Canadian Journal of Cardiology
-
(2015) 1e7
Clinical Research
Cardiovascular Risk Factors and In-Hospital Mortality in Acute Coronary Syndromes: Insights From the Canadian Global Registry of Acute Coronary Events Jenny Y. Wang, MD,a Shaun G. Goodman, MD, MSc,a,b Ilana Saltzman, MD,c Graham C. Wong, MD, MPH,d Thao Huynh, MD, MSc,e Jean-Pierre Dery, MD, MSc,f Lawrence A. Leiter, MD,g Deepak L. Bhatt, MD, MPH,h Robert C. Welsh, MD,i Frederick A. Spencer, MD,j Keith A.A. Fox, MB, ChB,k and Andrew T. Yan, MD;a for the Global Registry of Acute Coronary Events (GRACE/GRACE-2) and Canadian Registry of Acute Coronary Events (CANRACE) Investigators a
Division of Cardiology, St. Michael’s Hospital, University of Toronto, Toronto, Ontario, Canada b c
d
g
Vancouver General Hospital, University of British Columbia, Vancouver, British Columbia, Canada e
McGill University Health Centre, McGill University, Montreal, Quebec, Canada
f
Que bec Heart and Lung Institute, Laval Hospital, Que bec City, Quebec, Canada
Division of Endocrinology and Metabolism, St. Michael’s Hospital, University of Toronto, Toronto, Ontario, Canada h
i
Canadian Heart Research Centre, Toronto, Ontario, Canada
Toronto General Hospital, University of Toronto, Toronto, Ontario, Canada
Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA
Mazankowski Alberta Heart Institute, University of Alberta Hospital, Canadian VIGOUR Centre, Edmonton, Alberta, Canada j
Thrombosis and Atherosclerosis Research Institute, McMaster University, Hamilton, Ontario, Canada k
Centre for Cardiovascular Science, University of Edinburgh, Edinburgh, United Kingdom
ABSTRACT
RESUM E
Background: There are conflicting data regarding the relationship between the number of modifiable traditional risk factors and prognosis in acute coronary syndromes (ACS). This controversy might in part be explained by the differential use of prehospital medications. Methods: Using data from the Canadian, multicentre Global Registry of Acute Coronary Events (GRACE) (1999-2008), we stratified 13,686 ACS patients into 3 groups (0, 1-2, vs 3-4 risk factors) and compared
es sur le lien entre le nombre de facteurs de Introduction : Les donne risque traditionnels modifiables et le pronostic des syndromes coronariens aigus (SCA) sont contradictoires. Cette controverse pourrait en rentielle des me dicaments partie s’expliquer par l’utilisation diffe hospitalisation. pre thodes : À l’aide des donne es du registre multicentrique canadien Me GRACE (Global Registry of Acute Coronary Events; 1999-2008), nous
Prognostication is an important component of acute coronary syndrome (ACS) management. In the 2014 ACC/AHA guidelines for noneST-elevation ACS, the use of validated
risk scores to assess prognosis received a Class I recommendation.1 Therefore, optimizing the prognostic accuracy of existing risk scores by incorporating other clinical information that is readily available at presentation is of great clinical value. Although the importance of modifiable traditional cardiovascular risk factors (hypertension, diabetes, dyslipidemia, and smoking) in the pathogenesis of coronary artery disease (CAD) is well recognized,2,3 there are conflicting data regarding their prognostic value for short-term outcomes in patients with ACS. The Thrombolysis In Myocardial
Received for publication February 23, 2015. Accepted April 10, 2015. Corresponding author: Dr Andrew Yan or Dr Shaun Goodman, St. Michael’s Hospital, 30 Bond St, Donnelly Room 6-030, Toronto, Ontario M5B 1W8, Canada. Tel.: þ1-416-864-5465; fax: þ1-416-864-5159. E-mail: [email protected] or [email protected] See page 6 for disclosure information.
http://dx.doi.org/10.1016/j.cjca.2015.04.007 0828-282X/Ó 2015 Canadian Cardiovascular Society. Published by Elsevier Inc. All rights reserved.
