Usefulness of The CHADS2 and CHA2DS2-VASc Scores in Predicting In-Hospital Mortality in Acute Coronary Syndrome Patients: A Single-Center Retrospective Cohort Study

Background: Although the GRACE risk score is widely accepted as an established scoring system to predict in-hospital mortality in acute coronary syndrome (ACS) patients, this scoring system still depends on electrocardiography and laboratory findings to determine the results. Therefore, we aim to retrospectively evaluate the association between the CHADS2 and CHA2DS2-VASc score as an anamnesis-only mediated scoring system and in-hospital mortality in hospitalized ACS patients. Methods: This retrospective cohort study analyzed data of ACS patients from the ACS registry in Dr. Hasan Sadikin Central General Hospital from 2018 to 2021. The outcome of this study was in-hospital mortality. The association between this scoring system and in-hospital mortality were evaluated using binary logistic regression analysis. Receiver operating characteristics (ROC) analysis was also performed to assess the success rate of this scoring system in predicting in-hospital mortality. Results: A total of 1339 patients were included in this study, and 162 (12.1%) of them died in the hospital. High CHA2DS2-VASc score group (cut-off >2) was significantly associated with higher risk of in-hospital mortality before (OR=2.56 [1.75,3.75]; p<0.001) and after adjustment of several confounding factors (OR=3.39 [1.73,6.64]; p<0.001). Meanwhile, the high CHADS2 score (cutoff >2) only significantly increased the risk of in-hospital mortality in univariate analysis (OR=2.05[1.47,2.87];p<0.001), but was not significantly associated with in-hospital mortality after multivariate analysis (OR=1.31 [0.92,1.86];p=0.129). ROC analysis revealed that the predictive accuracy of the CHA2DS2-VASc score was significantly greater compared to the CHADS2 score (AUC: 0.653 vs 0.609, p<0.001). However, the predictive value of the CHA2DS2-VASc score was significantly lower than the GRACE risk score (AUC: 0.789 vs 0.653, p<0.001).. Conclusion: Our study showed that the CHA2DS2-VASc score >2 was significantly and independently associated with higher in-hospital mortality in ACS patients compared to the CHA2DS2-VASc score of 1 or lower. Despite its lower predictive accuracy compared to the GRACE risk score, the CHA2DS2-VASc score can still be used in practical situations as an alternative scoring system for predicting in-hospital mortality in ACS patients, especially in primary healthcare settings located in rural areas that lack the diagnostic facilities.


Introduction
A cute coronary syndrome (ACS), as a critical and severe manifestation of coronary artery disease, has high in-hospital mortality, numbering up to 20%. 1,2The Global Registry of an Acute Coronary Event (GRACE) risk score is widely known as a standard and validated scoring system to predict all-cause mortality in ACS patients within six months after admission. 3This scoring system can identify high-risk patients who will benefit from aggressive therapies. 3Additionally, KILLIP classification was also utilized to predict the risk of mortality in ACS patients.However, these scoring systems have limited applicability because they rely heavily on diagnostic modalities, including sphygmomanometer, electrocardiographic (ECG), and laboratory findings, a facility that not all rural healthcare centers carry. 4herefore, a simple and convenient scoring system may be useful for assessing in-hospital mortality risk in ACS patients.
2][13][14] These studies showed that high CHA2DS2-VASc scores were significantly associated with increased MACE and mortality risk in ACS patients.Of note, this scoring system was simple, feasible, and can be completed only through history taking.Hence, this scoring system was efficient in terms of time and resources and thus can be used in every healthcare facility.
This study will repurpose CHADS2 and CHA2DS2-VASc scores as prognostic tools to predict in-hospital mortality in ACS patients.However, a lack of studies investigated the association between CHADS2 and CHA2DS2-VASc scores with in-hospital mortality in ACS patients.Thus, the purpose of this cohort study was twofold: Firstly, to retrospectively evaluate the relationship between the CHADS2 and CHA2DS2-VASc score with in-hospital mortality of admitted ACS patients regardless of AF status.Secondly, to compare the predictive accuracy of CHADS2, CHA2DS2-VASc, and GRACE risk scores in predicting the in-hospital mortality in ACS patients

Study design and patient selection
This was a retrospective and single-center cohort study that included all ACS patients aged ≥ 18 years old hospitalized in Dr. Hasan Sadikin Central General Hospital from January 2018 to December 2021.The exclusion criteria were participants with missing at least one CHADS2 and CHA2DS2-VASc score component and in-hospital mortality data.This study was approved by the Medical Research Ethics Committee of Dr. Hasan Sadikin Central General Hospital, West Java, Indonesia.The informed patient consent was collected at the beginning of the study.

