SVEAT score outperforms HEART score in patients admitted to a chest pain observation unit

Daniel Antwi-Amoabeng, Chanwit Roongsritong, Moutaz Taha, Bryce David Beutler,Munadel Awad, AhmedHanfy, Jasmine Ghuman, Nicholas T Manasewitsch, Sahajpreet Singh, Claire Quang, Nageshwara Gullapalli

Abstract

Key Words: Acute chest pain; Risk stratification tool; Symptoms, history of Vascular disease, Electrocardiography, Age, and Troponin score; History, Electrocardiography, Age, Risk factors and Troponin score

lNTRODUCTlON

Acute chest pain is one of the most common presenting symptoms to the emergency department[1,2].Several non-cardiac conditions share clinical features with acute myocardial infarction and the emergency room clinician must rapidly identify those patients with chest pain who are most likely to have active coronary events from those who have chest pain due to other reasons. The key immediate task is to identify if a patient could benefit from being hospitalized for acute coronary syndrome (ACS)evaluation and those who can be safely discharged. This requires an estimation of the pretest probability of ACS. However, the accuracy of individual history, physical exam and electrocardiogram findings have been found to have limited utility for diagnosing ACS[3]. Therefore, multiple scoring systems and pathways have been proposed as risk stratification tools for these patients[4-7]. Among them, the History, Electrocardiography, Age, Risk factors and Troponin (HEART) score is arguably the most utilized particularly in the United States. Unfortunately, it has been shown in some studies to identify less than half of low-risk patients[4-6,8]. In an unselected population of chest pain patients in the emergency department, the HEART score and clinical gestalt had the same diagnostic accuracy for ACS[9]. The HEART score assigns a maximum score of 2 for chest pain deemed “highly suspicious” for ACS and suggests further inpatient evaluation for ACS for a score of 4 or more. By not clearly defining the classification of a patient’s chest pain, the score introduces subjectivity and considerable inter-rater variability[10]. The score also incorporates traditional cardiac risk factors such as diabetes, hypercholesterolemia, and hypertension, which have been shown to have limited value in diagnosis ACS especially in those older than 40 years[11]. To control health care utilization and cost, it is imperative to identify low risk patients with chest pain for discharge from the emergency department. However, it is perhaps more important to not miss real cases of ACS in otherwise low risk patients. Among patients without the traditional risk factors for ACS, the HEART score may not be sensitive in identifying those who would benefit from further evaluation. Thus, there is a need for alternative risk stratification for this patient group. Recently, a new scoring system based on five sets of clinical variables; characteristics of chest pain Symptoms, history of Vascular disease, Electrocardiography, Age, and Troponin (SVEAT score, Table 1) has been reported to outperform the HEART score[8]. The objective of this study is to assess the potential usefulness of SVEAT score as a risk stratification tool by comparing its performanceto HEART score in chest pain patients with low suspicion for acute coronary syndrome and admitted for overnight observation.

MATERlALS AND METHODS

The registry of patients admitted to our clinical decision units between January 1stto April 17th, 2019,were retrospectively reviewed. Our clinical decision unit allows for close observation of chest pain patients who are at low risk for true major adverse cardiovascular event (MACE). Admission to this unit allows for serial monitoring of the patient’s symptoms, cardiac enzymes, and electrocardiograms. To minimize any potential biases, one group of investigators was assigned to abstract relevant information necessary to calculate SVEAT score, and another was assigned to collect information for HEART score according to the published criteria[4,8]. The occurrence of MACE defined as all-cause mortality, acute myocardial infarction, confirmed coronary artery disease requiring revascularization or medical therapy at 30 d were then validated by two independent investigators who were blinded to the SVEAT and HEART score for each patient. The abstracted data were then provided to another set of investigators who were blinded to the outcome data to calculate the SVEAT and the HEART scores. Patients with ST segment elevation myocardial infraction were excluded from the study. The fourth-generation ultrahigh sensitivity troponin I assay was used in all participants at our institution during the study period like the original SVEAT score study. Acute myocardial infarction was diagnosed based on standard criteria[12]. The predictive power of the SVEAT and HEART scores for 30-d MACE were compared using c-statistic, based on area under the receiving-operator characteristic curve (AUC). Chi-squared test for equality of area under the curve was used to compare the performance of the SVEAT score to the HEART score. Categorical variables were summarized as counts (%) and between group comparisons were performed using Fisher’s exact test. Continuous variables were summarized as means ± SD and difference between means by outcome compared using Student’st-test. All analyses were performed at a two-tailed 5% level of significance using Stata version 16.1 (Stata Corporation, College Station, TX,United States).

