Clinical Cardiovascular Research

Abstract

Objective. To identify factors associated with in-hospital mortality and to develop and validate a formula for additional risk of early mortality in patients (pts) with high-risk pulmonary embolism (PE).

Materials and Methods. The study consisted of two phases. In Phase I, a retrospective analysis was conducted on the medical records of 113 pts with high-risk PE who were consecutively hospitalized at Kharkiv City Clinical Hospital No. 8 from 01.01. 2017 to 01.01.2023. The mean age of pts was 65.14±14.13 years, with 45 (38.8%) males and 68 (60.2%) females. Based on the course of the disease, the pts were divided into two groups: Group 1 – 63 pts (55.8%) who improved and were discharged, and Group 2 – 50 pts (44.2%) who died during hospitalization. Phase II of the study involved a multicenter cohort (9 centers in Ukraine that provided data on pts with high-risk pulmonary embolism hospitalized from January 2023 to December 2023). This phase included 75 pts with high-risk PE, with mean age of 60.60±13.40 years; 38 (50.7%) were male, and 37 (49.3%) were female. Group I consisted of 65 patients (80%) discharged with improvement, Group II comprised 15 patients (20%) who died during hospitalization. Clinical and anamnestic and laboratory-instrumental data were analyzed, statistical analysis of the resulting figures was performed.

Results and Discussion. In Phase I of the study, a formula for additional risk stratification of early mortality in pts with high-risk PE was developed. Using ROC analysis, the sensitivity of the resulting prognostic formula was found to be 64.0%, and the specificity was 92.1%. Based on this formula, we proposed a simplified scale called SBAFS (S - saturation, B - bifurcation, A - arterial hypotension, F - fraction of ejection, S - female sex). In Phase II of the study, the sensitivity and specificity of the obtained formula were validated using ROC analysis: the sensitivity was 93.3%, and the specificity was 74.6%, confirming the results of Phase I. When evaluating the proposed simplified SBAFS scale, it was also found that in patients of Group II, the mean score on the scale was (2.53±0.83) and was higher than that in Group I: (1.57±0.96), p=0.0006.

Conclusions. The main factors associated with the risk of in-hospital mortality in pts with high-risk PE are decreased oxygen saturation, thrombus localization in the bifurcation of the pulmonary artery as determined by СТPA angiography, arterial hypotension, female sex, and decreased left ventricular ejection fraction as assessed by echocardiography. The previously proposed SBAFS formula confirmed its significance upon validation involving data on PE pts from 9 regions of Ukraine, with a sensitivity of 93% and specificity of 75%.

Introduction

Pulmonary embolism (PE) ranks third, following myocardial infarction and stroke, in the structure of cardiovascular mortality [1,2]. A significant number of risk factors and the presence of comorbidities contribute to substantial variability in both the course and outcomes of PE [3,6]. In light of this, current guidelines recommend conducting risk stratification for PE, considering the predicted early mortality [1,6]. It has been established that in pts with high-risk PE, the clinical course of the disease is the most severe, and the prognosis is extremely unfavorable; hospital mortality rates in various studies can range from 25.4% to 70% [3,4]. Currently, patients with high-risk PE are classified as those with hemodynamic instability, which is defined by criteria such as cardiac arrest requiring cardiopulmonary resuscitation, the presence of obstructive shock, or hypotension lasting more than 15 minutes with a systolic blood pressure (SBP) of less than 90 mm Hg, or a decrease in SBP of 40 mm Hg from baseline without obvious reasons [1]. The question of additional clinical and instrumental factors related to in-hospital mortality in pts with high-risk PE remains insufficiently studied.

The aim of the study was to identify factors associated with the risk of in-hospital mortality and to develop and validate a formula for additional risk of early death in pts with high-risk pulmonary embolism.

Materials And Methods

In Phase I of our study, a retrospective analysis of the medical records of 113 patients with high-risk pulmonary embolism (PE) was conducted. These pts were consecutively hospitalized at Kharkiv City Clinical Hospital No. 8 from January 1, 2017, to January 1, 2023 (Table 1). The mean age of the pts was 65.14±14.13 years, with 45 men (38.8%) and 68 women (60.2%).

