Echocardiography-derived pulmonary vascular resistance outperforms coagulation profiles and thromboelastography in predicting in-hospital mortality among surgical intensive care unit patients

Introduction

Patients in the surgical intensive care unit (SICU) face high mortality risks due to critical conditions such as severe trauma, complex surgeries, or sepsis. Despite advancements in life support technologies, in-hospital mortality remains elevated. Early identification of high-risk patients and precise interventions are crucial for improving prognosis. Traditional prognostic assessments predominantly rely on scoring systems such as the Acute Physiology and Chronic Health Evaluation II and the Sequential Organ Failure Assessment. However, these tools lack sufficient sensitivity to dynamic changes in specific pathophysiological processes and often fail to reflect the prognostic impact of core mechanisms like coagulation disorders and cardiopulmonary interactions.

Coagulation dysfunction is prevalent in SICU patients, characterized by a ‘thrombotic-hemorrhagic biphasic imbalance’. Conventional coagulation tests can assess coagulation factor activity but fail to comprehensively evaluate platelet function, fibrinolytic activity, or the overall coagulation process. Thromboelastography (TEG), through dynamic monitoring of the entire coagulation process from initiation to fibrinolysis, can identify hypercoagulability, hypofibrinogenemia, and hyperfibrinolysis. Its parameters demonstrate significant advantages in guiding transfusion strategies and predicting thrombotic events. In recent years, parameters from both conventional coagulation tests and TEG have emerged as valuable biomarkers for assessing coagulation status, detecting thrombosis, and predicting mortality in critically ill patients.

Pulmonary vascular resistance (PVR), as a quantitative indicator of right ventricular afterload, exhibits close correlations with pulmonary hypertension, right heart failure, and multiple organ dysfunction. Critically ill patients with elevated PVR face an increased mortality risk, with mechanisms involving hypoxic pulmonary vasoconstriction, microcirculatory dysfunction, and inflammatory cytokine cascades (1-4). Our preliminary research has also confirmed echocardiography-derived PVR is a simple yet powerful predictor of in-hospital mortality in critically ill patients.

While these indicators individually demonstrate clinical value, the predictive efficacy of single parameters may be limited. The interaction between PVR and the coagulation system may exacerbate organ failure through the ‘cardiopulmonary-coagulation axis’: increased right heart afterload induces hepatic congestion, which impairs coagulation factor synthesis, while microcirculatory dysfunction activates the extrinsic coagulation pathway via tissue hypoxia, forming a vicious cycle. Dysregulation of the ‘cardiopulmonary-coagulation axis’—manifested by elevated PVR, dynamic coagulation-fibrinolysis imbalance, and abnormal TEG parameters—may represent the key pathophysiological mechanism driving in-hospital mortality in SICU patients.

Based on this background, this study aimed to (I) identify the independent risk factors among PVR, coagulation function, and TEG parameters for in-hospital mortality in SICU patients; (II) compare their predictive efficacy; and (III) develop a multivariate prediction model to optimize risk stratification. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-899/rc).

Methods

Study design and participants

The study protocol was approved by the Human Research Ethics Committee of Yangpu Hospital, School of Medicine, Tongji University (No. LL-2023-SCI-018). Written informed consent was obtained from all participants or their legally authorized representatives prior to enrollment. The investigation strictly adhered to the ethical principles outlined in the Declaration of Helsinki and its subsequent amendments. This prospective observational study was conducted in a SICU between January 2025 and April 2025. Sixty-eight consecutive critically ill patients who underwent concurrent transthoracic echocardiography (TTE), coagulation testing, and TEG on the same day were enrolled. Exclusion criteria included: (I) suboptimal quality of standardized echocardiographic images; (II) incomplete medical records; and (III) missing essential laboratory data.

