Abstract
Background:
Accurate risk stratification of non-ST segment elevation myocardial infarction (NSTEMI) patients is important due to great variability in mortality risk, but, to date, no prediction model has been available. The aim of this study was therefore to establish a risk score to predict in-hospital mortality risk in NSTEMI patients.
Methods and Results:
We enrolled 5,775 patients diagnosed with NSTEMI from the China Acute Myocardial Infarction (CAMI) registry and extracted relevant data. Patients were divided into a derivation cohort (n=4,332) to develop a multivariable logistic regression risk prediction model, and a validation cohort (n=1,443) to test the model. Eleven variables independently predicted in-hospital mortality and were included in the model: age, body mass index, systolic blood pressure, Killip classification, cardiac arrest, electrocardiogram ST-segment depression, serum creatinine, white blood cells, smoking status, previous angina, and previous percutaneous coronary intervention. In the derivation cohort, the area under curve (AUC) for the CAMI-NSTEMI risk model and score was 0.81 and 0.79, respectively. In the validation cohort, the score also showed good discrimination (AUC, 0.86). Diagnostic performance of CAMI-NSTEMI risk score was superior to that of the GRACE risk score (AUC, 0.81 vs. 0.72; P<0.01).
Conclusions:
The CAMI-NSTEMI score is able to accurately predict the risk of in-hospital mortality in NSTEMI patients.
Acute myocardial infarction (AMI) is the most severe manifestation of coronary artery disease (CAD), which annually causes >2.4 million deaths in the USA, and >4 million deaths in Europe and China.1
Traditionally, AMI is divided into ST-elevation or non-ST-elevation myocardial infarction (STEMI, or NSTEMI) based on electrocardiography (ECG) characteristics.
Risk stratification for patients with AMI, particularly for patients with NSTEMI, is of clinical significance. Approximately 60–75% of all MI patients are diagnosed with NSTEMI,2,3
which involves a wide spectrum of clinical presentations and, in turn, prognosis. Accurate prognosis is useful for selecting level of care and the appropriate pharmacological or invasive treatment, but there is no risk stratification model to date specially designed for NSTEMI. Thus, the aim of this study was to bridge the evidence gap and develop a novel risk score to predict the risk of in-hospital mortality in NSTEMI patients.
Methods
China Acute Myocardial Infarction (CAMI) Trial Design
The trial design of the CAMI registry has been previously reported in detail.4
Briefly, the CAMI registry is a prospective, nationwide, multicenter observational study for Chinese AMI patients. The registry includes 3 levels of hospital (representing typical multiple models in China) from all provinces and municipalities throughout mainland China except Hong Kong and Macau. All participating hospitals or cardiovascular centers were instructed to enroll consecutive patients with a primary diagnosis of AMI. Clinical data, treatment, outcome and cost were collected by local investigators and submitted electronically, with a standardized set of variables and definitions, and rigorous data quality control. Eligible patients must be admitted ≤7 days of acute ischemic symptoms with a primary clinical diagnosis of AMI, including STEMI or NSTEMI. Efforts were made to address potential bias. Consecutive patients were enrolled with broad coverage of geographical region (Table S1), which assured a good representation of all MI patients in China and minimized selection bias. All data were collected using standardized definitions and methods, and were submitted through a secure, password-based electronic data capture system, to ensure data quality and minimization of information bias. We reported all results in accordance with the protocol, with no significant aspects of the study omitted.
This project was approved by the institutional review board central committee at Fuwai Hospital, National Center for Cardiovascular Diseases, China. This study was registered on www.clinicaltrials.gov, and the registration number was NCT01874691. Written informed consent was obtained from eligible patients before registration.
Subjects
From January 2013 to September 2014, a total of 26,103 patients with AMI were registered. For the present study, the specific inclusion criterion was NSTEMI. Of the 26,103 patients, 6,209 were diagnosed with NSTEMI. We excluded 393 patients due to incomplete or invalid data on age, body mass index (BMI), admission diagnosis, and in-hospital outcome. We also excluded 41 patients with new-onset left bundle branch block (LBBB) because new LBBB is considered as a STEMI equivalent. Finally, a total of 5,775 patients with NSTEMI were included in the present study (Figure 1).
