Relationship between Selvester ECG Score and Cardiovascular Outcomes in Patients with Non-ST Elevation Myocardial Infarction

Abstract

Background

Early risk stratification plays a crucial role in the treatment of non-ST-elevation myocardial infarction (NSTEMI). Selvester score is an electrocardiography (ECG)-based method for estimating infarcted myocardial mass, however it has not been studied in NSTEMI before. In this study, we aimed to investigate the relationship between Selvester score and cardiovascular outcomes in a 1-year follow-up period in NSTEMI patients.

Methods

One hundred and forty-three consecutive patients with NSTEMI were analyzed. TIMI and GRACE risk scores were calculated accordingly. Selvester score was calculated on surface ECG as reported in prior studies. Syntax score was calculated using an online calculator. The study population was divided into two groups based on a cut-off value from receiver operating characteristic curve analysis for the discriminative ability of Selvester score for mortality: low score (≤ 4), and high score (> 4) groups.

Results

Age was higher, left ventricle ejection fraction and high-density lipoprotein-cholesterol levels were significantly lower, and TIMI, GRACE and SYNTAX scores were significantly higher in the high Selvester score group. In multivariate Cox regression analysis, ejection fraction [hazard ratio (HR): 0.926, 95% confidence interval (CI): 0.883-0.971, p = 0.002] and Selvester score > 4 (HR: 3.335, 95% CI: 1.306-8.503, p = 0.012) were found to be independent predictors of adverse events after 1 year of follow-up.

Conclusions

Selvester score is a fast and feasible method that has prognostic value for mortality and other major adverse outcomes in low and intermediate risk NSTEMI patients treated with urgent percutaneous coronary intervention for 12 months.

INTRODUCTION

Non-ST-elevation myocardial infarction (NSTEMI) is a manifestation of acute coronary syndrome associated with a high risk of mortality.¹ Patients with NSTEMI should undergo early risk stratification to identify the need for early invasive strategies.² Selvester score is not new,³ but it is a useful scoring system that calculates a percentile amount of infarcted left ventricle mass.⁴ Although previous studies^5-7 have assessed Selvester score in predicting infarct size and outcomes in ST-elevation myocardial infarction (STEMI), the predictive role of Selvester score on cardiovascular outcomes has not been studied in NSTEMI patients. Routine risk assessment scores such as global registry of acute coronary events (GRACE) and thrombolysis in myocardial infarction (TIMI) combine clinical and laboratory parameters to predict future outcomes. However, they do not include parameters to assess the amount of infarcted myocardium. The aim of this study was to evaluate the predictive value of Selvester score, a plain surrogate of infarcted myocardial mass, for adverse cardiovascular events in NSTEMI patients who had undergone percutaneous coronary interventions.

METHOD

Study population

This study was designed as a prospective and observational single-center study. Between November 2018 and February 2019, 164 consecutive patients with the diagnosis of NSTEMI who were admitted to the Cardiology Clinic of Samsun Research and Training Hospital were included in the study. NSTEMI was diagnosed if the patients had typical angina pain lasting 30 minutes, a positive troponin-I level (defined in our clinical laboratory as > 0.01 ng/mL) without any evidence of ST segment elevation on 12-lead electrocardiogram. The exclusion criteria were as follows: history of congestive heart failure, history of percutaneous coronary intervention and/or coronary artery bypass grafting, active infectious disease, inflammatory or immunologic disease, cirrhosis, peripheral arterial disease, chronic obstructive pulmonary disease, chronic kidney disease, malignancy, or cardiogenic shock on admission. Furthermore, 11 patients who did not undergo coronary angiography and 10 patients who had undergone bypass surgery for revascularization after NSTEMI were also excluded. Therefore, the final study cohort consisted of 143 patients with the diagnosis of NSTEMI. All of the enrolled patients underwent complete percutaneous coronary revascularization before hospital discharge. Initial (urgent) procedures for NSTEMIs were carried out in the first two hours of admission according to our institutional rule for being a 24-hour primary percutaneous coronary revascularization referral center. The hospital’s virtual archive was scanned for all available patient post-discharge follow-up data and major adverse cardiac events (MACEs) such as hospitalization due to heart failure, repeat coronary interventions, cerebrovascular accident and all-cause mortality. The local Institutional Ethics Committee reviewed and approved the study protocol. All procedures were carried out in accordance with the Declaration of Helsinki (2000).