Definition of variables and outcome
Regarding baseline characteristics variables, all diagnoses were collected from the patient's medical records.Cerebrovascular disease was defined as stroke or transient ischemic attack.Chronic obstructive pulmonary disease was defined as asthma or chronic obstructive pulmonary disease.
ACS was classified into STEMI, NSTEMI, and UAP, and its diagnosis was based on European Society of Cardiology guidelines. 15,16STEMI was diagnosed if the patient was admitted with acute chest pain with electrocardiographic findings of persistent ST-segment elevation >1 mm in 2 or more contiguous leads or new-onset left bundle branch block (LBBB). 15Patients with acute chest pain, together with no ST-segment elevation in electrocardiography and increased cardiac enzyme level (troponin or creatinine kinase myocardial band), were diagnosed as NSTEMI. 16Lastly, UAP was diagnosed if the patient presented with acute chest pain, with neither ST-segment elevation nor abnormal cardiac enzymes. 16ndonesian J Cardiol • Vol.44, Issue I • January -March 2023    5 In this study, CHADS 2 and CHA 2 DS 2 -VASc scores were categorized into two groups low and intermediate risk (score <2) and high risk (score ≥2).8][19] Moreover, the GRACE risk score was calculated according to its several components including age, heart rate, SBP, ST-segment deviation on electrocardiography, cardiac arrest at admission, abnormal cardiac enzymes, baseline creatinine levels, and Killip class.
The outcome of this study was in-hospital mortality which is defined as all ACS patients who died in the hospital before discharge regardless of the cause of death.

Statistical analysis
CHADS 2 and CHA 2 DS 2 -VASc scores were analyzed as numerical and categorical variables (based on a cutoff of 2).Age, body mass index, blood pressure, heart rate, and laboratory findings were interpreted as numerical variables.Sex category, smoking, comorbidities, revascularization history, ACS types, and Killip classification were reported as categorical variables.A Kolmogorov-Smirnov test was performed to assess the normality of data distribution.Numerical variables with parametric distributions are presented as the mean ± standard deviation, whilst numerical variables with nonparametric distributions are presented as the median and interquartile range.Categorical variables are reported with counts and percentages.A Mann-Whitney U test was used to evaluate the differences between two numerical variables.To compare the differences between two categorical variables, we performed the Chi-Square test or Fisher's exact test, as appropriate.The association between CHADS 2 and CHA 2 DS 2 -VASc score and inhospital mortality was evaluated using binary logistic regression analysis with the backward method according to each cutoff.Multivariate analysis was performed by adjusting several confounding factors with a p-value <0.25 based on univariate analysis.Statistical results were presented as odds ratio (OR) with a 95% confidence interval (CI) and a p-value.We used two-tailed p values with a significance set at ≤ 0.05.Receiver operating characteristics (ROC) analysis was also performed to assess the accuracy of these scoring systems in predicting in-hospital mortality.The area under the curve (AUC) between the two scoring systems was compared using De Long's method.All statistical analyses were performed using SPSS version 25.0 (IBM Corp., Armonk, NY).

Result Participants' baseline characteristics
From January 2018 until December 2021, a total of 1441 ACS patients were enrolled in this study.However, 62 patients were excluded because of missing data.Therefore, 1339 patients were included in this study.These participants had a median age of 58 (50.5-65.5),315 (23.5%) participants were female, and the total inhospital mortality calculated for all participants was 162 (12.1%).The patients' selection process is described in Figure 1.
In univariate analysis, a higher CHADS 2 score (≥2) increased the risk of in-hospital mortality about two times higher compared to the lower one (OR=2.05[1.47,2.87];p<0.001),However, a higher CHADS 2 score was no longer associated with inhospital mortality after performing the multivariate analysis (OR=1.31[0.92,1.86];p=0.129).
Additionally, we performed a univariate and multivariate analysis of every component of the CHADS  (1.93[1.10,3.37];p=0.022)turned out to be significantly associated with in-hospital mortality after multivariate analysis was performed.Odds ratios for in-hospital mortality according to CHADS 2 and CHA2DS2-VASc scores components were described in Table 3.
Receiver operating characteristics (ROC) analysis of CHADS 2 , CHA 2 DS 2 -VASc, TIMI, and GRACE risk scores in predicting in-hospital mortality in ACS patients In the ACS population, ROC analysis demonstrated that CHADS 2 , CHA 2 DS 2 -VASc, and GRACE risk scores were a good predictor of in-hospital mortality with the AUC of 0.609 (0.562-0.656), 0.653 (0.608-0.698), and 0.789 (0.748-0.830), respectively.This analysis showed that the predictive accuracy of CHA 2 DS 2 -VASc scores was significantly higher compared to the CHADS 2 score (p<0.001) and significantly lower than the GRACE risk score (p<0.001).ROC analysis is described in Figure 2.

