The Global Burden of Disease Results Tool of the Institute for Health Metrics and Evaluation reported that cardiovascular diseases accounted for approximately 10 383 550 deaths globally in 2017, representing 18.56% of all-cause mortality.1 Cardiothoracic surgery plays an important role in the treatment of these conditions and in reducing associated morbidity and mortality. However, surgery carries inherent risks that vary among patients, necessitating careful evaluation of risks and benefits before proceeding. A risk stratification tool is essential for effective patient triage and the consent process.

One widely used risk stratification tool is the European System for Cardiac Operative Risk Evaluation (EuroSCORE), a specialised scoring system that provides customised predictions of in-hospital mortality after cardiac surgery. The tool assigns scores based on various preoperative risk factors to stratify patients into different risk categories (low: EuroSCORE <4%, intermediate: 4-8%, high: >8%).2 In the UK, in-hospital mortality declined from 4.0% to 2.8% between 2002 and 2016 following implementation of EuroSCORE,3 supporting its value in cardiac surgical risk assessment. The EuroSCORE comprises three versions: the additive EuroSCORE,4 the logistic EuroSCORE,5 and EuroSCORE II.6 In 2012, the Society for Cardiothoracic Surgery in Great Britain and Ireland recommended the use of the latest version, EuroSCORE II.6

Prince of Wales Hospital (PWH) in Hong Kong has adopted the logistic EuroSCORE for risk assessment since 2007. However, several publications from different countries have raised concerns regarding the accuracy of the additive and logistic EuroSCORE models, leading to the development of EuroSCORE II.7 8 9 Consequently, EuroSCORE II has been proposed as the future risk adjustment tool of the Society for Cardiothoracic Surgery in Great Britain and Ireland following successful contemporary validation.6 10 11

Although EuroSCORE II has been widely used and validated, the underlying data were predominantly derived from Western populations undergoing cardiac surgery in Europe and the US.4 12 13 14 Therefore, studies evaluating the performance of EuroSCORE II in Asian populations remain limited,15 16 17 and none has been conducted specifically in Hong Kong. Furthermore, no studies have compared the performance of the logistic EuroSCORE and EuroSCORE II in the Hong Kong population. The present study aimed to address these gaps.

Moreover, Hong Kong has a higher proportion of aortic surgery than Western countries. PWH reported a surge in aortic surgeries, reaching 26% between 2021 and 2022,18 whereas the UK reported an aortic surgery prevalence of 3.47% between 2015 and 2016.3 We therefore sought to investigate whether this variation influences the validity of EuroSCORE II through subgroup analyses.

The primary objective of this study was to assess the discriminatory ability and calibration of EuroSCORE II in predicting postoperative mortality after the three main index cardiac surgeries (ie, coronary artery bypass grafting [CABG], valve surgery, and aortic surgery) at our centre.2 7 8 9 14 15 16 17 19 20 21 22 23 The secondary objective was to compare the discriminatory ability and calibration of EuroSCORE II with those of the logistic EuroSCORE in patients undergoing cardiac surgery.

This retrospective validation study included patients (aged ≥18 years) who underwent all types of cardiac surgery—including CABG, valve surgery (eg, aortic valve replacement, mitral valve replacement, and tricuspid valve repair), aortic surgery, isolated or combined procedures, and other procedures (eg, left atrial appendage closure) at PWH between 1 January 2013 and 31 December 2023 (inclusive) [Fig]. Because PWH does not perform certain cardiothoracic procedures—such as paediatric cardiac surgery, cardiac transplantation, and oesophageal surgery—records for these interventions were unavailable. For patients who underwent multiple cardiac surgeries during the same hospital admission, only the first index procedure was analysed. The minimum sample size of 225 was calculated based on estimates of area under the receiver operating characteristic curve (AUROC) from the literature7 and the estimated prevalence of the outcome (online supplementary Table),7 24 indicating that our primary cohorts for CABG, valve surgery, and aortic surgery exceeded the required sample size.

The Dendrite cardiac surgery database (Dendrite Clinical Systems, Oxford, United Kingdom)25 was utilised for secondary data collection (Fig). This database captures clinically relevant information, including preoperative medical records and postoperative complications for patients undergoing cardiac surgery. All key variables required to calculate EuroSCORE II6 and the logistic EuroSCORE5 were extracted. Mortality, the primary outcome, was defined as death within 30 days of the index operation (regardless of place of death), consistent with previous studies.2 9 14 15 21

Cases with missing or incomplete data required for calculation of EuroSCORE II were excluded from the analysis. Analyses were performed for the overall cohort and stratified by individual cardiac procedure. Complete data were available for validation of EuroSCORE II and comparison of its predictive performance with that of the logistic EuroSCORE for postoperative mortality.