Electrocardiography and echocardiography

Twelve-lead surface electrocardiography (ECG) (Nihon Kohden Corporation, Cardiofax M Model ECG-1250, Tokyo, Japan) with a 25 mm/s paper speed and a voltage of 10 mm/s was obtained in the supine position before coronary angiography was performed. All of the ECGs were scanned and transferred to a personal computer to decrease errors in measurements and then magnified by 400% magnification by using Adobe Photoshop software. The latest version of Selvester score analysis protocol was applied by two blinded cardiologists.^8,9 Initially, the 12-lead ECG was classified by the type of ventricular conduction/hypertrophy: left bundle branch block, left anterior fascicular block, right bundle branch block, right bundle branch block + left anterior fascicular block, and left ventricular hypertrophy or no confounders.¹⁰ The amplitudes of the Q, R, and S waves and the durations, amplitude ratios, and notches were then measured. Each point of the total score predicted a scar size involving 3% of the total left ventricular mass. The QRS score was calculated according to the 50-criteria 31-point Selvester QRS scoring system. It only took about three minutes for a single ECG assessment by checking a quantification algorithm chart which is available in reference articles.¹⁰

Echocardiography was performed by two experienced cardiologists who were blinded to other data. All measurements were performed in accordance to the latest guidelines.¹¹ Left ventricular ejection fraction (LVEF) was calculated according to the Modified Simpson’s method.

Risk stratification

Three established scoring systems were used to predict the risk in all patients: GRACE, TIMI and SYNTAX. GRACE and TIMI risk scores are mostly clinical-based systems, whereas SYNTAX is derived from coronary angiography findings. Clinical scoring systems help to predict in-hospital and short-term mortality independent of the vascular situation of patients. On the other hand, the angiography-based SYNTAX score is a tool for predicting adverse events depending on the type of revascularization procedure. Nonetheless, it is valuable in our case because we aimed to assess an ECG marker that could predict infarct size percentage of the left ventricle.

The GRACE risk score was developed to estimate the risk of in-hospital death.¹² Recent guidelines encourage the use of a GRACE risk calculator that provides a direct estimation, bypassing the calculation of a score, of mortality while in hospital, at 6 months, at 1 year and at 3 years. The combined risk of death or myocardial infarction at 1 year is also provided.¹³ All GRACE risk score models calculated at hospital presentation use the same eight variables (age, systolic blood pressure, pulse rate, serum creatinine, cardiac arrest at admission, elevated cardiac biomarkers, ST-segment deviation, and Killip class at presentation) for risk prediction. All related data were obtained from the hospital’s virtual data archive. All calculations were made using web calculators provided by the GRACE study group (www.outcomes-umassmed.org/grace/acs_risk2/index.html). The patients were classified into three categories according to GRACE scores: < 109 was defined as being low, 109-140 as intermediate and > 140 as high risk categoriesas advised by recent guidelines.¹⁴

Antman et al.¹⁵ used a merged database of 7,081 UA/NSTEMI patients in the TIMI 11B and ESSENCE trials for the original derivation and validation of the TIMI risk score for UA/NSTEMI. The TIMI risk score is calculated by taking seven factors into account. One point is given for each of the following: being older than 65 years, using aspirin within the last week, having at least two angina episodes in the last 24 hours, having elevated serum cardiac biomarkers, having ST-segment deviation ≥ 0.5 mm on electrocardiogram, having known coronary artery disease (≥ 50% stenosis), having at least three risk factors for heart disease (blood pressure greater than 140/90, being a current smoker, high-density lipoprotein-cholesterol (HDL-C) cholesterol less than 40 mg/dL, diabetes, a family history of heart disease). A total score of ≥ 2 indicates the need for urgent revascularization, and 6-7 points indicates a nearly 40% risk of adverse cardiac event. The TIMI risk scores in this study were calculated by a blinded physician from case files via an online tool at https://timi.org/calculators/timi-risk-score-calculator-for-ua-nstemi/.