Coagulation test

The coagulation tests—including prothrombin time (PT), prothrombin activity (PTA), international normalized ratio (INR), activated partial thromboplastin time (APTT), thrombin time (TT), fibrinogen, plasma antithrombin activity (PAA), fibrin degradation products (FDP), and D-dimer, were performed using fully automated coagulation analyzers (Sysmex CS-2500, Kobe, Hyogo, Japan). Samples of 3.2% sodium citrate-anticoagulated whole blood (1:9 blood-to-anticoagulant ratio) were centrifuged at 3,000 rpm for 15 minutes to obtain platelet-poor plasma. The analyzer automatically performed reagent mixing, incubation at 37 ℃, and clot detection via optical/mechanical methods. Results were reported in seconds, concentration units (g/L, mg/dL, or µg/mL), or as a percentage (%).

TEG

TEG was performed using a TEG^® 5000 or TEG^® 6s analyzer (Haemonetics Corporation, Boston, MA, USA) with kaolin-activated whole blood, recalcified by 0.2 M calcium chloride. Real-time parameters included: reaction time (R-time) (clot initiation; normal 5–10 min), kinetics time (K-time) (fibrin crosslinking; normal 1–3 min), alpha angle (clot formation rate; normal 53–72°), maximum amplitude (MA) (reflecting platelet function and fibrinogen levels; normal 50–70 mm), coagulation index (normal −3 to 3), and global clot strength (G value; normal 4,500–11,000 dyne/cm²).

Conventional TTE

All TTE studies were conducted by a board-certified radiologist (possessing three decades of expertise in echocardiography) utilizing a Philips EPIQ 7C ultrasound system (Amsterdam, Netherlands) with an X5-1 PureWave xMATRIX transducer (1–5 MHz). Examinations were triggered by the onset of sudden respiratory distress with hypoxemia accompanied by coagulation abnormalities (e.g., significant D-dimer elevation and/or TEG abnormalities). Assessments were typically performed upon SICU admission or within 24–48 hours postoperatively, and in some cases up to one week after surgery. Measurements were acquired in alignment with current European Association of Cardiovascular Imaging (EACVI)/American Society of Echocardiography (ASE) guidelines (5,6), including: left ventricular end-diastolic anteroposterior diameter (D_LV), right ventricular end-diastolic anteroposterior diameter (D_RV), left atrial end-systolic transverse diameter (D_LA), right atrial end-systolic transverse diameter (D_RA), left ventricular ejection fraction (LVEF), the ratio of the peak early diastolic transmitral filling velocity to the peak early diastolic lateral mitralannulus tissue velocity (mitral E/e’ ratio), the ratio of the peak early diastolic transtricuspid filling velocity to the peak early diastolic lateral tricuspid annulus tissue velocity (tricuspid E/e’ ratio), tricuspid annular plane systolic excursion (TAPSE), the peak tricuspid regurgitation velocity (TRV_max), the velocity time integral at right ventricular outflow tract (VTI_RVOT), the velocity time integral at left ventricular outflow tract (VTI_LVOT), the velocity time integral at pulmonary artery (VTI_PA), left ventricular outflow tract diameter (D_LVOT), and pulmonary artery diameter (D_PA). Derived hemodynamic parameters were calculated as follows: pulmonary artery systolic pressure (PASP) was estimated using the modified Bernoulli equation: PASP (mmHg) = (4 × TRV_max)² + right atrial pressure. Right atrial pressure was estimated based on inferior vena cava diameter and collapsibility. PVR was calculated using the empirically validated formula previously described by Abbas et al. (7): PVR (Wood units) = TRV_max/VTI_RVOT × 10 + 0.16. The methodology is further illustrated in Figure 1 of the manuscript. Left ventricular output (LVO) (mL/min) = left ventricular stroke volume × heart rate (HR) = (D_LVOT/2)² × π × VTI_LVOT × HR; right ventricular output (RVO) (mL/min) = right ventricular stroke volume × HR = (D_PA/2)² × π × VTI_PA × HR; right ventricular-pulmonary arterial (RV-PA) coupling (mm/mmHg) = TAPSE/PASP (8,9). RVO/LVO ratio was subsequently determined from these calculated values.

