one-on-one nursing intervention within 24 hours postoperatively, with nurses providing continuous care and using real-time electrocardiogram (ECG) monitoring of patient conditions to collect vital signs every hour. Any changes in vital signs were immediately recorded, and necessary interventions were promptly taken. Additionally, each shift was staffed with senior-level nurses to ensure care quality and timeliness.

However, our study’s single-center, retrospective design may limit the generalizability of its findings. Variations in ICU configurations, staffing, and nursing quality across centers could affect the applicability of our results. Furthermore, some key clinical variables, such as lactate and malperfusion, were excluded due to incomplete records. Future multicenter prospective studies are necessary to validate our findings and incorporate a broader range of clinical predictors, enhancing the external validity of our conclusions.

Conclusions

The MEWS is a simple, accessible, and valuable scoring system for predicting postoperative mortality in AAAD patients. Both the absolute MEWS score at 12 hours and its dynamic trajectory provide important prognostic information. Incorporating MEWS into routine postoperative monitoring protocols may enhance risk stratification and facilitate timely clinical interventions.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-898/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-898/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-898/prf

Funding: This work was supported by 2022 Guangdong Province Science and Technology Special Fund (No. 2022-124-6).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-898/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of The First Affiliated Hospital of Shantou University Medical College (No. B-2022-286) and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Harris KM, Nienaber CA, Peterson MD, et al. Early Mortality in Type A Acute Aortic Dissection: Insights From the International Registry of Acute Aortic Dissection. JAMA Cardiol 2022;7:1009-15. [Crossref] [PubMed]
2. Erdoes G, Ahmed A, Kurz SD, et al. Perioperative hemostatic management of patients with type A aortic dissection. Front Cardiovasc Med 2023;10:1294505. [Crossref] [PubMed]
3. Carrel T, Sundt TM 3rd, von Kodolitsch Y, et al. Acute aortic dissection. Lancet 2023;401:773-88. [Crossref] [PubMed]
4. Doukas P, Dalibor N, Keszei A, et al. Factors Associated with Early Mortality in Acute Type A Aortic Dissection-A Single-Centre Experience. J Clin Med 2024;13:1023. [Crossref] [PubMed]
5. Bhadra S, Drgastin RH, Song HK, et al. Lactate as a Preoperative Predictor of Mortality in Patients Undergoing Emergency Type A Aortic Dissection Repair. J Pers Med 2025;15:211. [Crossref] [PubMed]
6. Ede J, Teurneau-Hermansson K, Ramgren B, et al. Outcomes following repair of acute type A aortic dissection in patients with cerebral malperfusion. Scand Cardiovasc J 2025;59:2514742. [Crossref] [PubMed]
7. Hashimoto O, Saito Y, Sasaki H, et al. Treatment strategies and in-hospital mortality in patients with type A acute aortic dissection and coronary artery involvement. J Thorac Cardiovasc Surg 2024;167:596-601.e3. [Crossref] [PubMed]
8. Mok WQ, Wang W, Liaw SY. Vital signs monitoring to detect patient deterioration: An integrative literature review. Int J Nurs Pract 2015;21:91-8. [Crossref] [PubMed]
9. Xie X, Huang W, Liu Q, et al. Prognostic value of Modified Early Warning Score generated in a Chinese emergency department: a prospective cohort study. BMJ Open 2018;8:e024120. [Crossref] [PubMed]
10. Kim DK, Lee DH, Lee BK, et al. Performance of Modified Early Warning Score (MEWS) for Predicting In-Hospital Mortality in Traumatic Brain Injury Patients. J Clin Med 2021;10:1915. [Crossref] [PubMed]
11. Kim DK, Lee DH, Lee BK. Role of the National Early Warning score and Modified Early Warning score for predicting mortality in geriatric patients with non-traumatic coma. Heliyon 2024;10:e28338. [Crossref] [PubMed]
12. Liu Y, Li Q, Zhang Y, et al. Application of the national early warning score (NEWS) in patients with acute aortic dissection: A case-control study. J Clin Nurs 2022;31:1620-7. [Crossref] [PubMed]
13. Mahmoodpoor A, Sanaie S, Saghaleini SH, et al. Prognostic value of National Early Warning Score and Modified Early Warning Score on intensive care unit readmission and mortality: A prospective observational study. Front Med (Lausanne) 2022;9:938005. [Crossref] [PubMed]
14. Knoery C, Barlas RS, Vart P, et al. Modified early warning score and risk of mortality after acute stroke. Clin Neurol Neurosurg 2021;202:106547. [Crossref] [PubMed]
15. Aygun H, Eraybar S. The role of emergency department triage early warning score (TREWS) and modified early warning score (MEWS) to predict in-hospital mortality in COVID-19 patients. Ir J Med Sci 2022;191:997-1003. [Crossref] [PubMed]
16. Myrstad M, Ihle-Hansen H, Tveita AA, et al. National Early Warning Score 2 (NEWS2) on admission predicts severe disease and in-hospital mortality from Covid-19 - a prospective cohort study. Scand J Trauma Resusc Emerg Med 2020;28:66. [Crossref] [PubMed]
17. Mei T, Xia Y, Huang P, et al. The National Early Warning Score (NEWS) Predicts the 28-Day Mortality in Patients With Severe Fever With Thrombocytopenia Syndrome: A Cross-Sectional Study. J Clin Nurs 2025;34:2992-3002. [Crossref] [PubMed]
18. Mhd Yunin NS, Tan TL. National Early Warning Score 2 (NEWS2) as a prognostic tool for adult patients in emergency department: A retrospective observational study. PLoS One 2025;20:e0326058. [Crossref] [PubMed]
19. Evans D, Hodgkinson B, Berry J. Vital signs in hospital patients: a systematic review. Int J Nurs Stud 2001;38:643-50. [Crossref] [PubMed]
20. Yang L, Song LX, Zhang YM, et al. Application of national early warning score in assessing postoperative illness severity in elderly patients with gastrointestinal illnesses. Technol Health Care 2024;32:1393-402. [Crossref] [PubMed]
21. Lee YS, Choi JW, Park YH, et al. Evaluation of the efficacy of the National Early Warning Score in predicting in-hospital mortality via the risk stratification. J Crit Care 2018;47:222-6. [Crossref] [PubMed]
22. Weekes AJ, Calienescerpa F, Goonan KL, et al. Association of initial national early warning score with clinical deterioration in pulmonary embolism. Thromb J 2025;23:49. [Crossref] [PubMed]
23. Veldhuis LI, Kuit M, Karim L, et al. Optimal timing for the Modified Early Warning Score for prediction of short-term critical illness in the acute care chain: a prospective observational study. Emerg Med J 2024;41:363-7. [Crossref] [PubMed]
24. Tan ADA, Permejo CC, Torres MCD. Modified Early Warning Score vs Cardiac Arrest Risk Triage Score for Prediction of Cardiopulmonary Arrest: A Case-Control Study. Indian J Crit Care Med 2022;26:780-5. [Crossref] [PubMed]
25. Yu M, Huang B, Liu P, et al. Detection of deteriorating patients after Whipple surgery by a modified early warning score (MEWS). Ann Transl Med 2019;7:574. [Crossref] [PubMed]