All NSTEMI patients underwent diagnostic coronary angiography using the standard Judkins technique (Siemens Axiom Artis Zee 2011; Siemens Healthcare, Erlangen, Germany) via a femoral or radial artery. Significant coronary heart disease was defined using different image plans in cineangiography. SYNTAX score was computed using an online SYNTAX score calculator (current calculator version: 2.28; http://www.syntaxscore.com/calculator/start.htm) by two well-experienced cardiologists blinded to the study data. Each coronary lesion with greater than 50% diameter narrowing in vessels of greater than 1.5 mm diameter identified in the coronary tree was respectively scored and summed to obtain the final SYNTAX score (Figure 1 and 2). If there were conflicting results, the final score was determined by averaging the scores calculated by each cardiologist. Patients with previous coronary artery bypass grafting were excluded from the study because the SYNTAX score is a suitable tool only for patients with native coronary artery lesions.

Figure 1.

Electrocardiographic and angiographic findings of a patient who underwent urgent revascularization because of acute thrombosis of a circumflex stent which was electively implanted a month ago. Panel A, B and C shows coronary angiography projections, white arrows indicate critical lesions. This patient’s SYNTAX score was 20. On panel D, electrocardiogram at admission is seen. On lead aVF, there is Q wave (≥ 50 milliseconds) and R/Q is less than 1 that sums up Selvester score to 5. This patient admitted to emergency room because of ventricular tachycardia after 70 days from discharge.

Figure 2.

Electrocardiographic and angiographic findings of a patient who underwent urgent revascularization because of critical lesion with thrombus at middle segment of left anterior descending artery. Panel A, B and C shows coronary angiography projections, white arrows indicate critical lesions. This patient’s SYNTAX score was 9. On panel D, electrocardiogram at admission is seen. On lead aVL, there is Q wave (≥ 40 milliseconds) that sums up Selvester score to 1. This patient did not show any adverse events during 12 months follow-up period.

Laboratory measurements

All patients’ physical examination notes, demographic and clinical characteristics were recorded. Blood samples were taken from the antecubital vein after admission for the measurement of complete blood count, troponin level, liver and kidney function tests and bleeding profile.

Definition of risk factors

The following clinical and demographic parameters were recorded; age, sex, hypertension (known hypertension treated with antihypertensive drugs, ≥ 2 blood pressure recordings > 140/90 mmHg), diabetes mellitus (known diabetes treated with diet or drugs or both; or a fasting serum glucose level of > 126 mg/dl), hypercholesterolemia (known treated hypercholesterolemia or fasting or non-fasting serum cholesterol concentrations > 200 mg/dl). Current cigarette smoking was defined as active smoking within the past 12 months. Body mass index (BMI) was determined by the following formula: BMI = weight (kg)/height² (m²).

Statistical analysis

The Kolmogorov-Smirnov test was used to determine the distribution of continuous variables. The Student’s t-test or Mann-Whitney U test was used to compare two means or medians according to whether the data were normally distributed. The chi-square or Fisher’s exact test was used to examine categorical variables. Continuous variables were presented as mean ± standard deviation, whereas categorical variables were presented as count and percentages.

Regression modelling: Associations of different variables of MACEs were examined using the penalized maximum likelihood estimation Cox regression method to reduce overfitting the risk and bias.¹⁶ The results were represented as hazard ratio (HR) with 95% confidence interval (CI). Variables in multivariable Cox regression analysis were chosen according to univariable analysis with an alpha of 0.05. SYNTAX score, ejection fraction (EF), age, left anterior descending artery-infarction related artery (LAD-IRA), Selvester score, GRACE score were included in the multivariable model.

Sample size calculation: We used the log-rank test to calculate the power of the study. In another study which also used the Selvester score and adverse events by Tjandrawidjaja et al.,¹⁷ similar cut-off points were used to predict survival of 0.96 for a QRS score < 4 and 0.915 for a score > 4 with a HR of 2.1. In order to test sufficient sensitivity (power > 0.8 and type 1 error < 0.05) for survival probability, we used 0.96 for the low Selvester score group (≤ 4) and 0.915 for the high Selvester score group (> 4). The number of patients in the two groups differed, so we applied an allocation ratio of 1/4. When we applied the log-rank test over all 143 patients, we calculated a predicted power as 95% power to detect a difference between groups at an α level of 0.05. All statistical calculations were performed using IBM SPSS Statistics for Windows, Version 21.0 (IBM Corp., Armonk, NY, USA) and R Statistics software version 4.01 (Vienna, Austria) with the "coxphf" and "Hmisc" packages. A two-tailed p value < 0.05 was considered to be statistically significant.