Data Collection and Definitions
Patient demographics, clinical presentation, medical history, risk factors, triggering factors, physical examination, laboratory and imaging results, transfer facility therapy, reperfusion strategies, medication, clinical events, and cost were collected from the CAMI registry database. The diagnostic criteria of NSTEMI were based on the third Universal definition for myocardial infarction (2012):5
(1) ECG changes in accordance with AMI: new or presumed new significant ST-segment–T-wave (ST-T) changes or new LBBB or development of pathological Q waves on ECG; (2) absence of ST-segment elevation (ST-depression [new horizontal or down-sloping ST depression ≥0.05 mV in 2 contiguous leads] or T wave changes [T inversion ≥0.1 mV in 2 contiguous leads with prominent R wave or R/S ratio >1]. According to the classification of MI, types 1–3,4b,4c were included in the present registry. Types 4a,5 were not eligible for the CAMI registry. The primary endpoint of this study was in-hospital death, which was defined as all-cause death during hospitalization.
Statistical Analysis
Continuous data are presented as mean±SD and were compared using Student’s t-test. Categorical variables are summarized as count and percentage and were compared using chi-squared test or Fisher’s exact test as appropriate. Missing data were imputed by the average or the median of the available data. All P-values are 2-tailed, and P<0.05 was considered statistically significant. All analyses were performed with SAS 9.4 (SAS Institute, Cary, NC, USA).
The database was chronologically divided into 2 subsets according to procedure order: a derivation cohort of 4,332 patients that served for the construction of the risk model, and a validation subset of 1,443 patients for testing and validating the model. The CAMI-NSTEMI risk model was created by fitting clinical, laboratory analysis, and medical history variables into a logistic multivariable analysis for risk prediction of hospital death. To ensure that variables potentially correlated with the outcome were not excluded, univariable selection was first performed using the entry criterion P≤0.25. Then, the multivariable model was constructed using stepwise variable selection with entry and exit criteria P<0.05. The score was then derived by attributing integer numbers to the variables retained in the multivariable model. The variable with the smallest estimated coefficient was attributed 1 point and was considered as the reference variable. The scores of the other variables were determined by dividing their estimated coefficients by the coefficient of the reference variable (Supplementary Methods).6
Multicollinearity between variables was assessed using the variance inflation factor. Discrimination and calibration were determined by the C-statistic and the Hosmer-Lemeshow (HL) goodness-of-fit test, respectively,7,8
and the calibration and discrimination of the model were then assessed in the validation dataset. Additionally, the scoring system was classified by tertile to define 3 risk groups (low, intermediate, and high risk).
To compare the capability of risk prediction for in-hospital mortality between the Global Registry of Acute Coronary Events (GRACE) score and CAMI-NSTEMI score, receiver operating characteristic (ROC) analysis7
was performed to calculate the c-statistic of the overall model as well as the net reclassification improvement (NRI), and the integrated discriminatory index (IDI).9
Results
Baseline Characteristics
Hospital death occurred in 342 (5.92%) of 5,775 patients. The overall patient characteristics are listed in
Table 1. There were significant differences between the 2 groups in almost all baseline characteristics except medical history of hypertension, previous angina, previous coronary artery bypass grafting, previous stroke, previous peripheral vascular disease, and family history of premature CAD. The in-hospital death group was older, less often male or smokers, with lower BMI, and had a higher prevalence of diabetes mellitus and of hypertension.
Table 1.
Baseline NSTEMI Subject Characteristics vs. In-Hospital Mortality
| |
In-hospital survivors
(n=5,433) |
In-hospital deaths
(n=342) |
P-value |
| Age (years) |
64.92±11.98 |
72.13±11.16 |