Phase II of the study was conducted on a multicenter cohort involving centers from Kyiv, Sumy, Kremenchuk, Poltava, Mykolaiv, Dnipro, Lutsk and Kharkiv. These centers provided information on pts with high-risk PE who were treated from January 2023 to December 2023 (Table 4). In total, 75 pts with high-risk PE were included in Phase II, with a mean age of 60.60±13.40 years; among them, there were 38 men (50.7%) and 37 women (49.3%).

The diagnosis of PE was verified through computed tomography pulmonary angiography (CTPA) and/or post-mortem findings, while risk assessment was performed in accordance with the European guidelines for the treatment of pts with PE (2019) [1].

Statistical Analysis. The statistical analysis of the obtained data was performed using the Statistica 10.0 software package (StatSoft Inc, USA) and Microsoft Office Excel 2013. Since the preliminary analysis of quantitative characteristics did not reveal deviations from a normal distribution, they are presented as mean ± standard deviation (M ± σ). The Mann-Whitney U-test was used to compare the means of two samples. Differences were considered statistically significant at p < 0.05. Univariate and multivariate logistic regression analyses were employed to identify predictors of the combined clinical endpoint. The β coefficient, standard error, odds ratio (OR), and 95% confidence interval (CI) were calculated for each factor. Discrimination was assessed using the area under the receiver operating characteristic curve (AUC).

Results and Discussion

In the first phase pts with high-risk PE were divided into two groups: Group 1 – 63 (55.8%), who were discharged with improvement; Group 2 – 50 (44.2%), who died. (Table 1). Nearly 70% of pts from Group II died within 1,5 hours of arriving at the hospital.

Table 1: Comparative Characteristics of High-Risk Patients Based on Prognosis

Note. AF – atrial fibrillation, AH – arterial hypertension, CTPA – computed tomography of pulmonary arteries, DM – diabetes mellitus, DVT – deep vein thrombosis, HR – heart rate, LA – left atrium, LPA – left pulmonary artery, LV EDV – end diastolic volume of left ventricle, LV ESV – end systolic volume of left ventricle, LV EF – ejection fraction of left ventricle, MPAP – mean pressure in pulmonary artery, RA – right atrium, RPA – right pulmonary artery, RV – right ventricle , SBP – systolic blood pressure, SpO2 – blood oxygen saturation, SVT – superficial vein thrombosis.

There were more women in Group 2 (p=0,006). Also pts of Group 2 were more severe having significantly lower SpO2 values (p=0.000043) and SBP (p=0.00001). In Group 2, risk factors of PE, such as obesity (p=0.0005) and DVT (p=0.007), were more common than in Group 1. According to CTPA, pts of Group 2 had more bilateral segmental lesions (p=0.025) and (according to echocardiography) lower LVEF (p=0.008).

We used uni- and multivariate analyses to find out independent prognostic factors associated with in-hospital mortality in pts with high-risk PE (Table 4). The regression analysis revealed a significant independent association between in-hospital mortality and the presence of thrombus in the bifurcation of pulmonary trunk (p=0.015) and female sex (p=0.0048). Additionally, an inverse independent association was found between mortality and SBP (p=0.0008), SpO₂ (p=0.0014) and LVEF (p=0.011). There were more women in Group 2 (p=0,006). Also pts of Group 2 were more severe having significantly lower SpO2 values (p=0.000043) and SBP (p=0.00001). In Group 2, risk factors of PE, such as obesity (p=0.0005) and DVT (p=0.007), were more common than in Group 1. According to CTPA, pts of Group 2 had more bilateral segmental lesions (p=0.025) and (according to echocardiography) lower LVEF (p=0.008).

We used uni- and multivariate analyses to find out independent prognostic factors associated with in-hospital mortality in pts with high-risk PE (Table 4). The regression analysis revealed a significant independent association between in-hospital mortality and the presence of thrombus in the bifurcation of pulmonary trunk (p=0.015) and female sex (p=0.0048). Additionally, an inverse independent association was found between mortality and SBP (p=0.0008), SpO₂ (p=0.0014) and LVEF (p=0.011).