RESULTS

The study included 143 patients admitted to the emergency unit with a diagnosis of NSTEMI. The virtual archive of the hospital was scanned for MACEs. We could reliably document the following events: repeat coronary revascularization, cerebrovascular accident, hospitalization due to heart failure, and all-cause mortality. In total, adverse events were observed in 28 (19.5%) patients. When a comparison was done by grouping all enrolled patients by having an adverse event (Table 1), we detected that the patients who experienced a MACE during 12 months of follow-up after NSTEMI were older, had lower HDL cholesterol, lower EF, higher serum troponin-I peak level, higher Selvester ECG score and higher SYNTAX score. The other two risk classifications, GRACE and TIMI scores, did not differ between groups. To investigate the relationship between the studied parameters and adverse events, regression analysis was performed. Univariate Cox regression analysis identified that age, LVEF, left anterior descending artery being an infarct-related artery, Selvester, SYNTAX and GRACE scores were significantly associated with MACEs (Table 2). However, multivariate Cox regression analysis showed that only Selvester score (categorical variable: a score of > 4 or not) (HR: 3.335, 95% CI: 1.306-8.503, p = 0.012) and EF (HR: 0.926, 95% CI: 0.883-0.971, p = 0.002) were independent predictors of MACEs (Table 2). The time from the beginning of symptoms to hospital admission and the time from the beginning of symptoms to revascularization were insignificant in regression analysis.

Table 1. Comparison of patients who end up with major adverse cardiovascular events and remaining patients according to baseline clinical and laboratory features.

	MACE (-) (n = 115)	MACE (+) (n = 28)	p
Age, years	64.4 ± 12.5	72.2 ± 8.9	0.002
Male gender, n (%)	81 (71.1)	18 (64.3)	0.48
Hypertension, n (%)	61 (83.6)	12 (75.0)	0.41
Diabetes mellitus, n (%)	47 (64.4)	12 (75.0)	0.45
Smoking, n (%)	33 (28.7)	7 (25.0)	0.69
Heart rate, n (bpm)	75.4 ± 11.0	81.9 ± 14.3	0.61
Glucose, mg/dl	153.5 ± 76.7	183.2 ± 84.4	0.07
Creatinine, mg/dl	1.17 ± 1.09	1.21 ± 1.31	0.87
WBC, μl	9.71 ± 3.44	9.67 ± 3.89	0.95
Hemoglobin, g/dl	12.9 ± 1.8	13.6 ± 1.9	0.09
Platelet, 103/μl	237.3 ± 41.1	240.9 ± 61.6	0.75
LDL-C, mg/dl	114.8 ± 28.7	125.8 ± 42.1	0.19
HDL-C, mg/dl	42.8 ± 10.1	38.8 ± 7.9	0.03
Triglyceride, mg/dl	140.0 ± 100.2	162.9 ± 106.6	0.29
Troponin I peak, ng/dl	13.4 ± 11.8	28.2 ± 27.7	0.017
LAD is infarct related artery, n (%)	40 (34.8)	17 (60.7)	0.012
EF, %	50.7 ± 8.4	42.6 ± 7.3	0.001
Beta blocker usage, n (%)	109 (94.8)	27 (96.4)	0.71
ACE-I usage, n (%)	90 (78.3)	21 (75.0)	0.7
ASA usage, n (%)	111 (96.5)	26 (92.9)	0.38
P2Y12 usage, n (%)	107 (93.0)	28 (100)	0.15
Statin usage, n (%)	110 (95.7)	28 (100)	0.26
Long acting nitrate, n (%)	20 (17.4)	4 (14.3)	0.69
Time from symptom start to hospital admission, hours	3.58 ± 1.8	3.64 ± 1.9	0.86
Time from symptom start to initial revascularization, hours	4.83 ± 2.4	4.91 ± 2.6	0.87
TIMI risk score	2.78 ± 1.19	3.04 ± 1.29	0.32
GRACE risk score	107.0 ± 18.6	116.7 ± 17.4	0.014
Selvester QRS score	2.6 ± 2.0	5.6 ± 2.2	< 0.001
SYNTAX score	12.1 ± 7.7	19.9 ± 8.2	0.001

ACE-I, angiotensin-converting enzyme inhibitors; ASA, acetylsalicylic acid; EF, ejection fraction; bpm, beats per minute; GRACE, Global Registry of Acute Coronary Events 2.0; HDL-C, high density lipoprotein-cholesterol; LAD, left anterior descending artery; LDL-C, low density lipoprotein-cholesterol; MACE, major adverse cardiovascular events; TIMI, thrombolysis in myocardial infarction; WBC, white blood cells.

