Journal of Atherosclerosis and Thrombosis
Online ISSN : 1880-3873
Print ISSN : 1340-3478
ISSN-L : 1340-3478
Original Article
Validation of the Pulmonary Embolism Severity Index Risk Classification and the 2019 European Society of Cardiology Risk Stratification in the Southeast Asian Population with Acute Pulmonary Embolism
Chaiwat Bumroongkit, Atikun Limsukon, Chalerm Liwsrisakun, Athavudh Deesomchok, Chaicharn Pothirat, Theerakorn Theerakittikul, Konlawij Trongtrakul, Pattraporn Tajarernmuang, Nutchanok Niyatiwatchanchai, Juthamas Inchai, Warawut Chaiwong
著者情報
キーワード: Acute pulmonary embolism, PESI score, PESI classification, sPESI, 2019 ESC risk stratification
ジャーナル オープンアクセス HTML

2023 年 30 巻 11 号 p. 1601-1611

Browse “Advance online publication” version
詳細
PDFをダウンロード (408K)
メタデータをダウンロード RIS形式

(EndNote、Reference Manager、ProCite、RefWorksとの互換性あり)

BIB TEX形式

(BibDesk、LaTeXとの互換性あり)

テキスト
メタデータのダウンロード方法
発行機関連絡先
Abstract

Aim: To date, no studies representing the Southeast Asian population have validated the Pulmonary Embolism Severity Index (PESI) and 2019 European Society of Cardiology (ESC) risk stratification. Therefore, this study aimed to validate the PESI score, simplified PESI (sPESI), PESI risk classification, and 2019 ESC risk stratification in Southeast Asian patients with acute pulmonary embolism (APE).

Methods: The present study is a 10-year cross-sectional study. Here, risk regressions were conducted to identify the PESI risk classification, sPESI, and 2019 ESC risk stratification as predictors for 30-day all-cause and PE-related mortalities. Receiver operating characteristic (ROC) curves were constructed to determine the diagnostic ability of the PESI score, sPESI score, PESI risk classification, and 2019 ESC risk stratification to predict 30-day mortality.

Results: A total of 696 patients (male, 286; female, 410; mean age, 57.7±15.7 years) were included in this study from 2011 to 2020. The risk of 30-day all-cause mortality progressively increased with the 2019 ESC risk stratification, being approximately 6-fold higher in the high-risk than in the low-risk class [risk ratio: 6.24 (95% confidence interval (CI), 3.12, 12.47), P＜0.001]. The risk of 30-day all-cause mortality with the PESI risk classification also increased with the risk classes, being approximately 6-fold higher in class V than in class I [adjusted risk ratio: 5.91 (95% CI, 2.25, 15.51), P＜0.001]. The highest area under the receiver operating characteristic curve (AuROC) of the predictive model was the PESI score [AuROC=0.733 (95% CI, 0.685, 0.782)].

Conclusion: Our study represents a good validation of the PESI and 2019 ESC risk stratification to predict 30-day mortality after APE diagnosis in the Southeast Asian population.

Introduction

Acute pulmonary embolism (APE) has a broad spectrum of clinical presentation, ranging from asymptomatic (an incidental finding on computed tomography [CT]) to severe hemodynamic instability and sudden death1). APE patients presenting with hemodynamic instability have an average short-term mortality of 14%–50%1, 2). The prevalence of APE in Asian countries is not as high as in Western countries. The recently published population-wide estimates of annual symptomatic venous thromboembolism rates in Korea, Taiwan, and Hong Kong were 13.8, 15.9, and 19.9 per 100,000 people, respectively. These are approximately 15%–20% of the level recorded in Western countries (~100 per 100,000 people)3, 4).