<0.01 |
| Male |
3,754/5,433 (69.1) |
187/342 (54.7) |
<0.01 |
| BMI (kg/m2) |
24.09±3.05 |
23.02±3.11 |
<0.01 |
| DM |
1,249/5,418 (23.1) |
98/342 (28.7) |
0.02 |
| Hypertension |
3,154/5,423 (58.2) |
209/342 (61.1) |
0.28 |
| Hyperlipidemia |
456/5,421 (8.4) |
14/342 (4.1) |
<0.01 |
| LVEF (%) |
55.10±11.76 |
46.75±13.01 |
<0.01 |
| Previous angina |
2,074/5,408 (38.4) |
144/342 (42.1) |
0.17 |
| Previous MI (>1 month) |
585/5,411 (10.8) |
63/342 (18.4) |
<0.01 |
| Previous HF |
263/5,412 (4.9) |
48/342 (14.0) |
<0.01 |
| Previous PCI |
358/5,400 (6.6) |
12/342 (3.5) |
<0.01 |
| Previous CABG |
48/5,411 (0.9) |
4/342 (1.2) |
0.55 |
| Previous stroke |
542/5,410 (10.0) |
46/342 (13.5) |
0.05 |
| Previous renal dysfunction |
137/5,400 (2.5) |
15/341 (4.4) |
0.06 |
| Previous COPD |
131/5,382 (2.4) |
16/340 (4.7) |
0.02 |
| Family history of premature CAD |
168/5,417 (3.1) |
6/342 (1.8) |
0.13 |
| Previous peripheral vascular disease |
64/5,406 (1.2) |
5/342 (1.5) |
0.60 |
| Smoking status |
|
|
<0.01 |
| Current smoker |
1,967/5,406 (36.4) |
60/339 (17.7) |
|
| Previous smoker |
721/5,406 (13.3) |
46/339 (13.6) |
|
| Non-smoker |
2,718/5,406 (50.3) |
233/339 (68.7) |
|
| Prior use of medication (≤1 week) |
| Aspirin |
1,003/5,405 (18.6) |
69/339 (20.4) |
0.42 |
| Thienopyridines |
338/5,386 (6.3) |
29/339 (8.6) |
0.11 |
| Statins |
764/5,327 (14.3) |
54/334 (16.2) |
0.36 |
| HR (beats/min) |
79.19±19.90 |
89.97±25.39 |
<0.01 |
| SBP (mmHg) |
134.64±25.67 |
121.87±28.71 |
<0.01 |
| Killip classification |
|
|
<0.01 |
| I |
3,873/5,393 (71.8) |
127/339 (37.5) |
|
| II |
989/5,393 (18.3) |
93/339 (27.4) |
|
| III |
382/5,393 (7.1) |
52/339 (15.3) |
|
| IV |
149/5,393 (2.8) |
67/339 (19.8) |
|
| ST-segment depression |
2,917/5,340 (54.6) |
223/335 (66.6) |
<0.01 |
| Cardiac arrest |
29/5,404 (0.5) |
14/340 (4.1) |
<0.01 |
| Time to hospital |
|
|
0.68 |
| 1–7 days |
2,139/5,345 (40.0) |
140/334 (41.9) |
|
| 12–24 h |
758/5,345 (14.2) |
40/334 (12.0) |
|
| 6–12 h |
772/5,345 (14.4) |
54/334 (16.2) |
|
| <6 h |
1,676/5,345 (31.4) |
100/334 (29.9) |
|
| PLT (1012/L) |
208.26±72.23 |
217.53±117.26 |
0.16 |
| Hb (g/L) |
131.81±21.95 |
121.82±26.27 |
<0.01 |
| WBC (109/L) |
9.06±3.43 |
12.00±5.84 |
<0.01 |
| Cr (μmol/L) |
88.33±60.49 |
131.09±104.95 |
<0.01 |
| K+ (mmol/L) |
3.96±0.50 |
4.18±0.77 |
<0.01 |
Data given as mean±SD or n (%). BMI, body mass index; CABG, coronary artery bypass grafting; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease; Cr, creatinine; DM, diabetes mellitus; Hb, hemoglobin; HF, heart failure; HR, heart rate; LVEF, left ventricular ejection fraction; MI, myocardial infarction; NSTEMI, non-ST segment elevation myocardial infarction; PCI, percutaneous coronary intervention; PLT, platelet count; SBP, systolic blood pressure; WBC, white blood cells.
The in-hospital death group had higher heart rate, lower systolic blood pressure (SBP), higher Killip classification, and higher prevalence of ST segment depression on ECG, and of cardiac arrest.
On laboratory analysis, patients who died in hospital had higher platelet count, higher serum creatinine, white blood cell (WBC) count and serum potassium concentrations, but lower hemoglobin (Hb).
CAMI-NSTEMI Risk Model and Risk Score
Table 2
lists the association between baseline characteristics and in-hospital mortality. A total of 21 potential covariates with P≤0.25 were initially selected to fit the multivariable model, including age, gender, BMI, diabetes mellitus, hypertension, hyperlipidemia, left ventricular (LV) ejection fraction, previous MI, previous heart failure, previous percutaneous coronary intervention (PCI), previous chronic obstructive pulmonary disease, smoking status, heart rate, SBP, Killip classification, ST-segment depression, cardiac arrest, Hb, WBC count, serum creatinine, serum potassium concentration. After multivariable selection, 11 primary variables (age, BMI, SBP, Killip classification, cardiac arrest, ECG ST-segment depression, serum creatinine, WBC count, smoking status [non-smoker, previous smoker], previous MI, previous PCI) were retained in the model (Table 3). In the derivation cohort, the C-statistic for this model was 0.81 (95% CI: 0.78–0.83;
Figure 2) and excellent calibration was observed (HL P=0.59).
