Clinical characteristics of survival and in-hospital death group
Among 5063 STEMI patients, 392(7.7%) were died or withdrawal from treatment due to terminal status at discharge. Patients were divided into survivor and non-survivor (in-hospital death or treatment withdrawal). The clinical characteristics of the two groups were summarized in Table 1.
Table 1
Comparison of demographic and clinical characteristics between survivor and non-survivor group
variable
|
Survivor (N = 4671)
|
Non-survivor (N = 392)
|
P value
|
Demographic characteristic
|
|
|
Age (years)
|
62.3(52.1–70.1)
|
73.3(65.9–80.2)
|
< 0.001
|
≥ 75years (%)
|
680(14.6)
|
174(44.4)
|
< 0.001
|
Women (%)
|
1055(22.6)
|
179(45.7)
|
< 0.001
|
Tertiary hospital (%)
|
2337(50.0)
|
173(44.1)
|
0.025
|
Risk factors
|
|
|
|
Hypertension (%)
|
2047(43.8)
|
186(47.5)
|
0.17
|
Dyslipidemia (%)
|
2555(54.7)
|
166(42.4)
|
< 0.001
|
Diabetes (%)
|
785(16.8)
|
100(25.5)
|
< 0.001
|
Current smoker (%)
|
1922(41.2)
|
81(20.7)
|
< 0.001
|
Medical history
|
|
|
|
Heart failure (%)
|
41(0.9)
|
13(3.3)
|
< 0.001
|
Coronary heart disease (%)
|
286(6.1)
|
28(7.1)
|
0.42
|
Stroke (%)
|
583(12.5)
|
73(18.6)
|
< 0.001
|
CKD (%)
|
26(0.6)
|
6(1.5)
|
0.045
|
Clinical characteristic
|
|
|
|
Cardiac arrest (%)
|
126(2.7)
|
28(7.1)
|
< 0.001
|
Anterior myocardial infarction (%)
|
2601(55.7)
|
274(69.9)
|
< 0.001
|
Killip class (%)
|
|
|
< 0.001
|
I
|
3487(74.7)
|
156(39.8)
|
|
II
|
714(15.3)
|
72(18.4)
|
|
III
|
257(5.5)
|
59(15.1)
|
|
IV
|
213(4.6)
|
105(26.8)
|
|
SBP (mmHg)
|
130 (114–148)
|
116 (94–135)
|
< 0.001
|
DBP(mmHg)
|
80 (70–92)
|
73.00(59–85)
|
< 0.001
|
HR (BPM)
|
76 (65–86)
|
86 (70–104)
|
< 0.001
|
eGRF (ml/min/1.73m2)
|
103.6(85.2,123.4)
|
71.3(49.4,99.0)
|
< 0.001
|
LVEF (%)
|
56(48,61)
|
46(39,56)
|
< 0.001
|
Treatment therapy and time delay
|
|
|
|
Reperfusion strategy (%)
|
|
|
|
none
|
1903(40.7)
|
220(56.1)
|
< 0.001
|
PCI
|
1462(31.3)
|
55(14.0)
|
|
Fibrinolysis
|
1306(28.0)
|
117(29.9)
|
|
Successful reperfusion (%)
|
2616(56.0)
|
87(22.2)
|
< 0.001
|
Onset-to-FMC (min)
|
200(105–580)
|
250(120–949)
|
< 0.001
|
Medicine used in hospital
|
|
|
|
Aspirin (%)
|
4536(97.1)
|
305(77.8)
|
< 0.001
|
P2Y12 antagonists (%)
|
4543(97.3)
|
304(77.6)
|
< 0.001
|
Statin (%)
|
4450(95.3)
|
288(73.5)
|
< 0.001
|
Beta-blocker (%)
|
3362(72.0)
|
149(38.0)
|
< 0.001
|
ACEI/ARB (%)
|
2418(51.8)
|
87(22.2)
|
< 0.001
|
Diuretics (%)
|
1298(27.8)
|
110(28.1)
|
0.908
|
Digoxin (%)
|
89(1.9)
|
14(3.6)
|
0.025
|
Heparin (%)
|
3841(82.2)
|
244(62.2)
|
< 0.001
|
Data are expressed as n (%) or median (IQR), unless otherwise noted. |
Abbreviation: IQR, interquartile range; HR, heart rate; BPM, beat per minute; SBP, systolic blood pressure; Stoke contains ischemic and hemorrhagic. Coronary heart disease contains myocardial infarction, percutaneous coronary intervention and coronary artery bypass graft. CKD, chronic kidney disease. eGRF, estimated glomerular filtration rate. LVEF, left ventricular ejection fraction. PCI, percutaneous coronary intervention. FMC, first medical contact. P2Y12 antagonist, clopidogrel and ticagrelor. ACEI, Angiotensin-Converting Enzyme Inhibitors. ARB, Angiotensin Receptor Blockers.
The median age of the non-survivor group was significantly older than the survivor, and more than two fifths non-survivor were older than 75 years, the non-survivor were more likely admitted to secondary hospital. The non-survivor had a higher percentage of comorbidity such as diabetes, heart failure and stroke. The proportion of cardiac arrest, anterior myocardial infarction and Killip class IV level were higher in non-survivor. The eGRF and LVEF were significantly lower for non-survivor group.
Compared with the survived patients, the non-survivor had a lower proportion of reperfusion therapy, and only 22.2%were successfully reperfusion. The median of Onset-to-FMC time were significantly longer than survived patients. In the terms of medicine used in hospital, the non-survived patients were less likely to receive aspirin, P2Y12 antagonists, statin, beta-blocker, ACEI/ARB and heparin.