Various prediction rules based on clinical parameters have been shown to be helpful in the prognostic assessment of APE patients. Of those, the pulmonary embolism severity index (PESI) is the most extensively validated score to date5). The PESI has been used to help stratify the risk of unfavorable outcomes over the clinical course of APE6). The PESI is aimed at stratifying the risk of death after APE identification on the basis of objective clinical parameters to help guide treatment and identify low-risk patients who may be candidates for management as outpatient. The simplified pulmonary embolism severity index (sPESI ) score has been reported to be useful in predicting 30-day mortality for APE patients, which helps in the identification of low-risk patients for early hospital discharge7, 8). The PESI has been validated in population-based samples from different countries9-11). In 2019, the European Society of Cardiology (2019 ESC) guidelines for the diagnosis and treatment of APE included this tool in their treatment guidance flowchart12). The ESC 2019 guidelines proposed a prognostic model for early mortality (i.e., within 30 days) after APE diagnosis based on integrated clinical, laboratory, and instrumental parameters, which stratify into four categories of early mortality risk as high, intermediate-high, intermediate-low, and low risks12). Risk category is defined based on hemodynamic instability, PESI class, evidence of right ventricle dysfunction, and elevated cardiac biomarker levels, that is, troponins or pro-B-type natriuretic peptide (pro-BNP)12). This model allows the differentiation of patients who can undergo home treatment (low-risk category), reducing hospital costs, from those who need admission into intensive care units from the early stages of the acute event (high-risk category). Previous studies were focused on the predictive ability of the PESI score, sPESI, and 2019 ESC risk class stratification8, 13-15). For example, Soriano et al. suggested that the original version of the PESI showed a greater area under the receiver operating characteristic curve (AuROC) than the simplified version (0.70 vs. 0.60)14). Moor et al. demonstrated that the discriminatory power for overall mortality was lower in the 2019 ESC algorithm (AUC: 0.64) compared with the PESI (AUC: 0.75) but did not reach statistical significance13). However, to the best of our knowledge, there were no studies validating the PESI and 2019 ESC risk class stratification in Thailand, which might represent the Southeast Asian population. Therefore, this study aimed to validate the PESI score, sPESI, PESI risk classification, and 2019 ESC risk stratification in patients with a diagnosis of APE in tertiary care hospital in northern Thailand which represent the Southeast Asian population.

Methods

Study Design and Setting

This study was a 10-year cross-sectional study that involved retrospective data collection. It was conducted between January 2011 to December 2020 at Chiang Mai University Hospital, Chiang Mai, Thailand, a tertiary care university hospital.

Subject Selection

All patients aged 15 years or above who were admitted with ICD-10 coding I 26.0 (pulmonary embolism with mention of acute cor pulmonale) and I 26.9 (pulmonary embolism without mention of acute cor pulmonale) were recruited into the study. Medical records were reviewed to confirm the diagnosis of APE. APE diagnosis was made when one of the following criteria (confirmed by a radiologist) was present: thrombus in the pulmonary artery and its branches detected via computed tomographic angiography (CTPA) or multi-detector CT of the chest with contrast. Right ventricular dysfunction was diagnosed when the right ventricle (RV) and left ventricle (LV) diameter ratios were ＞1.

The patients with APE who had evidence of hemodynamic instability (defined as systolic blood pressure [SBP] of ＜90 mmHg for ＞15 min, drop in SBP of 40 mmHg or more from baseline, or requiring hemodynamic support) or cardiac arrest within 24 h of the APE diagnosis were defined as high-risk PE. Intermediate-high-risk, intermediate-low-risk, and low-risk PE were classified along with the 2019 ESC guideline in the absence of hemodynamic instability.

Variables, Data sources/Measurement

Patient’s data, including demographics, symptoms, signs (observed within the first 24 h of diagnosis), risk factors, co-morbidities, investigations (including electrocardiography, chest radiography, arterial blood gas, Troponin-T, pro-BNP, and echocardiography), treatments and outcomes, PESI score, and simplified PESI score (sPESI) were recorded and reviewed. The 2019 ESC risk stratifications were also evaluated, and any subjects with missing data required to calculate either the PESI or sPESI score were excluded. The variables in the PESI, sPESI, and 2019 ESC risk stratification are presented in Supplemental Tables 1 and 2. PE-related deaths were recorded when death was caused by PE. The study protocol was approved by the ethics committee of Chiang Mai University Hospital [Institutional Review Board approval number: MED-2564-08294].