Table 2.
Univariate Risk of NSTEMI In-Hospital Mortality
| Baseline variable |
OR |
95% CI |
P-value |
| Age |
1.047 |
1.035 |
1.06 |
<0.0001 |
| Male |
0.642 |
0.494 |
0.833 |
<0.0001 |
| BMI |
0.901 |
0.862 |
0.942 |
<0.0001 |
| DM |
1.252 |
0.938 |
1.67 |
0.12 |
| Hypertension |
1.196 |
0.917 |
1.559 |
0.19 |
| Hyperlipidemia |
0.506 |
0.281 |
0.913 |
0.0235 |
| LVEF |
0.979 |
0.969 |
0.989 |
<0.0001 |
| Previous MI |
1.515 |
1.065 |
2.154 |
0.02 |
| Previous HF |
2.645 |
1.760 |
3.973 |
<0.0001 |
| Previous PCI |
0.385 |
0.18 |
0.823 |
0.01 |
| Previous COPD |
1.928 |
1.019 |
3.648 |
0.04 |
| Smoking status |
| Non-smoker |
2.648 |
1.903 |
3.684 |
<0.0001 |
| Previous smoker |
1.999 |
1.267 |
3.154 |
0.0029 |
| HR |
1.017 |
1.012 |
1.022 |
<0.0001 |
| SBP |
0.982 |
0.976 |
0.987 |
<0.0001 |
| Killip classification |
2.098 |
1.859 |
2.368 |
<0.0001 |
| ST-segment depression |
1.776 |
1.347 |
2.343 |
0.0001 |
| Cardiac arrest |
6.654 |
3.046 |
14.546 |
<0.0001 |
| PLT |
1.001 |
0.999 |
1.002 |
0.4506 |
| Hb |
0.988 |
0.983 |
0.993 |
<0.0001 |
| WBC |
1.136 |
1.108 |
1.166 |
<0.0001 |
| Cr |
1.005 |
1.003 |
1.006 |
<0.0001 |
| K+ |
1.765 |
1.398 |
2.227 |
<0.0001 |
Abbreviations as in Table 1.
Table 3.
Independent Predictors of NSTEMI In-Hospital Mortality
| Predictors |
OR |
95% CI |
| Age |
1.029 |
1.015 |
1.042 |
| BMI |
0.944 |
0.901 |
0.990 |
| SBP |
0.983 |
0.978 |
0.988 |
| Killip classification |
1.566 |
1.369 |
1.792 |
| Cardiac arrest |
2.625 |
1.050 |
6.560 |
| ST-segment depression |
1.549 |
1.152 |
2.083 |
| Cr |
1.004 |
1.002 |
1.005 |
| WBC |
1.099 |
1.070 |
1.129 |
| Non-smoker |
2.050 |
1.434 |
2.932 |
| Previous smoker |
1.434 |
0.874 |
2.355 |
| Previous MI |
1.890 |
1.273 |
2.805 |
| Previous PCI |
0.331 |
0.146 |
0.752 |
Abbreviations as in Table 1.
The scores attributed to each variable according to their estimated coefficients from the derivation cohort are given in
Table 4. The corresponding risks of in-hospital mortality associated with each accumulated points are given in
Table S2. The C-statistic for the simplified risk score was only slightly worse than that of the original model (C-statistic, 0.79; 95% CI: 0.76–0.82; HL P=0.64;
Figure 2).
Table 4.
Scores Attributed to Each Variable
| Predictor |
Categories |
Score |
|
Predictor |
Categories |
Score |
| Age (years) |
<57 |
0 |
|
Cr (μmol/L) |
<64 |
0 |
| |
(57–66) |
9 |
|
|
(64–77) |
1 |
| |
(66–75) |
15 |
|
|
(77–94.3) |
3 |
| |
≥75 |
22 |
|
|
≥94.3 |
6 |
| BMI (kg/m2) |
<20.0 |
10 |
|
WBC (109/L) |
<6.8 |
0 |
| |
(20.0–23.9) |
7 |
|
|
(6.8–8.6) |
5 |
| |
(23.9–25.9) |
4 |
|
|
(8.6–10.74) |
9 |
| |
≥25.9 |
0 |
|
|
≥10.74 |
18 |
| SBP (mmHg) |
<118.5 |
24 |
|
Cardiac arrest |
No |
0 |
| |
(118.5–130) |
18 |
|
|
Yes |
25 |
| |
(130–150) |
10 |
|
Smoking status |
| |
≥150 |
0 |
|
|
Non smoker |
19 |
| Killip |
I |
0 |
|
|
Previous smoker |
9 |
| |
II |
12 |
|
|
Current smoker |
0 |
| |
III |
24 |
|
Prior MI |
No |
0 |
| |
IV |
35 |
|
|
Yes |
17 |
| ST-segment |
No |
0 |
|
Prior PCI |
No |
29 |
| depression |
Yes |
12 |
|
|
Yes |
0 |
Abbreviations as in Table 1.