Univariate and multivariate binary logistic regression analyses were conducted on all relevant variables included in the EuroSCORE II scale to identify significant covariates associated with an increased risk of mortality.

The discriminatory performance of the predictive models was evaluated using the AUROC; values of 0.8 or above indicated strong discrimination, and 1.0 indicated perfect discrimination. Pairwise comparisons of AUROCs for individual cardiac procedures were performed using the DeLong test, with the threshold for statistical significance set at P<0.05.

Calibration of the predictive model was evaluated using the Hosmer–Lemeshow goodness-of-fit test and calibration plots, through statistical and graphical assessment of agreement between observed and expected event rates within model subgroups. A P value >0.05 and a regression line approximating the 45-degree diagonal indicated good calibration, reflecting adequate agreement between observed and predicted event rates.

Model goodness of fit was further assessed using the coefficient of determination (R²), which quantifies the proportion of variance explained by the model, and the normalised root mean square error (NRMSE), which measures predictive accuracy by comparing predicted and observed values, normalised to the data range. Higher R² values and lower NRMSE values indicate better model fit.

Statistical analyses were performed using SPSS (Windows version 29.0; IBM Corp, Armonk [NY], United States), Microsoft Excel 2019, and R software (RStudio, version 2024.04.2).

The study cohort comprised 4180 patients (Fig). Table 1 summarises the characteristics of the overall cohort and relevant subgroups. The median age was 63 years (interquartile range, 56-69), and 29.2% (n=1222) were women. Aortic operations were performed in 21.1% (n=883) of patients and the majority underwent a single non-coronary procedure (36.9%, n=1541). For the overall cohort, the median logistic EuroSCORE value was 5.8 (interquartile range, 2.6-13.7), whereas median EuroSCORE II value was 2.4 (interquartile range, 1.2-5.3). The institutional 30-day mortality rate for all cardiac procedures was 4.2%.

The AUROC for EuroSCORE II was 0.829, indicating strong discriminatory ability. The Hosmer–Lemeshow P value for EuroSCORE II was 0.155, indicating no statistically significant difference between predicted and observed values (online supplementary Fig a). Accordingly, EuroSCORE II demonstrated acceptable calibration.

EuroSCORE II demonstrated a statistically significant improvement in discriminatory performance compared with the logistic EuroSCORE (DeLong P=0.006). Additionally, EuroSCORE II showed superior calibration, supported by a significant Hosmer–Lemeshow test result for the logistic EuroSCORE (P<0.001). Calibration curves comparing observed and predicted 30-day mortality were consistent with these findings, further indicating better calibration with EuroSCORE II than with the logistic EuroSCORE. More than 90% of the variation in 30-day mortality was explained by both models (R² for EuroSCORE II=98.7%; R² for logistic EuroSCORE=99.1%). Notably, EuroSCORE II demonstrated a substantially lower NRMSE (5.7%) than the logistic EuroSCORE (56.4%), indicating reduced dispersion and relative variability in predictions (online supplementary Fig a).

In this subgroup, EuroSCORE II demonstrated strong discriminatory performance (AUROC=0.847) and acceptable calibration (Hosmer–Lemeshow P=0.113). There was no statistically significant difference in discriminatory performance between EuroSCORE II and the logistic EuroSCORE (DeLong P=0.529). However, EuroSCORE II showed better calibration, supported by a significant Hosmer–Lemeshow test result for the logistic EuroSCORE (P<0.001) and calibration curves favouring EuroSCORE. More than 85% of the variation in 30-day mortality was explained by both models (R2 for EuroSCORE II=87.7%; R² for logistic EuroSCORE=91.4%). Compared with the logistic EuroSCORE (40.6%), EuroSCORE II demonstrated a lower NRMSE (13.0%), indicating reduced dispersion and relative variability (online supplementary Fig b).