Performance For The Prognosis Prediction Of Shock Index-based Risk Indices
As shown in Table2 and Fig. 1, the AUCs of SI, MSI, age SI, age MSI were slightly smaller than the TIMI risk score. The sensitivity of SI and MSI at the cut-off value were significantly lower than TIMI risk score, nevertheless, the specificity of SI and MSI at the cut-off value were slightly greater than TIMI risk score. The sensitivity and specificity of age SI and age MSI were in accordance with TIMI risk score.
Table 2
Performance for the prognosis prediction of shock index-based risk indices and TIMI risk score.
|
Discrimination
|
|
Calibration
|
Cut-off point
|
AUC
|
SE
|
95 CI
|
Sensitivity
|
Specificity
|
|
HL chi-square
|
HL P value
|
Nagelkerke-R2
|
Brier score
|
SI
|
0.75
|
0.69
|
0.02
|
0.68–0.71
|
0.47
|
0.85
|
|
38.19
|
< .0001
|
0.10
|
0.07
|
MSI
|
1.00
|
0.70
|
0.02
|
0.68–0.71
|
0.48
|
0.85
|
|
25.51
|
0.001
|
0.10
|
0.07
|
age SI
|
40.75
|
0.78
|
0.01
|
0.77–0.79
|
0.73
|
0.69
|
|
35.12
|
< .0001
|
0.15
|
0.06
|
age MSI
|
56.66
|
0.78
|
0.01
|
0.77–0.79
|
0.70
|
0.73
|
|
54.98
|
< .0001
|
0.13
|
0.06
|
TIMI risk score
|
6.00
|
0.82
|
0.01
|
0.81–0.83
|
0.73
|
0.77
|
|
8.74
|
0.27
|
0.23
|
0.06
|
Abbreviation: SI, shock index; MSI, modified shock index; TIMI, thrombolysis in myocardial infarction; AUC, area under the curve; SE, standard error; CI, confident interval; HL, Hosmer-Lemeshow.
Figure 2 shows the in-hospital mortality for each quartile of SI, MSI, age SI and age MSI. The in-hospital mortality of STEMI patients was increasing with the shock index-based risk indices. The STEMI patients within the Q4 of SI (Fig. 2A), MSI (Fig. 2B), age SI (Fig. 2C) and age MSI (Fig. 2D) had highest in-hospital mortality, which was significance higher than the other quartile (Q1, Q2 and Q3).
Figure 3 shows when STEMI patients was grouped by the cut-off value of SI, MSI, age SI, age MSI, in-hospital mortality in patients with the larger value were significance higher than the patients with smaller value. In-hospital mortality in STEMI patients with the SI ≥ 0.75 was 20.4%, with the MSI ≥ 1.00 was 20.9%, with age SI ≥ 40.75 was 16.6%, with age MSI ≥ 56.66 was 17.8%, which were relatively higher than the smaller value group.
Univariate And Multivariate Analysis Of Grouped Si Based Risk Indices
The association between SI, MSI, age SI, age MSI, TIMI risk score and in-hospital death was shown utilizing crude and adjusted odds ratio in Fig. 4. In univariate analysis, after accounting for clustering of patients within hospitals, categorized SI, MSI, ag SI, age MSI and TIMI risk score were associated with a higher rate of in-hospital mortality. In the multivariate analysis, categorized SI, MSI, age SI, age MSI and TIMI risk score were revealed to be predictors for in-hospital mortality, after adjusted the clustering of patients within hospitals and potential affecting factors(SI: OR = 2.95, 95% CI:2.25–3.87, P < 0.001; MSI: OR = 3.06, 95% CI:2.33–4.02, P < 0.001; age SI: OR = 3.19, 95% CI:2.39–4.26, P < 0.001; age MSI: OR = 3.03, 95% CI:2.27–4.05, P < 0.001). As for TIMI risk score, the high risk group was 9.19 fold than the low risk group.
Comparisons Of The Prognostic Performance Of Si Based Risk Indices With Timi Risk Score
The calibration of SI, MSI, age SI, and age MSI were not so good, the Hosmer-Lemeshow P values were little than 0.05 and the Nagelkerke-R2 were lower, which indicated bad model calibration. All the parameter of the study had a similar Brier score (Table 2).
As shown in Table3, the prognostic performance of SI, MSI, age SI, and age MSI were not equivalent to TIMI risk score for predicting in hospital death or treatment withdrawal, the P values for C-statistic, Pvalues for NRI and P values for IDI were statistical significantly, which indicates SI, MSI, age SI, and age MSI were not comparable to TIMI risk score in predicting in hospital death or treatment withdrawal.
Table 3
Comparisons of the prognostic performance of shock index-based risk indices versus TIMI risk score
|
z for C-statistic
|
P for C-statistic
|
NRI
|
P for NRI
|
IDI
|
P for IDI
|
SI vs TIMI risk score
|
7.94
|
< 0.001
|
0.75
|
< 0.001
|
0.07
|
< 0.001
|
MSI vs TIMI risk score
|
8.09
|
< 0.001
|
0.77
|
< 0.001
|
0.07
|
< 0.001
|
age SI vs TIMI risk score
|
3.36
|
< 0.001
|
0.53
|
< 0.001
|
0.03
|
0.006
|
age MSI vs TIMI risk score
|
3.58
|
< 0.001
|
0.68
|
< 0.001
|
0.04
|
< 0.001
|
Abbreviation: TIMI, thrombolysis in myocardial infarction; SI, shock index; MSI, modified shock index; NRI, net reclassification improvement; IDI, integrated discrimination improvement.