Supplemental Table 1. PESI, sPESI and PESI, sPESI classification6, 7)

Parameter Original PESI Simplified version (sPESI)
Age Age in years 1 point (if age ＞ 80 years)
Male sex +10 -
Cancer +30 1 point
Chronic heart failure +10 1 point
Chronic pulmonary disease +10
Heart rate ≥ 110 beats/min +20 1 point
Systolic blood pressure ＜100 mm Hg +30 1 point
Respiratory rate ＞30 breaths/min +20 -
Temperature ＜36℃ +20 -
Altered mental status +60 -
Arterial oxygen saturation ＜90% +20 1 point
Risk strata
Class I: ≤ 65 points 0 point = Low risk
Class II: 66-85 points ≥ 1 points = High risk
Class III: 86-105 points
Class IV: 106-125 points
Class V: ＞ 125 points

Abbreviations: PESI, Pulmonary Embolism Severity Index; sPESI, simplified pulmonary embolism severity index

Supplemental Table 2. 2019 ESC risk stratification of pulmonary embolism severity and the risk of early (in-hospital or 30 day) death12)

Early mortality risk Indicators of risk
Hemodynamic instability PESI class III-V or sPESI ≥ 1 RV dysfunction on TTE or CTPA Elevated cardiac troponin level
High + (+) + (+)
Intermediate Intermediate-high - + + +
Intermediate-low - + One (or none) positive
Low - - -

Assessment optional;

if assessed, negative

Abbreviations: PESI, Pulmonary Embolism Severity Index; RV, right ventricle; TTE, transthoracic echocardiography; CTPA, computed tomographic pulmonary angiography.

Statistical Analysis

Categorical data are expressed as numbers and percentages and continuous data as mean and standard deviation or median and interquartile range based on their distributions. Fisher’s exact test was employed to compare the categorical data between the survivor and non-survivor groups. Risk regressions were conducted to identify the PESI risk classification, sPESI, and 2019 ESC risk stratification as a predictor for 30-day all-cause mortality and PE-related death. The results were expressed as risk ratio (RR) together with a 95% confidence interval (CI) for RR. To determine the diagnostic ability of the PESI score, sPESI score, PESI risk classification, and 2019 ESC risk stratification to predict 30-day mortality, receiver operating characteristic (ROC) curves were constructed by plotting the true-positive rate (sensitivity) against the false-positive rate (1-specificity) for each scale level of the predictor variables for two dichotomous outcomes (non-survivor and survivor). The ROCs across the PESI score, sPESI score, PESI risk classification, and 2019 ESC risk stratification were compared. Statistical significance was accepted at P＜0.05. All statistical analyses were conducted using STATA version 16 (StataCorp, College Station, TX, USA).

Results

A total of 1,560 patient records were obtained from the ICD-10 coding diagnosis from January 2011 to December 2020, of which 864 were excluded due to age below 15 years, non-Asian ethnicity, repeated cases and lack of CT confirming the APE diagnosis, wrong or ambiguous diagnosis, incorrect coding, diagnosis of chronic thromboembolic pulmonary hypertension, and missing data required to calculate either the PESI or sPESI score (Fig.1). A total of 696 patients (male, 41.1%) were included in the study. Their age varied from 15 to 98 years old, and the mean age was 57.7±15.7 years. The baseline characteristics of the patients with APE divided according to the survival status at 30 days from the acute event are presented in Table 1. Non-survivors had higher PESI (132.3±36.0 vs. 103.9±31.6) and sPESI (2.5±1.0 vs. 1.7±1.0) scores. More data are presented in Table 1.

Fig.1. Study flow diagram

IPD, inpatient department; APE, acute pulmonary embolism; CT, computed tomography; DVT, deep vein thrombosis; CTEPH, chronic thromboembolic pulmonary hypertension

Table 1. Baseline characteristics of the 696 patients with Acute PE divided according to the survival status at 30 days from the acute event