Table 2: Factors Affecting In-Hospital Mortality in High-Risk Patients with Pulmonary Embolism (Stage I)

Note. AH – arterial hypertension, LV EF – ejection fraction of left ventricle, SBP – systolic blood pressure, SpO2 – blood oxygen saturation.

To determine the threshold levels of variable indicators associated with mortality, we also performed ROC analysis. The results indicate that the critical thresholds are: SpO₂ below 78% (sensitivity - 54.0%, specificity - 87.3%, p < 0.0001), SBP below 80 mmHg (sensitivity - 84.0%, specificity - 54.7%, p < 0.0001), and LVEF below 52% according to echocardiography (sensitivity - 66.0%, specificity - 82.5%, p < 0.0001) (Fig.1).

Fig 1: ROC Curve for the Predictive Formula of Logistic Regression to Determine (Positive or Negative) Prognosis.

Based on our results, we created a formula for risk stratification of in-hospital mortality for pts with high-risk PE:

Y=(exp(17,47-0,11*SpO2+1,58* Thrombus in bifurcation of PA-0,09*SBP

+1,66*female sex -0,08*LV EF))/(1+exp (17,47-0.11*SpO2+1.58* Thrombus in bifurcation of PA -0,09*SBP +1,66*female sex -0,08*LV EF)).

Calculating the individual prognosis for pts with high-risk PE, the additional risk of in-hospital mortality is higher if the value of Y > 0.5, and does not increase when Y < 0.5.

Using ROC analysis, it was established that the sensitivity of the obtained prognostic formula is 64.0%; specificity is 92.1% (area under the ROC curve is 0.7861; 95% confidence interval 0.784-0.93, p < 0.0001).

Based on the analysis and verification of the formula, we proposed the SBAFS scale (S – saturation, B – bifurcation, A – arterial hypotension, F – LV ejection fraction, S – female sex) for calculating the additional risk of in-hospital mortality in high-risk patients with PE (Table 3).

Table 3: Scale for Determining Additional Mortality Risk in High-Risk Pulmonary Embolism Patients (SBAFS)

Note. LV EF – ejection fraction of left ventricle, SpO2 – blood oxygen saturation.

In order to verify and check our formula we organized the second phase of our study based on a multicenter cohort. Patients were divided into two subgroups:  I - 65 (80%) pts who were discharged with improved status; II - 15 (20%) pts who died during hospitalization (Table 4). The SBAFS scale was evaluated in Phase 2.

Table 4: Indicators of the SBAFS Scale in High-Risk Pulmonary Embolism Patients Based on Prognosis (Stage II of the Study).

Note. SBP – systolic blood pressure, SpO2 – blood oxygen saturation.

As it was shown in the first stage of our study, SBP, LVEF and SpO2 were significantly lower in subgroup II pts compared to subgroup I (Table 5).

When verifying the formula, it was established that in subgroup II pts, the average value of the indicator Y was significantly higher than in subgroup I: (0.785 ± 0.250) vs. (0.306 ± 0.330), p = 0.00002. Additionally, the proportion of pts with Y > 0.5 was also significantly higher: 13 (86.7%) vs. 16 (26.7%), respectively, p = 0.00001, confirming the presence of an additional risk of early death in subgroup II pts.

To assess the sensitivity and specificity of the proposed formula, we performed ROC analysis for the Phase 2 of our study (Figure 2). It was shown that in the multicenter cohort, the sensitivity of our prognostic formula was 93.3%, and the specificity was 74.6% (area under the ROC curve was 0.803; 95% confidence interval 0.690-0.889, p < 0.001). Thus, during the verification of the formula in the second phase of our study, the sensitivity and specificity not only reproduced the results from the first stage but even exceeded the previous findings. Additionally, the sensitivity and specificity of indicators such as SBP, LVEF and SpO2, as evaluated by ROC analysis conducted in both stages of the study, were significant and comparable.

Fig 2: ROC Curve for the Predictive Formula of Logistic Regression to Determine (Positive or Negative) Prognosis (Phase2).