Figure 1 TTE with the measurement of (A) the TRV_max and (B) the VTI_RVOT, illustrating the measurement methodology of PVR: PVR = TRV_max/VTI_RVOT × 10 + 0.16 = [(2.59 + 2.65)/2]/[(16.2 + 16.0)/2] × 10 + 0.16 = 2.62/16.1 × 10 + 0.16 = 1.787 (Wood units). PVR, pulmonary vascular resistance; TRV_max, peak tricuspid regurgitation velocity; TTE, transthoracic echocardiography; VTI_RVOT, velocity time integral at right ventricular outflow tract.

Statistical analysis

Normally distributed continuous variables were expressed as mean ± standard deviation (SD) and compared via independent t-tests. Nonparametric data were reported as median (interquartile range) and analyzed using Mann-Whitney U tests. Categorical variables were summarized as counts (percentages) and evaluated via Fisher’s exact test or Chi-squared test. Multivariable logistic regression identified independent predictors of in-hospital mortality in critically ill patients. For continuous variables, collinearity diagnosis was performed using linear regression analysis, and variables with a variance inflation factor greater than 10 were excluded. Diagnostic performance of predictors was assessed using receiver operating characteristic (ROC) curve analysis, and the area under the curve (AUC) was reported. A two-tailed P value <0.05 denoted statistical significance. Analyses were performed using SPSS 19.0 (IBM Corp., Armonk, NY, USA) and MedCalc 16.8.4 (Ostend, Belgium).

Results

Characteristics of included patients

From an initial cohort of 68 SICU patients, seven were excluded due to insufficient image quality or missing data, yielding a final study population of 61 individuals. These patients were stratified into a survivor group (n=43) and a non-survivor group (n=18) based on in-hospital mortality status. Comprehensive demographic and clinical characteristics are summarized in Table 1. Non-survivors were significantly older and exhibited smaller body surface area (BSA) and lower body mass index (BMI) compared to survivors (P<0.05), though no sex-based disparities were observed (P>0.05). Both groups displayed similar prevalence rates of comorbidities, including cerebral hemorrhage, pneumonia, pulmonary embolism, fractures, myocardial infarction, gastrointestinal perforation, intestinal obstruction, acute hepatic failure, renal failure, diabetes, malignancy, and sepsis (all P>0.05). Characteristics of coagulation assay, TEG, and echocardiography are listed in Table 2. In coagulation assays, non-survivors demonstrated significantly prolonged PT and APTT, higher INR, elevated levels of fibrinogen, FDP, and D-dimer, as well as reduced platelet aggregation activity (PAA) and plasma thromboplastin activity (PTA) compared to survivors (P<0.05). TEG revealed that non-survivors exhibited longer K-time, larger alpha angle, and lower coagulation index than survivors (P<0.05). Echocardiographic analysis revealed that non-survivors had smaller cardiac chamber dimensions, impaired systolic function, and reduced cardiac output relative to survivors (P<0.05). Conversely, PASP and PVR were significantly elevated in non-survivors (P<0.05). These hemodynamic changes contributed to decreased RV-PA coupling.

Univariate/multivariate logistic regression analysis

Univariate logistic regression analysis identified all parameters showing significant differences in Tables 1,2—except for K-time, alpha angle, coagulation index, and tricuspid E/e’ ratio—as risk factors for in-hospital mortality (Table 3). However, multivariate logistic regression analysis demonstrated that only APTT, FDP, and PVR emerged as independent risk factors for in-hospital mortality in this population (Table 4).