Table 2. Univariate and multivariate Cox regression analysis.

	Univariate analysis			Multivariate analysis
	HR	95% CI	p	HR	95% CI	p
SYNTAX score	1.087	1.046-1.130	< 0.001	1.007	0.965-1.050	0.72
EF	0.908	0.872-0.944	< 0.001	0.926	0.883-0.971	0.002
HDL	0.958	0.917-1.002	0.06
Age	1.052	1.017-1.089	0.003	1.019	0.983-1.059	0.29
Troponin I	1.005	1.000-1.010	0.057
LAD IRA	2.429	1.138-5.186	0.022	1.127	0.503-2.431	0.76
Symptom start to revascularization period	1.010	0.873-1.168	0.89
GRACE risk score	1.030	1.008-1.052	0.007	1.000	0.977-1.028	0.96
TIMI risk score	1.195	0.886-1.611	0.24
Selvester score	12.632	5.331-29.934	< 0.001	3.335	1.306-8.503	0.012

CI, confidence interval; EF, ejection fraction; GRACE, Global Registry of Acute Coronary Events 2.0; HDL-C, high density lipoprotein-cholesterol; HR, hazard ratio; LAD-IRA, left anterior descending artery-infarction related artery; LDL-C, low density lipoprotein-cholesterol; TIMI, thrombolysis in miyocard infarction; WBC, white blood cells.

Selvester score is a categorical variable (> 4 or not) in this analysis.

In receiver-operating characteristic (ROC) curve analysis, the Selvester score to detect a MACE with a sensitivity of 75% and specificity of 85% was 4 in the NSTEMI patients (Figure 3). The area under the ROC curve was 0.86. All enrolled patients were divided into two groups again according to a Selvester score of ≤ 4 or > 4. Thirty-nine of the patients had a score > 4 and they were included in the high score group, while the remaining 104 patients had a score ≤ 4 and were categorized into the low score group. Table 3 summarizes comparisons of the low score and high score groups according to baseline, clinical and laboratory features. Patients’ compliance to optimal doses of the prescribed medicine was quite high and similar in both groups (Table 3). Seventy-four patients in the low score group and 25 patients in the high score group were men (p = 0.37). No statistical differences were observed in terms of gender, hypertension, diabetes mellitus and smoking. LVEF (p = 0.025) and HDL-C (p = 0.004) levels were lower in the high Selvester score group. Age was significantly higher in the high score group compared to the low score group (70.2 ± 10.4 vs. 64.3 ± 12.5 years, respectively; p = 0.01) (Table 3). Also, troponin I peak levels were similar between the two groups (p = 0.138). Furthermore, the high score group had a significantly higher SYNTAX score in coronary angiogram (18.7 ± 7.7 vs. 11.7 ± 8.2, p = 0.001). GRACE risk score (121.7 ± 15.7 vs. 104.1 ± 17.5, p = 0.001) and TIMI risk score (3.03 ± 1.34 vs. 2.76 ± 1.16, p = 0.028) were significantly higher in the high score group. There were no statistically differences in the time from beginning of symptoms to hospital admission and the time from beginning of symptoms to revascularization between the two groups (Table 3).

Figure 3.

Receiver operating characteristic curves indicating the discriminative ability of Selvester score for major adverse cardiac events. The value for Selvester score to detect major adverse cardiac event with a sensitivity of 75% and specificity of 85% was found 4 in NSTEMI patients. The area under the curve was 0.86.

Table 3. Comparison of low Selvester score and high Selvester score groups according to baseline clinical and laboratory features.