While evaluating the proposed SBAFS scale, we also found that the average score for subgroup II on the scale was (2.53 ± 0.83) and was significantly higher than in subgroup I: (1.57 ± 0.96), (p = 0.0006). In subgroup II there was also a significantly greater proportion of pts who had more than 3 points according to SBAFS.

As we mentioned before lower levels of SpO2, SBP and bifurcational lesions verified by CTPA are associated with increased in-hospital mortality in pts with high-risk PE. This statement is logical as those factors are directly pathophysiologically linked to the hemodynamic overload and dysfunction of the right ventricle. Moreover, these factors are also used in various scales for predicting the risk of death in patients with acute PE and are consistent with the results of other studies [3, 5, 13].

Taking into account female sex according to our previous studies (as well as other scientific works), it is not an additional risk factor in patients with non-high-risk PE [7, 14]. However, in pts with high-risk PE, female sex may significantly serve as an additional factor for unfavorable short-term prognosis, warranting further investigation. Analysis of registries has shown that high-risk PE occurs more frequently in women than in men (14.6% vs. 9.2%, respectively; p = 0.0002) [4]. Possible explanations for these findings may include sex differences in risk factors for VTE and anatomical characteristics of the cardiovascular and respiratory systems, leading to more significant right ventricular decompensation in women during PE [15, 16].

Thus, the results of our study suggest that, in addition to known factors associated with unfavorable 30-day prognosis in high-risk PE patients, additional factors associated with in-hospital mortality in such patients are: SpO2 below 78%, SBP below 80 mmHg, LVEF below 52%, saddle PE verified by CTPA, female sex. The formula and SBAFS scale, which we developed, can be used for determining additional mortality risk in pts with high-risk PE based in Ukraine.

Conclusion

1. The primary factors associated with the risk of hospital mortality in pts with high-risk pulmonary embolism include reduced oxygen saturation, thrombus localization in the bifurcation of the pulmonary arteries as determined by computed tomography of pulmonary arteries, arterial hypotension, female sex, and decreased left ventricular ejection fraction according to echocardiography.
2. A formula for stratifying the risk of in-hospital mortality in patients with high-risk PE has been developed, which, according to the results of the ROC analysis, provides a sensitivity of 69.6% and a specificity of 85.3%. The proposed simplified SBAFS scale indicates that a score of ≥2 points signifies an additional high risk of hospital mortality in patients with high-risk PE.
3. The feasibility of using the simplified assessment scale for patients with high-risk pulmonary embolism has been confirmed in a multicenter cohort. The total score was significantly higher in deceased patients—2.5 compared to 1.6 (p = 0.0006).

Declarations

Limitations: In this study, invasive treatment methods for patients with pulmonary embolism were not applied.

Conflict of Interest: The authors declare that there are no conflicts of interest.

Author Contributions: Concept and study design, article editing, conclusion formulation – Vira Tseluyko; statistical data analysis – L. Yakovleva; article writing – L. Yakovleva, N. Mishchuk, M. Kurinna, R. Askierov; data collection – M. Kurinna, L. Kharchenko, R. Askierov.