Characteristics Non-survivors (n = 133) Survivors (n = 563) p-value
Age (years) 59.2±14.3 57.4±15.9 0.228
Male sex, n (%) 67 (50.4) 219 (38.9) 0.019
Female sex, n (%) 66 (49.6) 344 (61.1)
Underlying diseases
Heart failure 8 (6.0) 34 (6.0) 1.000
Hypertension 54 (40.6) 251 (44.6) 0.438
DM 23 (17.3) 91 (16.2) 0.419
Obesity 2 (1.5) 10 (1.8) 1.000
OSA 0 (0.0) 15 (2.7) 0.088
COPD 11 (8.3) 27 (4.8) 0.135
Clinical APE
Suspected PE 101 (75.9) 356 (63.2) 0.006
Incidental PE 29 (21.8) 199 (35.4) 0.003
Risk factors
Active malignancy 100 (75.2) 283 (50.4) ＜0.001
Surgery (＜4 week) 21 (15.8) 120 (21.4) 0.187
Immobilization 50 (37.6) 161 (28.6) 0.047
Long travel 1 (0.8) 9 (1.6) 0.696
Unprovoked PE 13 (9.8) 123 (21.8) 0.004
Current DVT (Acute DVT proven by imaging) 51 (38.3) 194 (34.5) 0.674
Previous DVT, PE 7 (5.3) 38 (6.8) 0.100
Symptoms and signs
Dyspnea 109 (82.0) 344 (61.3) ＜0.001
Chest pain 14 (10.5) 98 (17.5) 0.018
Syncope 14 (10.5) 40 (7.1) 0.207
Tachycardia (Heart rate ≥ 110 ) 89 (66.9) 237 (42.2) ＜0.001
Tachypnea (Respiratory rate ≥ 30 ) 51 (38.3) 107 (19.0) ＜0.001
Systolic BP (mmHg) 110.4±22.2 117.8±21.7 0.001
Diastolic BP (mmHg) 68.9±14.3 73.1±13.7 0.002
Shock (BP＜90/60 mmHg ) 30 (22.6) 47 (8.3) ＜0.001
SpO2 ＜90% 89 (67.4) 244 (43.8) ＜0.001
Respiratory failure (need ETT & MV support) 48 (36.1) 81 (14.4) ＜0.001
Investigations
Sinus tachycardia
S1Q3T3 35 (26.3) 125 (22.2) 0.470
Right ventricular strain pattern EKG 21 (15.8) 91 (16.2) 0.728
RV dysfunction (n = 332) 26 (42.6) 106 (39.1) 0.665
RV/LV ＞1 43 (32.3) 147 (26.1) 0.160
Mediastinal (central) level emboli 51 (38.3) 259 (46.0) 0.121
Troponin-T (median, IQR) (N = 228) 122.0 (34.0, 233.3) 40.2 (16.0, 114.5) 0.002
D-dimer (median, IQR) (N = 129) 12,820.0 (7,058.0, 32,909.0) 7,316.0 (3,898.0, 14,389.0) 0.024
Clinical scores
PESI score 132.3±36.0 103.9±31.6 ＜0.001
Simplified PESI score 2.5±1.0 1.7±1.0 ＜0.001

Note: Data are presented as mean±SD or n (%) and median, IQR for D-dimer, Troponin-T

Abbreviations: DM, diabetes mellitus; OSA, obstructive sleep apnea; COPD, chronic obstructive pulmonary disease; PE, pulmonary emboli; DVT, deep vein thrombosis, BP, blood pressure; RV, right ventricle; LV, left ventricle; SpO2, pulse oximetry; PESI, Pulmonary Embolism Severity Index

The number of patients according to the PESI classification and 2019 ESC risk stratification are presented in Table 2. The lowest proportion was PESI class I (8.9%), and the highest proportion was PESI class V (27.9%). The 2019 ESC risk stratification was as follows: low-risk, 146 (21.0%); intermediate-low-risk, 376 (54.0%); intermediate-high-risk, 96 (13.8%); and high-risk, 78 (11.2%). More data are presented in Table 2.

Table 2. Number of patients according to PESI score, PESI classification and 2019 ESC risk stratification

PESI classification (n = 696) N (%)
I 62 (8.9)
II 98 (14.1)
III 191 (27.4)
IV 151 (21.7)
V 194 (27.9)
2019 ESC risk stratification (n = 696)
Low-risk 146 (21.0)
Intermediate-low 376 (54.0)
Intermediate-high 96 (13.8)
High-risk 78 (11.2)

Abbreviations: PESI, Pulmonary Embolism Severity Index; ESC, European Society of Cardiology

The PESI classification according to high-risk of the 2019 ESC stratification is presented in Table 3. Most of the high-risk cases of the 2019 ESC risk stratification were in PESI class V. There were no high-risk cases of the 2019 ESC risk stratification in PESI classes I and II. More data are presented in Table 3.

Table 3. PESI classification according to high-risk of 2019 ESC risk stratification

PESI classification High-risk (n = 78) p-value
I 0 (0.0)
II 0 (0.0)
III 1 (1.3) ＜0.001
IV 12 (15.4)
V 65 (83.3)

Abbreviations: PESI, Pulmonary Embolism Severity Index; ESC, European Society of Cardiology

The 30-day all-cause and PE-related mortalities are presented in Table 4. The all-cause mortality according to the PESI classification and the 2019 ESC risk stratification was the lowest and highest in the low- and high-risk classes, respectively. The PE-related mortality was also the lowest and highest in the low- and high-risk classes, respectively. More data are presented in Table 4.