Discussion
The main results of this study can be concluded as follows: first, the incrementation of each CHA 2 DS 2 -VASc score is independently associated with a 1.43 times higher risk of in-hospital mortality in ACS patients.Second, the high CHA2DS2-VASc score (cutoff > 2) was independently associated with a 3. 39 -fold increased risk of in-hospital mortality in patients with ACS, whilst the CHADS 2 score was not independently associated with in-hospital mortality.Third, the CHA 2 DS 2 -VASc score was a good predictor of in-hospital mortality.Its predictive value was significantly higher compared to the CHADS 2 score but still significantly lower than the GRACE risk score.
CHADS 2 and CHA 2 2 -VASc were previously known as risk stratification risk scores in AF patients.5,6Nonetheless, it is reasonable to use this scoring system to predict ACS patients' mortality because each of its components was prognostic factors associated with increased mortality risk in ACS. 20,21However, in this study, only hypertension and prior MI that independently associated with in-hospital mortality.Thus, although not all components of the CHA 2 DS 2 -VASc score independently increased the risk of mortality in this study, the combination of all these risk factors into one scoring system was confirmed to be useful in predicting in-hospital mortality in the ACS population.
A previous prospective cohort study conducted by Chua et al also provided a similar outcome to our study, which showed that the CHA 2 DS 2 -VASc score had better diagnostic performance compared to the CHADS 2 score in predicting the incidence of death in ACS patients. 12owever, a higher CHADS 2 score is still significantly associated with greater mortality risk after adjusting confounding factors in this study.In contrast, similar to our findings, a prospective cohort study by Ma et al. found that the CHADS 2 score was no longer associated with MACE after multivariate analysis.8It may explain that the addition of three variables in CHA 2 DS 2 -VASc score, including age 65-74 years, female sex, and vascular disease (MI, PAD, or aortic plaque), were significant risk factors of mortality in ACS. 20,22However, in this study, only prior MI was independently associated with in-hospital mortality.
Interestingly, this cohort showed that revascularization therapy was performed less often in the high CHA 2 DS 2 -VASc score group compared to the low scores group.This result is the same as that of Rozenbaum et al. and Chua et al. studies. 11,12 These studies showed that coronary intervention was significantly lower in the higher CHA 2 DS 2 -VASc score group.It may be caused by patient-related factors, including patient preference, vulnerability, functional and mental status, or clinicianrelated factors, including misjudgment of patient risk at baseline.However, since patient factors are fixed, the CHA 2 DS 2 -VASc score can help clinicians determine the best management approach for patients with ACS based on the CHA 2 DS 2 -VASc score (cutoff ≥2) as an additional scoring system to support the GRACE risk score.
The European Society of Cardiology (ESC) guideline recommends early risk stratification in ACS patients using the GRACE risk score to plan the appropriate treatment. 16However, unlike the GRACE risk score that required the electrocardiographic and laboratory findings to complete the calculation, the CHA 2 DS 2 -VASc score can be completed only through anamnesis, which was more convenient and feasible to use in the clinical setting, especially in primary health care facilities in a rural area which had lack of ancillary test.Thus, despite having less predictive value than the GRACE risk score, the CHA 2 DS 2 -VASc score may be a more practical model for early risk stratification in ACS patients.
Our study has several limitations.First, this study was limited by a single-center retrospective cohort study with all the shortcomings.Second, our center is the highest referral hospital in West Java; therefore, adverse events that could be potential confounders from previous hospitals were not recorded.Third, this study only included Indonesian populations, which does not reflect the racial heterogeneity of the global ACS population.Therefore, prospective multicentre cohort studies with larger numbers of participants, multi-ethnicity, and long-term duration follow-up are needed to understand better the association of CHADS 2 and CHA 2 DS 2 -VASc scores with mortality risk in ACS patients.

Conclusion
In ACS patients, a high CHA 2 DS 2 -VASc score (cutoffs ≥ 2) is independently and significantly associated with an increased risk of in-hospital mortality.Despite its lower predictive accuracy compared to the GRACE risk score, the CHA 2 DS 2 -VASc score still has a potential role in predicting in-hospital mortality, especially in primary healthcare settings located in rural areas that lack the necessary facilities.However, the importance of ECG and troponin measurements is not replaceable to truly predict the risk of mortality in ACS patients.

Figure 1 .
Figure 1.Patient selection process for this study. .

Table 1 .
Baseline characteristics of all participants that stratified by scoring system cutoffs.

Variable CHA 2 DS 2 -VASc score P value CHADS 2 score P value <2 (n=555) ≥2 (n=784) <2 (n=944) ≥2 (n=395)
CHA 2 DS 2 -VASc, and GRACE risk scores were collected accordingly based on patients' data in the medical record.We calculated the CHADS 2 score by assigning 1 point each for congestive heart failure, hypertension, age ≥75 years, and diabetes mellitus; and 2 points for prior stroke, TIA, or thromboembolism.6Wecalculated the CHA 2 DS 2 -VASc score by assigning 1 point each for congestive heart failure, hypertension, diabetes mellitus, vascular disease (including prior CAD, myocardial infarct, peripheral artery disease, or aortic plaque), age 65-74 years, and female sex; and 2 points each for age ≥75 years and prior stroke, TIA, or thromboembolism.

Table 2 .
Odds ratios for in-hospital mortality according to CHADS2 and CHA2DS2-VASc scores.

Table 3 .
Odds ratios for in-hospital mortality according to CHADS2 and CHA2DS2-VASc components.