References

1. Konstantinides, S. V., Meyer, G., Becattini, C., Bueno, H., Geersing, G. J., Harjola, V. P., ... & Zamorano, J. L. (2020). 2019 ESC Guidelines for the diagnosis and management of acute pulmonary embolism developed in collaboration with the European Respiratory Society (ERS) The Task Force for the diagnosis and management of acute pulmonary embolism of the European Society of Cardiology (ESC). European heart journal, 41(4), 543-603.
   View at Publisher | View at Google Scholar
2. Martin, K. A., Molsberry, R., Cuttica, M. J., Desai, K. R., Schimmel, D. R., & Khan, S. S. (2020). Time trends in pulmonary embolism mortality rates in the United States, 1999 to 2018. Journal of the American Heart Association, 9(17), e016784.
   View at Publisher | View at Google Scholar
3. Tak, T., Karturi, S., Sharma, U., Eckstein, L., Poterucha, J. T., & Sandoval, Y. (2019). Acute pulmonary embolism: contemporary approach to diagnosis, risk-stratification, and management. International Journal of Angiology, 28(02), 100-111.
   View at Publisher | View at Google Scholar
4. Tanabe, Y., Yamamoto, T., Murata, T., Mabuchi, K., Hara, N., Mizuno, A., ... & Nagao, K. (2018). Gender differences among patients with acute pulmonary embolism. The American Journal of Cardiology, 122(6), 1079-1084.
   View at Publisher | View at Google Scholar
5. THeroux, C. D., ALioTTA, J. M., & MuLLin, C. J. (2019). High-risk pulmonary embolism: current evidence-based practices. RI Med J, 102(10), 43-47.
   View at Publisher | View at Google Scholar
6. Authors/Task Force Members, Torbicki, A., Perrier, A., Konstantinides, S., Agnelli, G., Galiè, N., ... & Vachiery, J. L. (2008). Guidelines on the diagnosis and management of acute pulmonary embolism: the Task Force for the Diagnosis and Management of Acute Pulmonary Embolism of the European Society of Cardiology (ESC). European heart journal, 29(18), 2276-2315.
   View at Publisher | View at Google Scholar
7. Tseluyko, V. I., Yakovleva, L. M., Askierov, R. N., Kurinna, M. V., Kharchenko, L. V., & Shylo, N. G. (2023). Risk stratification of poor short-term prognosis in patients with high-risk pulmonary embolism. Emergency medicine, 19(5), 361-369.
   View at Publisher | View at Google Scholar
8. Wüstner, M., Radzina, M., Calliada, F., Cantisani, V., Havre, R. F., Jenderka, K. V., ... & Jenssen, C. (2022). Professional Standards in Medical Ultrasound–EFSUMB Position Paper (Long Version)–General Aspects. Ultraschall in der Medizin-European Journal of Ultrasound, 43(05), e36-e48.
   View at Publisher | View at Google Scholar
9. Keller, K., Hobohm, L., Ebner, M., Kresoja, K. P., Münzel, T., Konstantinides, S. V., & Lankeit, M. (2020). Trends in thrombolytic treatment and outcomes of acute pulmonary embolism in Germany. European heart journal, 41(4), 522-529.
   View at Publisher | View at Google Scholar
10. Ebner, M., Sentler, C., Harjola, V. P., Bueno, H., Lerchbaumer, M. H., Hasenfuß, G., ... & Lankeit, M. (2021). Outcome of patients with different clinical presentations of high-risk pulmonary embolism. European Heart Journal Acute Cardiovascular Care, 10(7), 787-796.
    View at Publisher | View at Google Scholar
11. THeroux, C. D., ALioTTA, J. M., & MuLLin, C. J. (2019). High-risk pulmonary embolism: current evidence-based practices. RI Med J, 102(10), 43-47.
    View at Publisher | View at Google Scholar
12. Surov, A., Akritidou, M., Bach, A. G., Bailis, N., Lerche, M., Meyer, H. J., ... & Wienke, A. (2021). A new index for the prediction of 30-day mortality in patients with pulmonary embolism: the pulmonary embolism mortality score (PEMS). Angiology, 72(8), 787-793.
    View at Publisher | View at Google Scholar
13. Becattini, C., & Agnelli, G. (2016). Risk stratification and management of acute pulmonary embolism. Hematology 2014, the American Society of Hematology Education Program Book, 2016(1), 404-412.
    View at Publisher | View at Google Scholar
14. Jiménez, D., Bikdeli, B., Rodríguez, C., Muriel, A., Ballaz, A., Soler, S., ... & RIETE Investigators. (2023). Identification of low-risk patients with acute symptomatic pulmonary embolism. Archivos de Bronconeumologia, 59(9), 575-580.
    View at Publisher | View at Google Scholar
15. Nordstrom, S. M., & Weiss, E. J. (2008). Sex differences in thrombosis. Expert Review of Hematology, 1(1), 3-8.
    View at Publisher | View at Google Scholar
16. Bajko, Z., Motataianu, A., Stoian, A., Barcutean, L., Andone, S., Maier, S., ... & Balasa, R. (2021). Gender differences in risk factor profile and clinical characteristics in 89 consecutive cases of cerebral venous thrombosis. Journal of Clinical Medicine, 10(7), 1382.
    View at Publisher | View at Google Scholar