Table 4. All-causes mortality and PE-related mortality according to PESI classification and 2019 ESC risk stratification at 1 month

Classification All-causes death at 1 month (n = 133) PE-related death at 1 month (n = 39) p-value
PESI classification (n = 696) ＜0.001
I (n = 62) 4 (6.5) 0 (0.0)
II (n = 98) 6 (6.1) 1 (1.0)
III (n = 191) 22 (11.5) 2 (1.0)
IV (n = 151) 27 (17.9) 3 (2.0)
V (n = 194) 74 (38.1) 33 (17.0)
2019 ESC risk stratification (n = 696) ＜0.001
Low-risk (n = 146) 9 (6.2) 1 (0.7)
Intermediate-low (n = 376) 72 (19.1) 3 (0.8)
Intermediate-high (n = 96) 22 (22.9) 12 (12.5)
High-risk (n = 78) 30 (38.5) 23 (29.5)

Note: Data are n (%)

Abbreviations: PESI, Pulmonary Embolism Severity Index; ESC, European Society of Cardiology

The risk of 30-day all-cause mortality progressively increased with the 2019 ESC risk stratification, being approximately 6-fold higher in the high-risk than in the low-risk class [adjusted risk ratio (ARR): 6.24 (95% CI; 3.12, 12.47), P＜0.001]. The risk of 30-day all-cause mortality with the PESI risk class also increased with the risk class, being approximately 6-fold higher in class V than in class I ((ARR: 5.91 (95% CI; 2.25, 15.51), P＜0.001), as presented in Table 5.

Table 5. Risk of all-causes mortality at 30-day after acute PE according to the 2019 ESC risk stratification, PESI classification, and sPESI classification

Classification Risk ratio 95%CI p-value
2019 ESC risk stratification
Low-risk 1 (reference)
Intermediate-low 3.11 1.59, 6.05 0.001
Intermediate-high 3.72 1.79, 7.73 ＜0.001
High-risk 6.24 3.12, 12.47 ＜0.001
PESI classification
I 1 (reference)
II 0.95 0.28, 3.23 0.933
III 1.79 0.64, 4.98 0.268
IV 2.77 1.01, 7.59 0.047
V 5.91 2.25, 15.51 ＜0.001
sPESI
sPESI score 1.66 1.50, 1.83 ＜0.001
sPESI classification
0 1 (reference)
≥ 1 4.46 1.46, 13.62 0.009

Notes: Data obtained from the 696 patients with acute PE (all the 133 Non-survivors and 563 of the survivors)

Abbreviations: PESI, Pulmonary Embolism Severity Index; sPESI, simplified Pulmonary Embolism Severity Index; ESC, European Society of Cardiology

The 1-month PE-related mortality also progressively increased with the 2019 ESC risk stratification and PESI risk classification, being 43-fold higher in the high-risk than in the low-risk class [ARR=43.05 (95% CI 5.93, 312.79), P＜0.001] being 27-fold higher in class V than in classes I and II [ARR=27.22 (95% CI; 3.76, 196.81), P=0.001] as presented in Table 6.

Table 6. Risk of 30-day PE-related death (n = 39) after acute PE according to the 2019 ESC risk stratification, PESI classification, and sPESI

Classification Risk ratio 95%CI p-value
2019 ESC risk stratification
Low-risk 1 (reference)
Intermediate-low 1.55 0.18, 13.78 0.693
Intermediate-high 19.77 2.63, 148.68 0.004
High-risk 43.05 5.93, 312.79 ＜0.001
PESI classification
I-II 1 (reference)
III 2.51 0.26, 23.92 0.423
IV 4.24 0.48, 37.49 0.194
V 27.22 3.76, 196.81 0.001
sPESI
sPESI score 2.19 1.74, 2.75 ＜0.001
sPESI classification
0 1 (reference)
≥ 1 4.12 0.58, 29.48 0.158

Notes: Data obtained from the 696 patients with acute PE (PE related dead 39 and 657 of the survivors)

Abbreviations: PESI, Pulmonary Embolism Severity Score score; sPESI, simplified Pulmonary Embolism Severity Index; ESC, European Society of Cardiology

To estimate the predictive capability of the model, an ROC analysis was conducted, as presented in Fig.2, and comparison of the AuROC is shown in Table 7. The highest AuROC of the predictive model was the PESI score [AuROC=0.733 (95% CI; 0.685, 0.782)], and the AuROC of the sPESI and PESI classification was greater than 0.7, indicating a good discriminative power, except for the ESC risk stratification [AuROC=0.643 (95% CI; 0.596, 0.688)]. The AuROC of the ESC 2019 risk stratification was significantly lower than the others. More data are presented in Table 7.