In this subgroup, EuroSCORE II demonstrated strong discriminatory performance (AUROC=0.810) and acceptable calibration (Hosmer–Lemeshow P=0.162). There was no statistically significant difference in discriminatory performance between EuroSCORE II and the logistic EuroSCORE (DeLong P=0.160). Nevertheless, EuroSCORE II demonstrated superior calibration, supported by a significant Hosmer–Lemeshow test result for the logistic EuroSCORE (P<0.001) and calibration curves favouring EuroSCORE II. More than 90% of the variation in 30-day mortality was explained by both models (R² for EuroSCORE II=94.7%; R² for logistic EuroSCORE=94.4%). Compared with the logistic EuroSCORE (80.4%), EuroSCORE II demonstrated a lower NRMSE (21.8%), indicating reduced dispersion and relative variability (online supplementary Fig c).

In this subgroup, EuroSCORE II demonstrated satisfactory discriminatory performance (AUROC=0.735) and good calibration (Hosmer–Lemeshow P=0.549). It also showed a statistically significant improvement in discrimination compared with the logistic EuroSCORE (DeLong P<0.001). Calibration was also superior, supported by a significant Hosmer–Lemeshow test result for the logistic EuroSCORE (P<0.001) and calibration curves favouring EuroSCORE II. More than 90% of the variation in 30-day mortality was explained by EuroSCORE II (R² for EuroSCORE II=96.1%; R² for logistic EuroSCORE=76.6%). EuroSCORE II also demonstrated a lower NRMSE (6.6%) than the logistic EuroSCORE (98.8%), indicating reduced dispersion and relative variability (online supplementary Fig d).

In this subgroup, EuroSCORE II demonstrated fair discriminatory performance (AUROC=0.694) and good calibration (Hosmer–Lemeshow P=0.606). There was no statistically significant difference in discriminatory performance between EuroSCORE II and the logistic EuroSCORE (DeLong P=0.913). Both models exhibited adequate calibration (EuroSCORE II P=0.606; logistic EuroSCORE P=0.280) [online supplementary Fig e].

In this subgroup, EuroSCORE II demonstrated strong discriminatory performance (AUROC=0.862) and acceptable calibration (Hosmer–Lemeshow P=0.159). There was no statistically significant difference in discriminatory performance between EuroSCORE II and the logistic EuroSCORE (DeLong P=0.248). However, EuroSCORE II exhibited superior calibration compared with the logistic EuroSCORE (Hosmer–Lemeshow P=0.062) [online supplementary Fig f].

In this subgroup, EuroSCORE II demonstrated strong discriminatory performance (AUROC=0.872) but poor calibration (Hosmer–Lemeshow P<0.001). There was no statistically significant difference in discriminatory performance between EuroSCORE II and the logistic EuroSCORE (DeLong P=0.626). Notably, calibration curves favoured EuroSCORE II over the logistic EuroSCORE (online supplementary Fig g).

Furthermore, comparison of EuroSCORE II variables with multivariable analyses from PWH database identified ‘dialysis’ as an additional significant predictor of increased 30-day mortality (adjusted odds ratio=3.401) among patients undergoing cardiac surgery (Table 2).

In the present study, EuroSCORE II demonstrated strong discriminatory performance and good calibration in the overall cohort and three key subgroups (isolated CABG, isolated valve surgery and aortic surgery). Moreover, EuroSCORE II outperformed the logistic EuroSCORE in both discrimination and calibration across the overall cohort and these principal subgroups.

Our results are consistent with validation studies conducted in several European countries (Italy,26 Greece,27 Serbia,28 Spain,29 and Hungary30), which demonstrated strong discriminatory performance (AUROC >0.7) for EuroSCORE II.26 27 28 29 30 31 These findings reaffirm the robust predictive performance of EuroSCORE II for mortality in patients undergoing cardiac surgery.

In addition to European populations, our findings align with those of validation studies conducted in Asian cohorts.15 16 17 23 Specifically, Liu et al15 demonstrated strong discriminatory performance for EuroSCORE II, with an AUROC of 0.792 in a single-centre setting. This concordance further supports the consistency and reliability of EuroSCORE II as a mortality prediction tool in Asian cardiac surgery populations.

However, Kurniawaty et al19 reported considerably different findings, demonstrating only fair discriminatory performance, with evidence of miscalibration and underprediction in an Indonesian population. This discrepancy may be attributable to differences in patient age. Both our cohort and the European cohorts had substantially higher median (63 years) or mean (64.6 years)6 ages compared with the mean age in the Indonesian cohort (44 years)19. Given the younger age profile and lower prevalence of risk factors included in the EuroSCORE II model among Indonesian patients, its predictive performance may be limited in that population. Accordingly, these findings may be less generalisable to the Hong Kong population.