2
Canadian Journal of Cardiology Volume - 2015
their baseline characteristics, in-hospital treatments, and outcomes. Multivariable logistic regressions were performed to adjust for the components of the GRACE risk score and preadmission statin and acetylsalicylic acid (ASA) use. Results: Among these patients (ST-elevation myocardial infarction 28.3%), 14.5%, 62.6%, and 22.9% had 0, 1-2, and 3-4 risk factors, respectively. Patients with fewer risk factors were less likely to be on ASA, statin, and other prehospital medications. Unadjusted in-hospital mortality was significantly different across risk factor groups (4.9%, 3.0%, and 3.1% for 0, 1-2, and 3-4 risk factor groups, respectively, P for trend ¼ 0.002). This difference was no longer significant after adjusting for the components of the GRACE risk score (P for trend ¼ 0.088) and further adjusting for preadmission statin and ASA use (P for trend ¼ 0.96). For in-hospital mortality, there was no significant interaction between risk factor categories and ACS type (P ¼ 0.26). Conclusions: The lower mortality observed in patients with ACS with more risk factors may be partially attributed to the protective effect of prehospital ASA and statin use. The number of risk factors does not provide incremental prognostic value beyond the validated GRACE risk score.
13 686 patients atteints d’un SCA en 3 groupes (0, 1-2 avons stratifie leurs caracte ristiques initiales, vs 3-4 facteurs de risque) et compare sultats cliniques. Nous avons les traitements intrahospitaliers et les re alise les re gressions logistiques multivarie es pour les ajuster aux re composantes du score de risque GRACE et à l’utilisation d’une statine tylsalicylique (AAS) avant l’admission. et de l’acide ace sultats : Parmi ces patients (28,3 % d’infarctus du myocarde avec Re calage du segment ST), 14,5 %, 62,6 % et 22,9 % avaient sus-de respectivement 0, 1-2 et 3-4 facteurs de risque. Les patients ayant taient moins susceptibles de prendre de moins de facteurs de risque e dicaments pre hospitalisation. l’AAS, une statine ou d’autres me intrahospitalière non ajuste e e tait significativement La mortalite rente entre les groupes de facteurs de risque (4,9 %, 3,0 % et diffe 3,1 % pour 0, 1-2 et 3-4 groupes de facteurs de risque, respectiverence n’e tait plus signifiment, P de tendance ¼ 0,002). Cette diffe cative après l’ajustement des composantes du score de risque GRACE (P de tendance ¼ 0,088) et les autres ajustements de l’utilisation d’une statine et d’AAS avant l’admission (P de tendance ¼ 0,96). Pour intrahospitalière, il n’existait aucune interce qui est de la mortalite gories de facteurs de risque et le type action significative entre les cate de SCA (P ¼ 0,26). observe e chez les patients Conclusions : La plus faible mortalite atteints d’un SCA qui ont plus de facteurs de risque peut être e à l’effet protecteur de l’utilisation de l’AAS et de parellement attribue hospitalisation. Le nombre de facteurs de risque la tistatine en pre mentaire au-delà de la valin’offre pas de valeur pronostique supple dation du score de risque GRACE.
Infarction risk score includes 3 risk factors as an adverse prognosticator in ACS.4 In contrast, registry data suggest an inverse relationship between the number of risk factors and mortality.5,6 This discrepancy may be partially accounted for by the confounding effect of preadmission medication use, such as acetylsalicylic acid (ASA) or statin, which has been associated with improved ACS outcomes.7 The objective of this study was to characterize the relationship between the number of modifiable traditional cardiovascular risk factors and in-hospital ACS management and outcomes, and potentially improve the prognostic accuracy of existing risk scores. Specifically, we examined the relationship between the number of risk factors and in-hospital mortality, after adjusting for the components of the Global Registry of Acute Coronary Events (GRACE) risk score8-10 and preadmission medication use.
GRACE/GRACE2 enrolled patients from 1999 to 2007, and was continued in Canada as CANRACE in 2008. Patient demographics, clinical presentation, investigations, treatments, and outcomes were collected by trained personnel using standardized case report forms. Our study included data from Canadian sites only. Similar to prior studies,5,6 we focused on 4 traditional, potentially modifiable risk factors (smoking, diabetes, hypertension, and dyslipidemia); patients with missing data for any of these selfreported and/or chart-abstracted risk factors were excluded. Smoking refers to current smoking. Diabetes mellitus refers to a previous diagnosis of diabetes treated with diet, oral antihyperglycemic agents and/or insulin before hospitalization. Hypertension refers to previously diagnosed hypertension, treated or untreated. Dyslipidemia refers to previously diagnosed dyslipidemia or preadmission use of lipid-lowering medication. A total of 13,686 patients were included and classified into one of 3 groups based on the number of documented modifiable risk factors: 0, 1 to 2, or 3 to 4 risk factors. These 3 groups were selected to avoid group size and permitted evaluation of trends based on the number of risk factors.