Fig.2. Receiver operating characteristic curves for 30-day mortality for the PESI score, PESI risk classification, simplified PESI score, and 2019 ESC risk stratification

PESI, Pulmonary Embolism Severity Index; sPESI, simplified Pulmonary Embolism Severity Index; ESC, European Society of Cardiology

Table 7. Discriminatory power of the PESI score, sPESI score, PESI risk classification and 2019 ESC risk stratification

Comparison AuROC (95%CI) p-value
PESI score (0.733; 0.685, 0.782) sPESI score (0.713; 0.666, 0.759) 0.173
PESI score (0.733; 0.685, 0.782) PESI risk classification (0.713; 0.667, 0.759) 0.010
PESI score (0.733, 95%CI; 0.685, 0.782) 2019 ESC risk stratification (0.643; 0.596, 0.688) ＜0.001
sPESI score (0.713; 0.666, 0.759) PESI risk classification (0.713; 0.667, 0.759) 0.963
sPESI score (0.713; 0.666, 0.759) 2019 ESC risk stratification (0.643; 0.596, 0.688) 0.002
PESI risk classification (0.713; 0.667, 0.759) 2019 ESC risk stratification (0.643; 0.596, 0.688) ＜0.001

Abbreviations: PESI, Pulmonary Embolism Severity Score score; sPESI, simplified Pulmonary Embolism Severity Index; ESC, European Society of Cardiology

Discussion

The prediction of 30-day mortality after an APE diagnosis proposed by the 2019 ESC risk stratification, original PESI, sPESI, and PESI risk classification was validated in our study. We found that the all-cause mortality rate in 1 month was 19.1%, which is comparable to those in previous studies in Thailand (10.0%–21.3%)16-18) but is higher than those in previous studies in Europe and North America (9%–17%)19-24) as well as Korea (7.8%)25). The causes of death were mostly pneumonia, sepsis, and advanced cancers. The PE-related mortality was only 5.9%, and most of the PE-related deaths were massive PE cases (56%) and intermediate-high-risk PE cases (31.7%).

The PESI, sPESI, PESI risk classification, and 2019 ESC risk stratification were validated for the prediction of 30-day all-cause mortality in the Asian population with APE. We found that the 30-day risk of death proportionally increased with the 2019 ESC stratification. Furthermore, we demonstrated that the 30-day all-cause mortality increased by 3.1, 3.7, and 6.2 times in the intermediate-low-risk, intermediate-high-risk, and high-risk categories, respectively, as compared with the low-risk category. We also found that the 30-day risk of death proportionally increased with the PESI severity classification. For example, the 30-day all-cause mortality increased by 5.9 times in the PESI class V compared with the PESI class I. For the 30-day PE-related mortality, we found that it proportionally increased with the 2019 ESC stratification and PESI classification. The 1-month PE-related mortality progressively increased with the 2019 ESC risk stratification and PESI risk classification, being 43-fold higher in the high-risk than in the low-risk class and being 27-fold higher in class V than in classes I and II. Our results were supported by previous findings indicating that the mortality proportionally increased with the higher level of severity according to the PESI classification, sPESI, and 2019 ESC risk stratification8, 13, 14).

According to the AuROC, the predictive model including the PESI, sPESI, and PESI risk class exhibited a good discriminatory power, except for the 2019 ESC risk stratification. The prognostic accuracy from the AuROC was the highest in the original PESI scores (AuROC=0.733), whereas the simplified PESI scores (AuROC=0.713) and PESI risk classification (AuROC=0.713) did not differ but were higher than the 2019 ESC risk stratification (AuROC=0.643). Our results were comparable with the previous studies suggesting that the AuROC of the original PESI and simplified PESI for mortality prediction ranged from 0.70 to 0.78 and from 0.60 to 0.72, respectively13-15). Previous study in Brazil demonstrated that the AuROC of sPESI for mortality prediction was 0.60, which was lower than that in our study14). The AuROC of the 2019 ECS risk stratification for mortality prediction was 0.63 in the previous study, which is consistent with our results13).