CAMI-NSTEMI score ranged from 0 to 217. The IQR and the frequency distribution of each variable across tertiles of CAMI-NSTEMI score are given in
Table 5. Event rates in the derivation cohort across tertiles of CAMI-NSTEMI score were as follows: 1.12% in tertile I (score, 0–79); 2.97% in tertile II (score, 80–101); and 12.75% in tertile III (score, ≥102; P<0.001). Thus, tertiles I–III were determined as the low, intermediate, and high risk groups, respectively (Table 5). Event rate was significantly different between the intermediate- and low-risk groups (OR, 2.697; 95% CI: 1.509–4.819, P<0.001;
Table S3).
Table 5.
In-Hospital Mortality vs. CAMI-NSTEMI Risk Groups
| |
Low risk
(tertile I) |
Intermediate risk
(tertile II) |
High risk
(tertile III) |
P-value |
| Score range |
0–79 |
80–101 |
≥102 |
|
| In-hospital mortality rate (%, n) |
| Training dataset (n=4,332) |
1.12 (16/1,427) |
2.97 (42/1,415) |
12.7 (190/1,490) |
<0.001 |
| Validation dataset (n=1,443) |
0.86 (4/466) |
2.30 (11/479) |
15.86 (79/498) |
<0.001 |
CAMI, China Acute Myocardial Infarction; NSTEMI, non-ST segment elevation myocardial infarction.
For the 1,443 patients included in the validation cohort, the CAMI-NSTEMI model had excellent prognostic accuracy with a C-statistic of 0.88 (95% CI: 0.83–0.92;
Figure 2; HL P=0.16). The CAMI-NSTEMI score also had good prognostic accuracy with a C-statistic of 0.86 (95% CI: 0.82–0.90;
Figure 2; HL P=0.20). As shown in
Table 5, in-hospital mortality in the validation cohort increased significantly across different risk groups: 0.86% in the low-risk group; 2.30% in the intermediate-risk group; and 15.86% in the high-risk group (P<0.001).
For the whole cohort, the C-statistic was 0.83 (95% CI: 0.80–0.85) for the model and 0.81 (95% CI: 0.78–0.83) for the score, and HL P-value was 0.84 and 0.35 for the model and the score, respectively.
Comparison With GRACE Score
For all 5,775 patients, the area under the ROC curve for CAMI-NSTEMI score was 0.81, which was significantly higher than that of the GRACE score (0.73; P<0.01 for comparison;
Figure 3). The NRI of the CAMI-NSTEMI score over the GRACE score was 56.8% (P<0.01) and the IDI of the CAMI-NSTEMI score over the GRACE score was 7.7% (P<0.01).
Discussion
Major Finding
Using data from a large-scale prospective registry of AMI patients, we developed and validated a risk score, the CAMI-NSTEMI risk score, to predict in-hospital mortality particularly for patients with NSTEMI. The risk score included the following 11 variables: age, BMI, SBP, Killip classification, cardiac arrest, ECG ST-segment depression, serum creatinine, WBC count, smoking status [non-smoker, previous smoker], previous MI, and previous PCI. The CAMI-NSTEMI risk score showed good discrimination and calibration and its diagnostic performance was superior to that of the GRACE risk score.
Rationale for the Novel Score
Risk stratification is important for NSTEMI patients because of the great variability in individual prognosis. Many models have been proposed to predict the risk of death in the setting of acute coronary syndrome (ACS), of which the GRACE and TIMI risk scores are the most commonly used by clinicians. Both risk scores, however, have limitations. TIMI risk score was the first to be implemented in clinical practice. It was based on clinical trial data, for which rigorous inclusion and exclusion criteria are required, thereby limiting its diagnostic performance in the general population.10
GRACE risk score was based on patients from America, Europe and Australia but not Asia. In addition, since its creation, there has been significant change in patient characteristics and management.11
Finally, there is still no model, to date, developed specifically for predicting in-hospital mortality for NSTEMI patients. For these reasons, a new score system is needed.