Methods Study design and population The Global Registry of Acute Coronary Events (GRACE/ Expanded GRACE [GRACE2]) and Canadian Registry of Acute Coronary Events (CANRACE) were multicentre prospective registries of patients presenting with ACS.11-13 Briefly, inclusion criteria for these registries included age 18 years old, being alive at hospital presentation, and presumed ACS diagnosis. Patients who were already hospitalized at the time of development of ACS and those with a significant acute comorbidity such as trauma were excluded. Patients were categorized as having ST-elevation myocardial infarction (STEMI), noneSTEMI (NSTEMI), or unstable angina.11
Statistical analysis Categorical variables were reported as percentages and continuous variables as medians with interquartile ranges. Pearson-c2 and Kruskal-Wallis tests were used for categorical and continuous variables, respectively. Multivariable logistic regression was performed to estimate the adjusted odds ratio (OR) and 95% confidence interval (95% CI) for in-hospital mortality for each risk factor group. The candidate predictor variables were the components of the GRACE
Wang et al. CV Risk Factors and ACS In-hospital Mortality
3
Table 1. Baseline demographics
Age, years* Male CAD risk factors Diabetes Hypertension Dyslipidemia Current smoker Body mass index* Previous myocardial infarction Previous angina Prior CABG Prior PCI Prior heart failure Previous transient ischemic attack/stroke Peripheral artery disease Atrial fibrillation History of bleeding Prehospitalization medications Statin ASA Clopidogrel b-Blocker ACE inhibitor ARB Calcium channel blocker
No risk factors (n ¼ 1984)
1-2 risk factors (n ¼ 8574)
3-4 risk factors (n ¼ 3128)
P value for trend
66 (56, 78) 70.4%
67 (56, 77) 66.7%
66 (57, 75) 65.4%
0.006
-
(2015) 1e7
Clinical Research
Cardiovascular Risk Factors and In-Hospital Mortality in Acute Coronary Syndromes: Insights From the Canadian Global Registry of Acute Coronary Events Jenny Y. Wang, MD,a Shaun G. Goodman, MD, MSc,a,b Ilana Saltzman, MD,c Graham C. Wong, MD, MPH,d Thao Huynh, MD, MSc,e Jean-Pierre Dery, MD, MSc,f Lawrence A. Leiter, MD,g Deepak L. Bhatt, MD, MPH,h Robert C. Welsh, MD,i Frederick A. Spencer, MD,j Keith A.A. Fox, MB, ChB,k and Andrew T. Yan, MD;a for the Global Registry of Acute Coronary Events (GRACE/GRACE-2) and Canadian Registry of Acute Coronary Events (CANRACE) Investigators a
Division of Cardiology, St. Michael’s Hospital, University of Toronto, Toronto, Ontario, Canada b c
d
g
Vancouver General Hospital, University of British Columbia, Vancouver, British Columbia, Canada e
McGill University Health Centre, McGill University, Montreal, Quebec, Canada
f
Que bec Heart and Lung Institute, Laval Hospital, Que bec City, Quebec, Canada
Division of Endocrinology and Metabolism, St. Michael’s Hospital, University of Toronto, Toronto, Ontario, Canada h
i
Canadian Heart Research Centre, Toronto, Ontario, Canada
Toronto General Hospital, University of Toronto, Toronto, Ontario, Canada
Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA
Mazankowski Alberta Heart Institute, University of Alberta Hospital, Canadian VIGOUR Centre, Edmonton, Alberta, Canada j
Thrombosis and Atherosclerosis Research Institute, McMaster University, Hamilton, Ontario, Canada k
Centre for Cardiovascular Science, University of Edinburgh, Edinburgh, United Kingdom
ABSTRACT
RESUM E
Background: There are conflicting data regarding the relationship between the number of modifiable traditional risk factors and prognosis in acute coronary syndromes (ACS). This controversy might in part be explained by the differential use of prehospital medications. Methods: Using data from the Canadian, multicentre Global Registry of Acute Coronary Events (GRACE) (1999-2008), we stratified 13,686 ACS patients into 3 groups (0, 1-2, vs 3-4 risk factors) and compared
es sur le lien entre le nombre de facteurs de Introduction : Les donne risque traditionnels modifiables et le pronostic des syndromes coronariens aigus (SCA) sont contradictoires. Cette controverse pourrait en rentielle des me dicaments partie s’expliquer par l’utilisation diffe hospitalisation. pre thodes : À l’aide des donne es du registre multicentrique canadien Me GRACE (Global Registry of Acute Coronary Events; 1999-2008), nous