Our results indicated that the original PESI, sPESI, PESI risk classification, and 2019 ESC risk stratification could predict 30-day all-cause and PE-related mortalities. The original PESI score had the highest AuROC for mortality prediction. Therefore, we preferred to recommend the use of the PESI score for classifying the 30-day all-cause and 30-day PE-related mortalities. However, the AuROC of the ECS 2019 risk stratification for mortality prediction was lower than 0.7. Thus, the PESI score should be used as the primary tool for the identification of low-risk patients who may be candidates for outpatient management.

The mortality data according to the PESI and sPESI risk classifications between the Western population and our study are presented in Table 8. Both the all-cause and PE-related mortalities according to the PESI and sPESI risk classification in our study were comparable with those in the study of the Western population6, 7). More data are presented in Table 8.

Table 8. The mortality data according to PESI and sPESI risk classification between the Western people and our study

PESI (Aujesky et al)6) sPESI (Jiminez et al)7) PESI (Bumroongkit et al) sPESI (Bumroongkit et al)

Class I: ≤ 65 points

very low 30 day mortality risk (0.0-1.6%)

0 points = 30 day mortality risk

1.0% (95%CI; 0.0, 2.1)

Class I: ≤ 65 points

PE-related death at 1 month = 0%

0 points

PE-related death at 1

month = 1.5% (95%CI; 0.0,

8.3)

Class II: 66-85 points

low mortality risk (1.7-3.5%)

Class II: 66-85 points

PE-related death at 1 month = 1.0%

(95%CI; 0.0, 5.5)

Class III:

86-105 points

moderate mortality risk (3.2- 7.1%)

≥ 1 point(s) = 30 day mortality

risk 10.9% (95%CI; 8.5, 13.2)

Class III: 86-105 points

PE-related death at 1 month = 1.0%

(95%CI; 0.0, 3.7)

≥1 point(s)

PE-related death at 1 month = 6.3%

(95%CI; 4.6, 8.5)

Class IV:

106-125 points

high mortality risk (4.0-11.4%)

Class IV: 106-125 points

PE-related death at 1 month = 2.0%

(95%CI; 0.0, 5.7)

Class V: ＞125 points

very high mortality risk (10.0- 24.5%)

Class V: ＞125 points

PE-related death at 1 month = 17.0%

(95%CI; 12.0, 23.0)

Abbreviations: PESI, Pulmonary Embolism Severity Score score; sPESI, simplified Pulmonary Embolism Severity Index

The strength of our study is that we enrolled only Southeast Asian patients. Therefore, the heterogeneity of the study population might be minimal. However, our study involved retrospective data collection, which limited the possibility to retrieve all patient information regarding some parameters for the correct definition of the ESC risk class for early mortality, such as biomarkers (for example, some cases in intermediate-risk patients would be classified as intermediate-high- or intermediate-low-risk patients).

Conclusions

Our study represents a good validation of the predictive models for 30-day mortality after APE diagnosis proposed by the sPESI, PESI score, PESI risk classification, and 2019 ESC risk stratification in the Southeast Asian population. However, the prognostic accuracy from the AuROC for the 30-day mortality prediction was higher in the PESI risk classification than in the 2019 ESC risk class stratification. The original PESI score has the highest accuracy.

Acknowledgements

The authors wish to acknowledge the contribution of the staff of the Division of Pulmonary, Critical Care and Allergy, Department of Internal Medicine, Faculty of Medicine, Chiang Mai University to this trial.

Conflict of Interest

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Author Contributions

Study conception and design: (C.B.), data acquisition: (C.B., W.C.), data analysis: (C.B., W.C.), manuscript preparation and revision: (C.B., A.L., W.C.). The remaining authors conducted acquisition and interpretation of data and critical reviews of the manuscript. All authors read and approved the final manuscript.

Funding

This research did not receive any grant from funding agencies in the public, commercial, or not-for-profit sectors.

References
 

This article is licensed under a Creative Commons [Attribution-NonCommercial-ShareAlike 4.0 International] license.
https://creativecommons.org/licenses/by-nc-sa/4.0/
feedback
 Top