ROC curve analysis

As shown in Table 5 and Figure 2, both PVR (AUC =0.865; cutoff, 2.175 Wood units; sensitivity, 97.94%; specificity, 73.36%) and the combined model incorporating PVR, APTT, and FDP (AUC =0.978; cutoff, 0.593; sensitivity, 92.59%; specificity, 97.40%) showed significantly stronger predictive performance for in-hospital mortality compared to APTT (AUC =0.749; cutoff, 29.500 s; sensitivity, 83.52%; specificity, 71.18%) and FDP (AUC =0.645; cutoff, 26.000 µg/L; sensitivity, 70.21%; specificity, 69.87%) alone (z-statistic, P<0.01). Additionally, APTT exhibited stronger predictive efficacy than FDP for in-hospital mortality in SICU patients (z-statistic, P<0.05).

Figure 2 ROC curve demonstrating the ability of PVR, APTT, FDP and the combined model incorporating PVR, APTT, and FDP to predict in-hospital mortality among patients in SICU. APTT, activated partial thromboplastin time; AUC, area under the curve; FDP, fibrin degradation products; PVR, pulmonary vascular resistance; ROC, receiver operating characteristic; SICU, surgical intensive care unit.

Discussion

Our study showed that PVR, APTT, and FDP independently predict in-hospital mortality in SICU patients. Importantly, PVR outperformed APTT or FDP alone in predicting outcomes. These results highlight new approaches for improving risk assessment and personalized care in critically ill surgical patients.

Right heart catheterization (RHC) remains the gold standard for measuring PVR. However, its invasive nature carries risks of vascular injury, pneumothorax, thrombosis, and even life-threatening arrhythmias, compounded by procedural complexity, limited dynamic monitoring capability, high costs, and challenges in real-time bedside implementation. In this study, we employed bedside echocardiography to noninvasively estimate PVR based on tricuspid regurgitation velocity and right ventricular outflow tract flow parameters. This approach offers convenience, rapidity, high reproducibility, and zero procedural risks-particularly advantageous for hemodynamically unstable critically ill patients or those requiring dynamic therapeutic response monitoring. In clinical practice, echocardiography is widely utilized for initial screening, serial follow-up, and managing critically ill patients (e.g., in SICU/ICU settings) who are unfit for invasive catheterization.

Our study demonstrated that echocardiography-derived PVR outperforms conventional coagulation tests (e.g., FDP, APTT) and TEG parameters in predicting in-hospital mortality among SICU patients. The superiority of PVR can be attributed to three key aspects: (I) physiological integration: PVR quantifies right ventricular afterload, directly reflecting the core mechanism of cardiopulmonary interaction dysfunction (10,11). Elevated PVR induces hepatic congestion (impairing coagulation factor synthesis) and microcirculatory dysfunction (activating extrinsic coagulation pathways), thereby driving a vicious ‘cardiopulmonary-coagulation axis’ (12-16). In contrast, conventional coagulation tests assess isolated aspects (e.g., fibrinolysis or clotting factor activity), while TEG lacks sensitivity to cardiopulmonary impairments despite its dynamic evaluation of coagulation. (II) Operational and monitoring advantages: PVR is measured non-invasively via echocardiography, avoiding repeated blood sampling or complex workflows, making it ideal for hemodynamically unstable patients. Coagulation tests require frequent blood draws and are susceptible to anticoagulant interference, whereas TEG interpretation demands expertise. PVR dynamically responds to therapeutic interventions (e.g., fluid management or vasoactive drugs), while coagulation markers and TEG parameters often lag behind clinical deterioration, limiting their early warning utility. (III) Clinical validation and limitations: consistent with previous studies, our findings indicate that a PVR >2.175 Wood units increased the risk of in-hospital mortality. Although coagulation abnormalities (e.g., elevated FDP) correlate with thrombotic events, their mortality association is confounded by comorbidities (e.g., infection or organ dysfunction). TEG parameters (e.g., MA value) aid transfusion strategies but exhibit limited predictive power for multi-organ failure.