For the overall cohort, EuroSCORE II demonstrated superior performance in both discrimination and calibration compared with the logistic EuroSCORE. This difference may reflect the tendency of the logistic EuroSCORE to overestimate mortality risk, particularly in high-risk emergency patients.10 Consequently, EuroSCORE II appears to provide more accurate risk stratification than the logistic EuroSCORE.

For isolated CABG procedures, the discriminatory performance of EuroSCORE II was strong in our study, supported by a non-significant Hosmer–Lemeshow statistic, consistent with findings from a large UK validation cohort.7 Studies conducted in Finland32 (AUROC=0.852) and China16 (AUROC=0.762) similarly demonstrated robust discriminatory performance of EuroSCORE II in predicting operative mortality among high-risk isolated CABG patients and those undergoing CABG with or without concomitant major cardiac surgery. However, a study from Singapore reported poor discrimination and calibration, particularly in moderate- and high-risk cohorts.33 Comparable findings were reported in studies from Indonesia22 and Malaysia,23 which demonstrated fair discrimination but underestimation of mortality after isolated CABG. These discrepancies suggest that additional caution may be warranted when applying EuroSCORE II in isolated CABG populations. Differences in demographic characteristics or study design may contribute to variability in model performance, warranting further investigation.

For aortic procedures, EuroSCORE II demonstrated higher AUROC values and more favourable Hosmer–Lemeshow P values than the logistic EuroSCORE. Nevertheless, caution is warranted because the model does not incorporate specific procedural variables (eg, open surgery vs minimally invasive approaches) as risk factors, which may limit precision in mortality prediction for aortic surgery.7

The adoption of contemporary machine learning and artificial intelligence techniques, rather than logistic regression, may offer more effective modelling approaches for capturing complex, non-linear interactions among established risk factors. Furthermore, incorporating the statistically significant variable identified through multivariate analysis of the PWH database, specifically dialysis, into a future EuroSCORE III model may further enhance its predictive performance.

First, the robustness of this validation study is supported by its substantial sample size (n=4180), which increases statistical power, enables detection of smaller effects, and enhances generalisability. Second, the absence of missing data strengthens measurement completeness and the credibility of the validation process, reduces information bias, and facilitates a more precise evaluation of EuroSCORE II predictive performance within this large cohort.

First, reliance on data from a single institution may introduce sampling bias. Therefore, multi-centre analyses should be conducted in future, provided sufficient resources are available. Second, the retrospective design limited the study by precluding long-term follow-up after patient discharge. Consequently, the analysis did not capture longer-term outcomes that may be influenced by baseline EuroSCORE II risk estimates. Additionally, the cohort demonstrated a skewed distribution across risk categories, with a substantial proportion (>85.2%) categorised as low or intermediate risk, thereby limiting generalisability to high-risk populations.

First, in aortic surgery, the discrepancy in EuroSCORE II performance observed between Hong Kong and the UK indicates a need for further investigation.10 A meta-analysis focusing on validation of EuroSCORE II in aortic procedures could help refine risk assessment in this subgroup. Second, although EuroSCORE II is a valuable risk stratification tool in cardiac surgery, minimally invasive cardiac procedures34 and certain established risk factors (eg, diffuse coronary artery disease and aortic calcification)15 are not included in the model. Accordingly, there may be a need for in-depth evaluation of their relevance to EuroSCORE II calculation. Third, multi-centre studies would enable validation of these findings on a broader scale. Collaboration with the other two cardiac centres in Hong Kong would enhance generalisability and support more robust conclusions.

In our cohort, EuroSCORE II demonstrated strong discriminatory performance and good calibration for predicting 30-day postoperative mortality among patients undergoing cardiac surgery. It also shows superior calibration and comparable or improved discrimination in the three principal subgroups—isolated CABG, isolated valve surgery, and aortic surgery—compared with the logistic EuroSCORE. Accordingly, EuroSCORE II represents a risk stratification tool superior to the logistic EuroSCORE and is well suited for use in Hong Kong.

Concept or design: KHL Ng, T Fujikawa, K Wang, RHL Wong.
Acquisition of data: KHL Ng, MWT Kwok, JYK Ho, SCY Chow, JWY Chan, K Lim, ATC Chang, ICH Siu, T Fujikawa, RHL Wong.
Analysis or interpretation of data: KHL Ng, T Fujikawa, K Wang, RHL Wong.
Drafting of the manuscript: KHL Ng, T Fujikawa, K Wang, RHL Wong.
Critical revision of the manuscript for important intellectual content: KHL Ng, T Fujikawa, RHL Wong.