Risk Factors
Many variables in the present model are consistent with previous models, including age, SBP, Killip classification, cardiac arrest, ST-segment depression, and serum creatinine. Compared with a previous study, the present model incorporated a new variable: WBC count. Evidence regarding the prognostic value of WBC count in patients with STEMI was controversial. WBC count was previously reported to be associated with adverse events in AMI patients, including heart failure, cardiac shock,12
and death,13
as well as infarct size on magnetic resonance imaging.14,15
The association, however, was not found to be significant at either 1-month or 1-year follow-up.16,17
The WBC behavior in the context of MI is complex, involving both protective and deleterious effects on the myocardium.18
This may explain the aforementioned difference in conclusions. In the present multivariable logistic analysis, WBC count was an independent risk factor of in-hospital mortality. The OR, however, was close to 1, indicating that the effect of WBC on mortality was limited and that WBC count alone cannot be used to determine the risk of adverse events. Nonetheless, it is reasonable to add WBC count into the risk model in order to improve the diagnostic performance of the risk score.
Obesity and Smoker’s Paradoxes
In the present study, higher BMI was associated with lower in-hospital mortality, a phenomenon known as the obesity paradox. Many studies have been carried out on the obesity paradox, and a recent meta-analysis of 26 studies and >210,000 patients also confirmed the existence of the obesity paradox in patients with ACS.19
Several mechanisms have been proposed for the obesity paradox. Obese patients were 1–10 years younger than normal-weight patients,19,20
more likely to receive optimal medical therapy,19
coronary angiogram21
and invasive coronary intervention.22
In addition, multi-vessel disease and complex lesions as assessed on SYNTAX score were less common in obese patients.23,24
Finally, overweight patients may have had more nutritional reserves with which to survive MI when metabolic demands increased sharply.25
Similarly, evidence for smoker’s paradox was also seen in the present study after adjustment for age and comorbidity, consistent with previous studies.26
Possible mechanisms include greater clopidogrel-induced platelet suppression,27,28
and lower risk of adverse LV remodeling after infarction.29
Because it is well-established that obesity and smoking are associated with long-term adverse event risk, we still recommend that patients maintain a BMI between 18.5 and 24.9 kg/m2
and stop smoking after AMI, in accordance with the guideline.30
It is appropriate, however, to incorporate smoking and BMI into the present model due to their additive prognostic value.
Risk Categories
To determine the risk categories of CAMI risk score, we initially divided patients into 3 groups based on guideline recommendations and GRACE risk score.31
In the derivation cohort, in-hospital mortality risk increased significantly in tertile II and tertile III. In the validation cohort, there was a trend towards higher incidence rate in tertile II. We also divided patients into 4 groups to determine which classification method (as compared with 3 groups). We observed, however, that the scores clustered close in quartile III and quartile IV, and mortality rate was similar between quartile III and quartile IV. Therefore, we believed that it was more reasonable to divide patients into 3 groups.
Study Limitations
There were several limitations of this study. First, although the discrimination of the CAMI-NSTEMI risk score was confirmed in a separate cohort of patients, its predictive accuracy for hospital death should be further validated in a different study with a larger dataset. Second, all the patients were from China. Whether this risk score can be extrapolated to different ethnicities needs further investigation.
Conclusions
We developed and validated the first risk model and risk score to predict in-hospital mortality particularly for NSTEMI patients, in whom prognosis varies greatly. We believe that this model is simple and useful for guiding clinicians to select the appropriate treatment and level of care. Further study, involving larger sample sizes and different ethnicities, is needed to further validate this score.
Acknowledgments
We thank the TIMI Study Group and the Duke Clinical Research Institute for their contributions to the design, conduct, and data analysis of the CAMI registry. We also thank all the investigators and coordinators for their active participation and great work.
Disclosures
The authors declare no conflicts of interest.
Grants
This work was supported by CAMS Innovation Fund for Medical Sciences (CIFMS) (2016-I2M-1-009), the Twelfth Five-Year Planning Project of the Scientific and Technological Department of China (2011BAI11B02), and 2014 Special fund for scientific research in the public interest by National Health and Family Planning Commission of the People’s Republic of China (No. 201402001).
Supplementary Files
Supplementary File 1
Supplementary Methods
Table S1.