In our study, the exceptionally high odds ratio (OR) for PVR (284.280) primarily stems from two key factors. First, in our multivariate logistic regression model, PVR was entered as a continuous variable (in Wood units), rather than as a categorical variable. Therefore, the OR represents the increase in mortality odds for each 1-unit increase in PVR. Given that PVR in critically ill patients can range from near-normal to severely elevated values (e.g., from 2 to over 8 Wood units), a single-unit change represents a substantial physiological insult, which is reflected in the high OR. Second, and most importantly, our study had a limited number of mortality events (n=18). In logistic regression with a small sample size and a strong predictor, the maximum likelihood estimation can produce inflated ORs with very wide confidence intervals (CIs) (23.887–3,383.278), indicating a point estimate that is statistically significant but imprecise. This suggests that while PVR is an extremely powerful predictor, the exact magnitude of its effect needs to be validated in a larger cohort.

The relatively small cohort, particularly the limited number of outcome events (18 deaths), directly impacts the stability and generalizability of our statistical models. It increases the risk of overfitting in our multivariate logistic regression, where the model may describe random error in our specific dataset rather than the underlying true relationships. This is a plausible explanation for the implausibly high OR for PVR. The small sample size leads to reduced statistical power, increasing the likelihood of type II errors (failing to identify other true predictors of mortality). Lastly, it results in wide CIs for our effect estimates, as seen with PVR, indicating substantial uncertainty about the precise strength of the association. Therefore, our findings, while highly suggestive, should be considered preliminary and require confirmation in a larger, prospective multicenter study.

Several other limitations warrant consideration. First, the accuracy of PVR measurement depends on the sonographer’s expertise, and measurements in some patients (e.g., those with severe obesity, emphysema, or forced posture) may yield unreliable measurements due to poor acoustic windows. Second, this study did not fully exclude other factors affecting mortality (such as underlying cardiopulmonary diseases and duration of mechanical ventilation), which may partially obscure the independent predictive value of PVR. Third, this is a single-center retrospective study, and multicenter prospective cohort studies are required to validate the generalizability of the PVR model and the combined model incorporating PVR, APTT, and FDP, particularly regarding performance variations among different subgroups (e.g., sepsis vs. trauma).

Conclusions

This is the first study report on PVR, coagulation tests, and TEG parameters as predictors of in-hospital mortality among SICU patients. Echocardiography-derived PVR, as a non-invasive and dynamic cardiopulmonary functional assessment parameter, demonstrates superior clinical utility over coagulation tests and TEG in predicting in-hospital mortality among SICU patients. Its advantages stem from its holistic reflection of the pathophysiological mechanisms underlying the ‘cardiopulmonary-coagulation axis’ and procedural simplicity. The PVR model and its combination with coagulation/TEG parameters hold significant clinical value for optimizing risk stratification in SICU patients and guiding personalized interventions.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-899/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-899/dss

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-899/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 protocol was approved by the Human Research Ethics Committee of Yangpu Hospital, School of Medicine, Tongji University (No. LL-2023-SCI-018). Written informed consent was obtained from all participants or their legally authorized representatives prior to enrollment. The investigation strictly adhered to the ethical principles outlined in the Declaration of Helsinki and its subsequent amendments.