	QRS score ≤ 4 (n = 104)	QRS score > 4 (n = 39)	p
Age (years)	64.3 ± 12.5	70.2 ± 10.4	0.01
Male gender, n (%)	74 (71.8)	25 (64.1)	0.37
Hypertension, n (%)	86 (82.7)	34 (87.2)	0.42
Diabetes mellitus, n (%)	58 (63.3)	42 (70)	0.51
Smoking, n (%)	28 (26.9)	12 (30.8)	0.64
Heart rate, n (bpm)	75.8 ± 11.5	77.0 ± 13.3	0.60
Beta blocker usage, n (%)	98 (94.2)	38 (97.4)	0.43
ACE-I usage, n (%)	79 (76)	32 (82.1)	0.43
ASA usage, n (%)	99 (95.2)	38 (97.4)	0.55
P2Y12 usage, n (%)	98 (94.2)	37 (94.9)	0.88
Statin usage, n (%)	99 (95.2)	39 (100).	0.16
Long acting nitrate, n (%)	16 (15.4)	8 (20.5)	0.46
Glucose, mg/dl	161.4 ± 80.5	154.0 ± 74.1	0.62
Creatinine, mg/dl	1.19 ± 1.11	1.27 ± 1.22	0.51
WBC, μl	9.99 ± 3.64	8.93 ± 3.09	0.11
Hemoglobin, g/dl	13.6 ± 1.9	13.1 ± 1.7	0.11
Platelet, 103/μl	244 ± 60	228 ± 48	0.14
LDL-C, mg/dl	122 ± 42	127 ± 33	0.43
HDL-C, mg/dl	43.4 ± 10.0	38.3 ± 8.5	0.004
Triglyceride, mg/dl	155.4 ± 95.7	167.2 ± 117.6	0.55
LAD is infarct related artery, n (%)	26 (31)	31 (49)	0.15
EF, %	50.0 ± 8.1	46.6 ± 7.2	0.025
Troponin I peak, ng/dl	11.1 ± 12.3	25.3 ± 57.7	0.13
Time from symptom start to hospital admission, hours	3.58 ± 1.85	3.64 ± 1.92	0.87
Time from symptom start to initial revascularization, hours	4.83 ± 2.49	4.91 ± 2.6	0.88
TIMI risk score	2.76 ± 1.16	3.03 ± 1.34	0.028
GRACE risk score	104.1 ± 17.5	121.7 ± 15.7	0.001
SYNTAX score	11.7 ± 8.2	18.7 ± 7.7	0.001

ACE-I, angiotensin-converting enzyme inhibitors; ASA, acetylsalicylic acid; bpm, beats per minute; EF, ejection fraction; GRACE, Global Registry of Acute Coronary Events 2.0; HDL-C, high density lipoprotein-cholesterol; LAD, left anterior descending artery; LDL-C, low density lipoprotein-cholesterol; TIMI, thrombolysis in miyocard infarction; WBC, white blood cells.

Information on long-term follow-up for 12 months was available in the hospital’s virtual archive for all patients. In total, adverse events were observed in 28 (19.5%) patients. In the high score group, repeat coronary revascularization was observed in 3 (7.6%) patients, cerebrovascular accident in 2 (5.1%) patients, hospitalization due to heart failure in 2 (5.1%) patients, and all-cause mortality in 10 (25.6 %) patients. In the low score group, repeat coronary revascularization was observed in 3 (2.8%) patients, cerebrovascular accident in 1 (0.9%) patient, hospitalization due to heart failure in 3 (2.8%) patients, and all-cause mortality in 4 (3.8%) patients. The occurrence of adverse events was significantly higher (43.5% vs. 21.1%; p < 0.001) in the high score group. Kaplan-Meier estimates of cumulative survival in the total population according to Selvester score are shown in Figure 4.

Figure 4.

Kaplan-Meier curve analysis demonstrating significant differences in long term all-cause mortality events in patients with low and high Selvester score.

DISCUSSION

Based on the literature, this study is the first to apply Selvester score in NSTEMI patients and also to demonstrate a relationship between Selvester score and MACEs. Furthermore, this is the first study to shown that a high Selvester score (> 4) was associated with 12-month mortality.

In the early 1970s, a QRS score was proposed by Selvester et al. to calculate infarct size using only surface ECG.³ The prognostic value of this scoring system has been demonstrated in various cardiac conditions.^18-21 The Selvester score has been significantly correlated with the anatomically detected size of single MIs in the anterior, inferior, and posterolateral divisions of the left ventricle, and it can be used for patients admitted with first episodes of chest pain suggestive of an acute coronary syndrome.^22-24 The prognostic value for mortality was tested by Jones et al.,²⁵ who reported that a higher QRS score was associated with death. Furthermore, Kalogeropoulos et al.²⁶ indicated that the QRS score was a predictor of heart failure in patients with STEMI. For patients surviving initial STEMI, the pre-discharge QRS score demonstrated prognostic value for predicting short-term mortality and/or re-hospitalization due to heart failure.