Prognostication is an important component of acute coronary syndrome (ACS) management. In the 2014 ACC/AHA guidelines for noneST-elevation ACS, the use of validated
risk scores to assess prognosis received a Class I recommendation.1 Therefore, optimizing the prognostic accuracy of existing risk scores by incorporating other clinical information that is readily available at presentation is of great clinical value. Although the importance of modifiable traditional cardiovascular risk factors (hypertension, diabetes, dyslipidemia, and smoking) in the pathogenesis of coronary artery disease (CAD) is well recognized,2,3 there are conflicting data regarding their prognostic value for short-term outcomes in patients with ACS. The Thrombolysis In Myocardial
Received for publication February 23, 2015. Accepted April 10, 2015. Corresponding author: Dr Andrew Yan or Dr Shaun Goodman, St. Michael’s Hospital, 30 Bond St, Donnelly Room 6-030, Toronto, Ontario M5B 1W8, Canada. Tel.: þ1-416-864-5465; fax: þ1-416-864-5159. E-mail: [email protected] or [email protected] See page 6 for disclosure information.
http://dx.doi.org/10.1016/j.cjca.2015.04.007 0828-282X/Ó 2015 Canadian Cardiovascular Society. Published by Elsevier Inc. All rights reserved.
2
Canadian Journal of Cardiology Volume - 2015
their baseline characteristics, in-hospital treatments, and outcomes. Multivariable logistic regressions were performed to adjust for the components of the GRACE risk score and preadmission statin and acetylsalicylic acid (ASA) use. Results: Among these patients (ST-elevation myocardial infarction 28.3%), 14.5%, 62.6%, and 22.9% had 0, 1-2, and 3-4 risk factors, respectively. Patients with fewer risk factors were less likely to be on ASA, statin, and other prehospital medications. Unadjusted in-hospital mortality was significantly different across risk factor groups (4.9%, 3.0%, and 3.1% for 0, 1-2, and 3-4 risk factor groups, respectively, P for trend ¼ 0.002). This difference was no longer significant after adjusting for the components of the GRACE risk score (P for trend ¼ 0.088) and further adjusting for preadmission statin and ASA use (P for trend ¼ 0.96). For in-hospital mortality, there was no significant interaction between risk factor categories and ACS type (P ¼ 0.26). Conclusions: The lower mortality observed in patients with ACS with more risk factors may be partially attributed to the protective effect of prehospital ASA and statin use. The number of risk factors does not provide incremental prognostic value beyond the validated GRACE risk score.
13 686 patients atteints d’un SCA en 3 groupes (0, 1-2 avons stratifie leurs caracte ristiques initiales, vs 3-4 facteurs de risque) et compare sultats cliniques. Nous avons les traitements intrahospitaliers et les re alise les re gressions logistiques multivarie es pour les ajuster aux re composantes du score de risque GRACE et à l’utilisation d’une statine tylsalicylique (AAS) avant l’admission. et de l’acide ace sultats : Parmi ces patients (28,3 % d’infarctus du myocarde avec Re calage du segment ST), 14,5 %, 62,6 % et 22,9 % avaient sus-de respectivement 0, 1-2 et 3-4 facteurs de risque. Les patients ayant taient moins susceptibles de prendre de moins de facteurs de risque e dicaments pre hospitalisation. l’AAS, une statine ou d’autres me intrahospitalière non ajuste e e tait significativement La mortalite rente entre les groupes de facteurs de risque (4,9 %, 3,0 % et diffe 3,1 % pour 0, 1-2 et 3-4 groupes de facteurs de risque, respectiverence n’e tait plus signifiment, P de tendance ¼ 0,002). Cette diffe cative après l’ajustement des composantes du score de risque GRACE (P de tendance ¼ 0,088) et les autres ajustements de l’utilisation d’une statine et d’AAS avant l’admission (P de tendance ¼ 0,96). Pour intrahospitalière, il n’existait aucune interce qui est de la mortalite gories de facteurs de risque et le type action significative entre les cate de SCA (P ¼ 0,26). observe e chez les patients Conclusions : La plus faible mortalite atteints d’un SCA qui ont plus de facteurs de risque peut être e à l’effet protecteur de l’utilisation de l’AAS et de parellement attribue hospitalisation. Le nombre de facteurs de risque la tistatine en pre mentaire au-delà de la valin’offre pas de valeur pronostique supple dation du score de risque GRACE.