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. Norderfeldt J, Liliequist A, Eksborg S, Frostell C, Eriksson MJ, Adding C, Agvald P, Lönnqvist PA. Severe Covid-19 and acute pulmonary hypertension: 24-month follow-up regarding mortality and relationship to initial echocardiographic findings and biomarkers. Acta Anaesthesiol Scand 2023;67:206-12. [Crossref] [PubMed]
2. Dandel M. Heart-lung interactions in COVID-19: prognostic impact and usefulness of bedside echocardiography for monitoring of the right ventricle involvement. Heart Fail Rev 2022;27:1325-39. [Crossref] [PubMed]
3. Maron BA, Brittain EL, Hess E, Waldo SW, Barón AE, Huang S, Goldstein RH, Assad T, Wertheim BM, Alba GA, Leopold JA, Olschewski H, Galiè N, Simonneau G, Kovacs G, Tedford RJ, Humbert M, Choudhary G. Pulmonary vascular resistance and clinical outcomes in patients with pulmonary hypertension: a retrospective cohort study. Lancet Respir Med 2020;8:873-84. [Crossref] [PubMed]
4. Crawford TC, Leary PJ, Fraser CD 3rd, Suarez-Pierre A, Magruder JT, Baumgartner WA, Zehr KJ, Whitman GJ, Masri SC, Sheikh F, De Marco T, Maron BA, Sharma K, Gilotra NA, Russell SD, Houston BA, Ramu B, Tedford RJ. Impact of the new pulmonary hypertension definition on heart transplant outcomes: expanding the hemodynamic risk profile. Chest 2020;157:151-61. [Crossref] [PubMed]
5. Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, Flachskampf FA, Foster E, Goldstein SA, Kuznetsova T, Lancellotti P, Muraru D, Picard MH, Rietzschel ER, Rudski L, Spencer KT, Tsang W, Voigt JU. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. Eur Heart J Cardiovasc Imaging 2015;16:233-70. [Crossref] [PubMed]
6. Mitchell C, Rahko PS, Blauwet LA, Canaday B, Finstuen JA, Foster MC, Horton K, Ogunyankin KO, Palma RA, Velazquez EJ. Guidelines for performing a comprehensive transthoracic echocardiographic examination in adults: recommendations from the american society of echocardiography. J Am Soc Echocardiogr 2019;32:1-64. [Crossref] [PubMed]
7. Abbas AE, Fortuin FD, Schiller NB, Appleton CP, Moreno CA, Lester SJ. A simple method for noninvasive estimation of pulmonary vascular resistance. J Am Coll Cardiol 2003;41:1021-7. [Crossref] [PubMed]
8. Rako ZA, Kremer N, Yogeswaran A, Richter MJ, Tello K. Adaptive versus maladaptive right ventricular remodelling. ESC Heart Fail 2023;10:762-75. [Crossref] [PubMed]
9. Guazzi M, Dixon D, Labate V, Beussink-Nelson L, Bandera F, Cuttica MJ, Shah SJ. RV contractile function and its coupling to pulmonary circulation in heart failure with preserved ejection fraction: stratification of clinical phenotypes and outcomes. JACC Cardiovasc Imaging 2017;10:1211-21. [Crossref] [PubMed]
10. Steingrub JS, Tidswell M, Higgins TL. Hemodynamic consequences of heart-lung interactions. J Intensive Care Med 2003;18:92-9. [Crossref] [PubMed]
11. Romand JA, Donald FA, Suter PM. Cardiopulmonary interactions in acute lung injury: clinical and prognostic importance of pulmonary hypertension. New Horiz 1994;2:457-62.
12. Wang Y, Wang M, Zhang GH, He FL, Ou XJ, Wang XM, Jia JD. Liver manifestation of circulatory disorders. Zhonghua Gan Zang Bing Za Zhi 2022;30:362-6. [Crossref] [PubMed]
13. Møller S, Bernardi M. Interactions of the heart and the liver. Eur Heart J 2013;34:2804-11. [Crossref] [PubMed]
14. Xanthopoulos A, Starling RC, Kitai T, Triposkiadis F. Heart failure and liver disease: cardiohepatic interactions. JACC Heart Fail 2019;7:87-97. [Crossref] [PubMed]
15. Savale L, Kularatne M, Roche A, Pichon J, Baron A, Boucly A, Sitbon O, Humbert M. Management of Acutely Decompensated Pulmonary Hypertension. Semin Respir Crit Care Med. 2023;44:762-70. [Crossref] [PubMed]
16. Xie R, Tan D, Liu B, Xiao G, Gong F, Zhang Q, Qi L, Zheng S, Yuan Y, Yang Z, Chen Y, Fei J, Xu D. Acute respiratory distress syndrome (ARDS): from mechanistic insights to therapeutic strategies. MedComm (2020) 2025;6:e70074.