RESULTS

A total of 330 subjects were included in the study. Baseline patient characteristics are shown in Table 2.There were slightly more male (52.1%) than female subjects. The mean age was 59.5 ± 13.9 years. The incidence of 30-d MACE in our population was 3.33%. The subjects who suffered 30-d MACE were significantly older than those who did not (74.3 ± 13.2 yearsvs59.0 ± 13.6 years,P＜ 0.0001). There were however no other significant differences in baseline characteristics between the two groups (Table 2).

Figure 1 illustrates the receiver-operator-characteristic curves of the SVEAT and HEART scores in predicting 30-d MACE. The AUC of the SVEAT score (0.8876, 95%CI: 0.82-0.96) is significantly higher than AUC of the HEART score (0.7962, 95%CI: 0.71-0.88),P= 0.03. Using logistic model, SVEAT score ≤4 significantly predicted 30-d MACE (odds ratio 1.52, 95%CI: 1.19-1.95,P= 0.001) but the HEART score≤ 3 did not (odds ratio 1.29, 95%CI: 0.78-2.14,P= 0.32) (Table 3).

DlSCUSSlON

Currently, despite numerous risk stratification protocols, most low-risk patients presenting with acute chest pain are not being released from emergency department. The 2020 European Society of Cardiology Guideline for ACS recommends using an ultrahigh sensitivity troponin (hs-Tn) assay with 0/1-h hs-Tn protocol for ruling out acute coronary syndrome but also emphasizes the importance of incorporating clinical information into the decision-making process[12]. It additionally proposes using Global Registry of Acute Coronary Events score for prognostic purposes but does not recommend any specific clinical risk score for initial risk stratification[12]. The American College of Cardiology/American Heart Association has not updated their guideline since 2014 when they stated that none of available risk prediction tools at the time was definitively demonstrated to be superior to clinician judgement[13].

The HEART score is perhaps the most widely used risk stratification tool in the United States due to its simplicity and large amount of supporting evidence[6,14,15]. The criteria for its History and EKG component are however somewhat subjective. Consequently, inter-observer variability and scoring inconsistency have been reported[16-18]. More importantly, it has been shown to be able to identify merely less than half of low-risk patients[5,17,18]. One of the potential contributing factors for the latter issue is that the HEART score does not incorporate some of the useful clinical information readily available on initial evaluation. To circumvent some of the pitfalls of the HEART score, the SVEAT score was developed. There are a few differences between the SVEAT and HEART scores. First, larger weight(higher points) is assigned to the findings associated with higher likelihood of subsequent acute coronary event clinically and negative point for those traditionally associated with negative likelihood of the events in a stepwise manner. This approach allows wider range of potential scores, and we believe theoretically could help better discriminating among various risk group of patients. Secondly,the criteria for EKG changes and assigned point for each change are much more clearly defined.Moreover, the presence of vascular disease is included in the SVEAT score instead of risk factor which has been shown to be only a weak predictor in acute chest pain evaluation[19]. In fact, the SVEAT score has recently been shown to outperform the HEART score[8]. Like the previous study, this analysis found SVEAT score to be superior risk stratification tool to HEART score for acute chest pain evaluation in low-risk patients.