CAMI registry participating hospitals
Table S2.
In-hospital NSTEMI mortality risk associated with points
Table S3.
OR of in-hospital NSTEMI mortality
Table S4.
In-hospital mortality vs. CAMI-NSTEMI quartiles
Please find supplementary file(s);
http://dx.doi.org/10.1253/circj.CJ-17-1078
References
- 1.
Nichols M, Townsend N, Scarborough P, Rayner M. Cardiovascular disease in Europe 2014: Epidemiological update. Eur Heart J 2014; 35: 2950–2959.
- 2.
Yeh RW, Sidney S, Chandra M, Sorel M, Selby JV, Go AS. Population trends in the incidence and outcomes of acute myocardial infarction. N Engl J Med 2010; 362: 2155–2165.
- 3.
O’Gara PT, Kushner FG, Ascheim DD, Casey DE Jr, Chung MK, de Lemos JA, et al. 2013 ACCF/AHA guideline for the management of ST-elevation myocardial infarction: A report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation 2013; 127: e362–e425.
- 4.
Xu H, Li W, Yang J, Wiviott SD, Sabatine MS, Peterson ED, et al. The China Acute Myocardial Infarction (CAMI) Registry: A national long-term registry-research-education integrated platform for exploring acute myocardial infarction in China. Am Heart J 2016; 175: 193–201.e193.
- 5.
Thygesen K, Alpert JS, Jaffe AS, Simoons ML, Chaitman BR, White HD, et al. Third universal definition of myocardial infarction. J Am Coll Cardiol 2012; 60: 1581–1598.
- 6.
Sullivan LM, Massaro JM, D’Agostino RB Sr. Presentation of multivariate data for clinical use: The Framingham Study risk score functions. Stat Med 2004; 23: 1631–1660.
- 7.
Cook NR. Use and misuse of the receiver operating characteristic curve in risk prediction. Circulation 2007; 115: 928–935.
- 8.
Hosmer DW, Hosmer T, Le Cessie S, Lemeshow S. A comparison of goodness-of-fit tests for the logistic regression model. Stat Med 1997; 16: 965–980.
- 9.
Steyerberg EW, Vickers AJ, Cook NR, Gerds T, Gonen M, Obuchowski N, et al. Assessing the performance of prediction models: A framework for traditional and novel measures. Epidemiology 2010; 21: 128–138.
- 10.
Morrow DA, Antman EM, Charlesworth A, Cairns R, Murphy SA, de Lemos JA, et al. TIMI risk score for ST-elevation myocardial infarction: A convenient, bedside, clinical score for risk assessment at presentation: An intravenous nPA for treatment of infarcting myocardium early II trial substudy. Circulation 2000; 102: 2031–2037.
- 11.
GRACE Investigators. Rationale and design of the GRACE (Global Registry of Acute Coronary Events) Project: A multinational registry of patients hospitalized with acute coronary syndromes. Am Heart J 2001; 141: 190–199.
- 12.
Seropian IM, Sonnino C, Van Tassell BW, Biasucci LM, Abbate A. Inflammatory markers in ST-elevation acute myocardial infarction. Eur Heart J Acute Cardiovasc Care 2016; 5: 382–395.
- 13.
Barron HV, Cannon CP, Murphy SA, Braunwald E, Gibson CM. Association between white blood cell count, epicardial blood flow, myocardial perfusion, and clinical outcomes in the setting of acute myocardial infarction: A thrombolysis in myocardial infarction 10 substudy. Circulation 2000; 102: 2329–2334.
- 14.
Palmerini T, Mehran R, Dangas G, Nikolsky E, Witzenbichler B, Guagliumi G, et al. Impact of leukocyte count on mortality and bleeding in patients with myocardial infarction undergoing primary percutaneous coronary interventions: Analysis from the Harmonizing Outcome with Revascularization and Stent in Acute Myocardial Infarction trial. Circulation 2011; 123: 2829–2837, 2827.
- 15.
Palmerini T, Brener SJ, Genereux P, Maehara A, Della Riva D, Mariani A, et al. Relation between white blood cell count and final infarct size in patients with ST-segment elevation acute myocardial infarction undergoing primary percutaneous coronary intervention (from the INFUSE AMI trial). Am J Cardiol 2013; 112: 1860–1866.
- 16.
Smit JJ, Ottervanger JP, Slingerland RJ, Kolkman JJ, Suryapranata H, Hoorntje JC, et al. Comparison of usefulness of C-reactive protein versus white blood cell count to predict outcome after primary percutaneous coronary intervention for ST elevation myocardial infarction. Am J Cardiol 2008; 101: 446–451.