All authors had full access to the data, contributed to the study, approved the final version for publication, and take responsibility for its accuracy and integrity.

All authors have disclosed no conflicts of interest.

The authors thank Dr Simon KS Yau from Department of Family Medicine of the New Territories East Cluster for his insightful contributions to data interpretation and manuscript revision.

This research was presented at The Hospital Authority Convention 2025 (26 May 2025, Hong Kong).

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

This research was approved by the Institutional Review Board of The Chinese University of Hong Kong/Hospital Authority New Territories East Cluster, Hong Kong (Ref No.: 2024.571). The requirement for informed patient consent was waived by the Board due to the retrospective nature of the study.

The supplementary material was provided by the authors and some information may not have been peer reviewed. Accepted supplementary material will be published as submitted by the authors, without any editing or formatting. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by the Hong Kong Academy of Medicine and the Hong Kong Medical Association. The Hong Kong Academy of Medicine and the Hong Kong Medical Association disclaim all liability and responsibility arising from any reliance placed on the content.

1. Vervoort D, Swain JD, Pezzella AT, Kpodonu J. Cardiac surgery in low- and middle-income countries: a state-of-the-art review. Ann Thorac Surg 2021;111:1394-400. Crossref

2. Silverborn M, Nielsen S, Karlsson M. The performance of EuroSCORE II in CABG patients in relation to sex, age, and surgical risk: a nationwide study in 14,118 patients. J Cardiothorac Surg 2023;18:40. Crossref

3. Society for Cardiothoracic Surgery in Great Britain and Ireland. Blue Books. Available from: https://scts.org/professionals/reports/resources/. Accessed 10 Sep 2024.

4. Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R. European System for Cardiac Operative Risk Evaluation (EuroSCORE). Eur J Cardiothorac Surg 1999;16:9-13. Crossref

5. Roques F, Michel P, Goldstone AR, Nashef SA. The logistic EuroSCORE. Eur Heart J 2003;24:881-2. Crossref

6. Nashef SA, Roques F, Sharples LD, et al. EuroSCORE II. Eur J Cardiothorac Surg 2012;41:734-44. Crossref

7. Chalmers J, Pullan M, Fabri B, et al. Validation of EuroSCORE II in a modern cohort of patients undergoing cardiac surgery. Eur J Cardiothorac Surg 2013;43:688-94. Crossref

8. Zheng Z, Li Y, Zhang S, et al. The Chinese coronary artery bypass grafting registry study: how well does the EuroSCORE predict operative risk for Chinese population? Eur J Cardiothorac Surg 2009;35:54-8. Crossref

9. Yap CH, Reid C, Yii M, et al. Validation of the EuroSCORE model in Australia. Eur J Cardiothorac Surg 2006;29:441-6. Crossref

10. Grant SW, Hickey GL, Dimarakis I, et al. Performance of the EuroSCORE models in emergency cardiac surgery. Circ Cardiovasc Qual Outcomes 2013;6:178-85. Crossref

11. Grant SW, Hickey GL, Dimarakis I, et al. How does EuroSCORE II perform in UK cardiac surgery; an analysis of 23 740 patients from the Society for Cardiothoracic Surgery in Great Britain and Ireland National Database. Heart 2012;98:1568-72. Crossref

12. Roques F, Nashef SA, Michel P, et al. Risk factors and outcome in European cardiac surgery: analysis of the EuroSCORE multinational database of 19030 patients. Eur J Cardiothorac Surg 1999;15:816-22. Crossref

13. Gogbashian A, Sedrakyan A, Treasure T. EuroSCORE: a systematic review of international performance. Eur J Cardiothorac Surg 2004;25:695-700. Crossref

14. Nashef SA, Roques F, Hammill BG, et al. Validation of European System for Cardiac Operative Risk Evaluation (EuroSCORE) in North American cardiac surgery. Eur J Cardiothorac Surg 2002;22:101-5. Crossref

15. Liu PH, Shih HH, Kang PL, Pan JY, Wu TH, Wu CJ. Performance of the EuroSCORE II model in predicting short-term mortality of general cardiac surgery: a single-center study in Taiwan. Acta Cardiol Sin 2022;38:495-503. Crossref