Infarction risk score includes 3 risk factors as an adverse prognosticator in ACS.4 In contrast, registry data suggest an inverse relationship between the number of risk factors and mortality.5,6 This discrepancy may be partially accounted for by the confounding effect of preadmission medication use, such as acetylsalicylic acid (ASA) or statin, which has been associated with improved ACS outcomes.7 The objective of this study was to characterize the relationship between the number of modifiable traditional cardiovascular risk factors and in-hospital ACS management and outcomes, and potentially improve the prognostic accuracy of existing risk scores. Specifically, we examined the relationship between the number of risk factors and in-hospital mortality, after adjusting for the components of the Global Registry of Acute Coronary Events (GRACE) risk score8-10 and preadmission medication use.
GRACE/GRACE2 enrolled patients from 1999 to 2007, and was continued in Canada as CANRACE in 2008. Patient demographics, clinical presentation, investigations, treatments, and outcomes were collected by trained personnel using standardized case report forms. Our study included data from Canadian sites only. Similar to prior studies,5,6 we focused on 4 traditional, potentially modifiable risk factors (smoking, diabetes, hypertension, and dyslipidemia); patients with missing data for any of these selfreported and/or chart-abstracted risk factors were excluded. Smoking refers to current smoking. Diabetes mellitus refers to a previous diagnosis of diabetes treated with diet, oral antihyperglycemic agents and/or insulin before hospitalization. Hypertension refers to previously diagnosed hypertension, treated or untreated. Dyslipidemia refers to previously diagnosed dyslipidemia or preadmission use of lipid-lowering medication. A total of 13,686 patients were included and classified into one of 3 groups based on the number of documented modifiable risk factors: 0, 1 to 2, or 3 to 4 risk factors. These 3 groups were selected to avoid group size and permitted evaluation of trends based on the number of risk factors.
Methods Study design and population The Global Registry of Acute Coronary Events (GRACE/ Expanded GRACE [GRACE2]) and Canadian Registry of Acute Coronary Events (CANRACE) were multicentre prospective registries of patients presenting with ACS.11-13 Briefly, inclusion criteria for these registries included age 18 years old, being alive at hospital presentation, and presumed ACS diagnosis. Patients who were already hospitalized at the time of development of ACS and those with a significant acute comorbidity such as trauma were excluded. Patients were categorized as having ST-elevation myocardial infarction (STEMI), noneSTEMI (NSTEMI), or unstable angina.11
Statistical analysis Categorical variables were reported as percentages and continuous variables as medians with interquartile ranges. Pearson-c2 and Kruskal-Wallis tests were used for categorical and continuous variables, respectively. Multivariable logistic regression was performed to estimate the adjusted odds ratio (OR) and 95% confidence interval (95% CI) for in-hospital mortality for each risk factor group. The candidate predictor variables were the components of the GRACE
Wang et al. CV Risk Factors and ACS In-hospital Mortality
3
Table 1. Baseline demographics
Age, years* Male CAD risk factors Diabetes Hypertension Dyslipidemia Current smoker Body mass index* Previous myocardial infarction Previous angina Prior CABG Prior PCI Prior heart failure Previous transient ischemic attack/stroke Peripheral artery disease Atrial fibrillation History of bleeding Prehospitalization medications Statin ASA Clopidogrel b-Blocker ACE inhibitor ARB Calcium channel blocker
No risk factors (n ¼ 1984)
1-2 risk factors (n ¼ 8574)
3-4 risk factors (n ¼ 3128)
P value for trend
66 (56, 78) 70.4%
67 (56, 77) 66.7%
66 (57, 75) 65.4%
0.006