There are certainly a few limitations in our study. Firstly, the overall 30-d MACE incidence of 3.3% in this study is rather low and substantially lower than in the previous report of 19.6%[13]. This may unfavorably increase the possibility of our finding to be due to statistical chance. An extremely low event rate in this study is likely explained by our study design to include only those retrospectively identified from a low-risk chest pain registry at our institution. The incidence of MACE in our population however is in line with the recent report of real-world data in the United States where ED visit for acute chest pain exceeds 8 million annually[20]. Among these patients, ＜ 5% of them subsequently experienced acute coronary syndrome. Second, the sample size of our study is relatively small for a retrospective design. As indicated in the methodology section, we did try to design our study to minimize potential biases. Lastly, this is a single center study and therefore future confirmation in a multicenter study in wider range of population, and larger sample size will be needed.

Table 2 Baseline patient characteristics

Table 3 Logistic model of major adverse cardiovascular events with HEART and SVEAT scores as covariates using cut-off of ≤ 4 points for SVEAT and ≤ 3 points for HEART for low-risk

Figure 1 Receiver operating characteristics for HEART and SVEAT scores. There area under the curve was significantly larger for SVEAT than HEART(P = 0.03). SVEAT: Symptoms, history of Vascular disease, Electrocardiography, Age, and Troponin; HEART: History, Electrocardiography, Age, Risk factors and Troponin; ROC: Receiver operating characteristics.

CONCLUSlON

In conclusion, our study suggests potential usefulness of the newly developed SVEAT score as a risk stratification tool among low-risk patients admitted to clinical decision unit for evaluation of acute chest pain. We found that SVEAT score significantly outperforms the commonly used HEART score.Incorporating SVEAT score as part of a clinical assessment of these patients may help improve resource utilization while maintaining minimal risk of future cardiovascular events in low-risk patients presenting to emergency department with acute chest pain.

ARTlCLE HlGHLlGHTS

FOOTNOTES

Author contributions:Antwi-Amoabeng D and Roongsritong C helped design the research study and wrote the original draft of the manuscript; Taha M, Beutler BD, Awad M and Hanfy A contributed to data curation, validation,and formal analysis; Ghuman J, Manasewitsch NT, Singh S and Quang C contributed to data curation and helped review and edit the manuscript; Gullapalli N supervised the project from initiation to completion.

lnstitutional review board statement:The study protocol was reviewed and approved by the University of Nevada,Reno School of Medicine Institutional Review Board.

lnformed consent statement:The study was conducted in accordance with the policies of the Institutional Review Board of the University of Nevada, Reno School of Medicine. The trial was conducted as a retrospective cohort study using anonymized data from existing records. Therefore, informed consent was not required.

Conflict-of-interest statement:The authors declare no actual or potential conflicts of interest or relationship with industry.

Data sharing statement:Technical appendix, statistical code, and dataset available from the corresponding author upon reasonable request.

STROBE statement:The authors have read the STROBE Statement—checklist of items, and the manuscript was prepared and revised according to the STROBE Statement—checklist of items.

Open-Access:This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution NonCommercial (CC BYNC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is noncommercial. See: https://creativecommons.org/Licenses/by-nc/4.0/

Country/Territory of origin:United States

ORClD number:Daniel Antwi-Amoabeng 0000-0001-8594-004X; Chanwit Roongsritong 0000-0001-9158-4715; Moutaz Taha 0000-0002-8999-0034; Bryce David Beutler 0000-0002-5071-1826; Munadel Awad 0000-0001-6633-005X; Ahmed Hanfy 0001-6895-5551-004X; Jasmine Ghuman 0000-0002-3015-2377; Nicholas T Manasewitsch 0001-6895-5120-004X; Sahajpreet Singh 0000-0001-5454-0039; Claire Quang 0000-0002-3015-2392; Nageshwara Gullapalli 0000-0002-9974-555X.

S-Editor:Gong ZM

L-Editor:A

P-Editor:Gong ZM