16. Shen L, Chen X, Gu J, Xue S. Validation of EuroSCORE II in Chinese patients undergoing coronary artery bypass surgery. Heart Surg Forum 2018;21:E036-9. Crossref

17. Zhang GX, Wang C, Wang L, et al. Validation of EuroSCORE II in Chinese patients undergoing heart valve surgery. Heart Lung Circ 2013;22:606–11. Crossref

18. Prince of Wales Hospital, The Chinese University of Hong Kong. Cardiac Surgery Report 2021–22. Available from: https://www.surgery.cuhk.edu.hk/cts/Cardiac_Surgery_Report_2021-22.pdf. Accessed 4 Sep 2024.

19. Kurniawaty J, Setianto BY, Widyastuti Y, Supomo S, Boom CE, Ancilla C. Validation for EuroSCORE II in the Indonesian cardiac surgical population: a retrospective, multicenter study. Expert Rev Cardiovasc Ther 2022;20:491-6. Crossref

20. Sembiring YE, Ginting A, Puruhito, Budiono. Validation of EuroSCORE II to predict mortality in post-cardiac surgery patients in East Java tertiary hospital. Med J Indones 2021;30:54-9. Crossref

21. Atashi A, Amini S, Tashnizi MA, et al. External validation of European System for Cardiac Operative Risk Evaluation II (EuroSCORE II) for risk prioritization in an Iranian population. Braz J Cardiovasc Surg 2018;33:40-6. Crossref

22. Zahara R, Soeharto DF, Widyantoro B, Sugisman, Herlambang B. Validation of EuroSCORE II scoring system on isolated CABG patient in Indonesia. Egypt Heart J 2023;75:86. Crossref

23. Musa AF, Cheong XP, Dillon J, Nordin RB. Validation of EuroSCORE II in patients undergoing coronary artery bypass grafting (CABG) surgery at the National Heart Institute, Kuala Lumpur: a retrospective review. F1000Res 2018;7:534. Crossref

24. Riley RD, Ensor J, Snell KI, et al. Calculating the sample size required for developing a clinical prediction model. BMJ 2020;368:m441. Crossref

25. Dendrite Clinical Systems. Databases for hospitals and clinics. Available from: https://www.e-dendrite.com/hospital-systems. Accessed 5 Sep 2024.

26. Paparella D, Guida P, Di Eusanio G, et al. Risk stratification for in-hospital mortality after cardiac surgery: external validation of EuroSCORE II in a prospective regional registry. Eur J Cardiothorac Surg 2014;46:840-8. Crossref

27. Stavridis G, Panaretos D, Kadda O, Panagiotakos DB. Validation of the EuroSCORE II in a Greek cardiac surgical population: a prospective study. Open Cardiovasc Med J 2017;11:94-101. Crossref

28. Nezic D, Spasic T, Micovic S, et al. Consecutive observational study to validate EuroSCORE II performances on a single-center, contemporary cardiac surgical cohort. J Cardiothorac Vasc Anesth 2016;30:345-51. Crossref

29. Garcia-Valentin A, Mestres CA, Bernabeu E, et al. Validation and quality measurements for EuroSCORE and EuroSCORE II in the Spanish cardiac surgical population: a prospective, multicentre study. Eur J Cardiothorac Surg 2016;49:399-405. Crossref

30. Koszta G, Sira G, Szatmári K, Farkas E, Szerafin T, Fülesdi B. Performance of EuroSCORE II in Hungary: a single-centre validation study. Heart Lung Circ 2014;23:1041-50. Crossref

31. Barili F, Pacini D, Capo A, et al. Does EuroSCORE II perform better than its original versions? A multicentre validation study. Eur Heart J 2013;34:22-9. Crossref

32. Biancari F, Vasques F, Mikkola R, Martin M, Lahtinen J, Heikkinen J. Validation of EuroSCORE II in patients undergoing coronary artery bypass surgery. Ann Thorac Surg 2012;93:1930-5. Crossref

33. Luo HD, Teoh LK, Gaudino MF, Fremes S, Kofidis T. The Asian system for cardiac operative risk evaluation for predicting mortality after isolated coronary artery bypass graft surgery (ASCORE-C). J Card Surg 2020;35:2574-82. Crossref

34. Ilcheva L, Risteski P, Tudorache I, et al. Beyond conventional operations: embracing the era of contemporary minimally invasive cardiac surgery. J Clin Med 2023;12:7210. Crossref