- 17.
Pellizzon GG, Dixon SR, Stone GW, Cox DA, Mattos L, Boura JA, et al. Relation of admission white blood cell count to long-term outcomes after primary coronary angioplasty for acute myocardial infarction (The Stent PAMI Trial). Am J Cardiol 2003; 91: 729–731.
- 18.
Swirski FK, Nahrendorf M. Leukocyte behavior in atherosclerosis, myocardial infarction, and heart failure. Science 2013; 339: 161–166.
- 19.
Niedziela J, Hudzik B, Niedziela N, Gasior M, Gierlotka M, Wasilewski J, et al. The obesity paradox in acute coronary syndrome: A meta-analysis. Eur J Epidemiol 2014; 29: 801–812.
- 20.
Madala MC, Franklin BA, Chen AY, Berman AD, Roe MT, Peterson ED, et al. Obesity and age of first non-ST-segment elevation myocardial infarction. J Am Coll Cardiol 2008; 52: 979–985.
- 21.
Diercks DB, Roe MT, Mulgund J, Pollack Jr CV, Kirk JD, Gibler WB, et al. The obesity paradox in non-ST-segment elevation acute coronary syndromes: Results from the Can Rapid risk stratification of Unstable angina patients Suppress ADverse outcomes with Early implementation of the American College of Cardiology/American Heart Association Guidelines Quality Improvement Initiative. Am Heart J 2006; 152: 140–148.
- 22.
Mahaffey KW, Tonev ST, Spinler SA, Levine GN, Gallo R, Ducas J, et al. Obesity in patients with non-ST-segment elevation acute coronary syndromes: Results from the SYNERGY trial. Int J Cardiol 2010; 139: 123–133.
- 23.
Das SR, Alexander KP, Chen AY, Powell-Wiley TM, Diercks DB, Peterson ED, et al. Impact of body weight and extreme obesity on the presentation, treatment, and in-hospital outcomes of 50,149 patients with ST-segment elevation myocardial infarction results from the NCDR (National Cardiovascular Data Registry). J Am Coll Cardiol 2011; 58: 2642–2650.
- 24.
Cepeda-Valery B, Chaudhry K, Slipczuk L, Pressman GS, Figueredo VM, Lavie CJ, et al. Association between obesity and severity of coronary artery disease at the time of acute myocardial infarction: Another piece of the puzzle in the “obesity paradox”. Int J Cardiol 2014; 176: 247–249.
- 25.
Bucholz EM, Beckman AL, Krumholz HA, Krumholz HM. Excess weight and life expectancy after acute myocardial infarction: The obesity paradox reexamined. Am Heart J 2016; 172: 173–181.
- 26.
Aune E, Roislien J, Mathisen M, Thelle DS, Otterstad JE. The “smoker’s paradox” in patients with acute coronary syndrome: A systematic review. BMC Med 2011; 9: 97.
- 27.
Reed GW, Cannon CP, Waalen J, Teirstein PS, Tanguay JF, Berger PB, et al. Influence of smoking on the antiplatelet effect of clopidogrel differs according to clopidogrel dose: Insights from the GRAVITAS trial. Catheter Cardiovasc Interv 2017; 89: 190–198.
- 28.
Zhang M, Liu X, Wang L, Wang Y, Ju L, Li J. Cigarette smoking might weaken the prognostic significance of cytochrome P450 2C19*2 polymorphism in acute myocardial infarction patients. J Cell Mol Med 2016; 20: 1247–1254.
- 29.
Symons R, Masci PG, Francone M, Claus P, Barison A, Carbone I, et al. Impact of active smoking on myocardial infarction severity in reperfused ST-segment elevation myocardial infarction patients: The smoker’s paradox revisited. Eur Heart J 2016; 37: 2756–2764.
- 30.
Smith SC Jr, Benjamin EJ, Bonow RO, Braun LT, Creager MA, Franklin BA, et al. AHA/ACCF Secondary Prevention and Risk Reduction Therapy for Patients with Coronary and other Atherosclerotic Vascular Disease: 2011 update: A guideline from the American Heart Association and American College of Cardiology Foundation. Circulation 2011; 124: 2458–2473.
- 31.
National Clinical Guideline Centre (UK). Unstable angina and NSTEMI: The early management of unstable angina and non-ST-segment-elevation myocardial infarction. London: Royal College of Physicians, 2010.