Data on the relationship between Selvester score and mortality/morbidity are sparse.^5,17,25-27 The vast majority of study patients in these articles had STEMI. Many years ago, Uyarel et al.²⁷ demonstrated that the presence of a high QRS score was an independent predictor of incomplete ST segment recovery and 30-day MACE risk in STEMI patients treated with primary percutaneous intervention. The authors defined a high QRS score as ≥ 4, and further analysis revealed that the high QRS group had significantly lower EF, higher peak cardiac enzyme levels, higher proportions of anterior wall infarction and higher proportions of no-reflow phenomenon. In addition, these patients had a significantly higher incidence of MACEs in post-discharge 30 days. The findings of this study are compatible with ours, and we also demonstrated that the patients with a higher QRS score had a significantly lower EF, higher troponin levels and higher incidence of left anterior descending artery being an infarct-related artery. Tjandrawidjaja et al.¹⁷ conducted a prospective study of 5745 patients with PCI-treated STEMI, and found that a higher QRS score at hospital discharge was an independent and prognostically relevant metric. Recently, a change in baseline Selvester score was found to predict the prognosis in patients with STEMI.⁵ Both baseline and predischarge changes in Selvester scores independently predicted poor outcomes in 2 years of follow up. Similarly, we found that the rate of adverse events was significantly higher in the NSTEMI patients if their baseline Selvester score was > 4. We also found that the presence of a high QRS score was related to adverse events, whereas SYNTAX score was not. At first glance this result looks unexpected, however the mean SYNTAX scores in both groups were in the low risk category (< 22 points). Finally, our median QRS score was relatively low for an acute coronary syndrome group, which is also not unexpected for NSTEMI,²⁸ and confirms the finding that the mean LVEF was near normal or mildly reduced in both groups due to low percentage of infarcted cardiac mass.

To the best of our knowledge, Selvester score has not previously been compared with traditional risk classification methods, namely GRACE and TIMI risk scores. Our results demonstrated that Selvester score showed a stronger relationship with the occurrence of MACEs than GRACE and TIMI risk scores. Although univariate Cox regression analysis showed that age, LVEF, Selvester, GRACE, TIMI, and SYNTAX scores were significantly associated with MACEs, only Selvester score and LVEF were independent predictors of MACEs. GRACE risk score has been validated and is a valuable risk predictor. Of note, the GRACE risk score model versions 1.0 and 2.0 were both derived from populations enrolled more than 10 years ago and likely overestimate the risk, although discrimination into low and high risk categories remains good.^29-31 However, our study patients mainly fell into intermediate and low risk categories. Poor predictive results may be related to this issue. An inadequate number of high-risk patients may have affected our results, which may not be applicable for high and very high-risk patients. In addition, we suggest that the study population needs to be large enough to distinguish the adverse event prediction value of GRACE risk score between low and intermediate risk groups.

TIMI risk score is not a long-term risk predictor and is valid for 14 days, and studies concerning longer follow-up periods are inconclusive.^32,33 Therefore, it is not unexpected for TIMI risk score to show such an insufficient predictive value compared to GRACE and Selvester scores. However, classification of study patients by TIMI risk score confirmed that our group consisted of largely low and intermediate risk patients.

Limitations

The main limitation of this study was the relatively small study population based on a single center experience. The follow-up period was relatively short and the number of adverse events was relatively low. This may have had an effect on the absence of a significant relationship between SYNTAX score and adverse event rates. Furthermore, patients with moderate to severe left ventricular dysfunction were excluded from analysis, and thus our findings may not be applicable for this group. Lastly, the patients mainly fell into low and intermediate risk groups as defined by traditional risk calculators, and so our results may be less applicable for high-risk patients. In this respect, our study can be considered as a hypothesis producing study. Prospective clinical studies with larger study populations and long-term follow-up to investigate the predictive value of Selvester score for MACEs are needed to reach definite conclusions.

CONCLUSION

Our findings indicate that Selvester score has prognostic value for long-term mortality and other major adverse outcomes in patients with NSTEMI. It is vital to define higher risk patients for early intervention and close follow-up after discharge in acute coronary syndromes. Baseline Selvester score may be a feasible and fast option for recognizing low and intermediate risk NSTEMI patients with a higher probability of adverse events.

CONFLICT OF INTEREST

All the authors declare no conflicts of interest.