Abstract
Aims:
The purpose of this study was to investigate the relationship between heart rate at admission and in-hospital mortality in patients with ST-segment elevation myocardial infarction (STEMI) and non-ST-segment elevation acute coronary syndrome (NSTE-ACS).
Methods:
Consecutive ACS patients admitted in 2008–2010 across 58 hospitals in six participant countries of the European Hospital Benchmarking by Outcomes in ACS Processes (EURHOBOP) project (Finland, France, Germany, Greece, Portugal and Spain). Cardiogenic shock patients were excluded. Associations between heart rate at admission in categories of 10 beats per min (bpm) and in-hospital mortality were estimated by logistic regression in crude models and adjusting for age, sex, obesity, smoking, hypertension, diabetes, known heart failure, renal failure, previous stroke and ischaemic heart disease. In total 10,374 patients were included.
Results:
In both STEMI and NSTE-ACS patients, a U-shaped relationship between admission heart rate and in-hospital mortality was found. The lowest risk was observed for heart rates between 70–79 bpm in STEMI and 60–69 bpm in NSTE-ACS; risk of mortality progressively increased with lower or higher heart rates. In multivariable models, the relationship persisted but was significant only for heart rates >80 bpm. A similar relationship was present in both patients with or without diabetes, above or below age 75 years, and irrespective of the presence of atrial fibrillation or use of beta-blockers.
Conclusion:
Heart rate at admission is significantly associated with in-hospital mortality in patients with both STEMI and NSTE-ACS. ACS patients with admission heart rate above 80 bpm are at highest risk of in-hospital mortality.
Keywords
Introduction
Heart rate is a basic vital parameter which is assessed at admission in every patient admitted with an acute coronary syndrome (ACS).
Heart rate at rest has been shown to be predictive of adverse outcomes in both general populations1–3 and patient populations4–6 across several prognostic studies, and heart rate at discharge has been shown to be a predictor of long-term outcomes in patients with coronary artery disease.7,8
In the acute setting, admission heart rate has been shown to be predictive of mortality across several previous studies; in the US Global Registry of Acute Coronary Events (GRACE) registry, 9 admission heart rate was shown to be one of the eight most important predictors of in-hospital mortality, and in the Can Rapid Risk Stratification of Unstable Angina Patients Suppress ADverse Outcomes with Early Implementation of the ACC/AHA Guidelines (CRUSADE) population of US patients with non-ST-segment elevation acute coronary syndrome (NSTE-ACS) 10 there was also an association between heart rate and prognosis.
Nonetheless, data on the association between admission heart rate and outcome in a contemporary population of patients with both NSTE-ACS and ST-segment elevation myocardial infarction (STEMI) admitted to European hospitals is scarce.
In the present study we therefore examined the relationship between heart rate at admission and in-hospital mortality in patients with ACS admitted across the 58 hospitals from six participant countries of the European Hospital Benchmarking by Outcomes in ACS Processes (EURHOBOP) collaboration.
Methods
The EURHOBOP project is a collaborative, multicentre and multinational retrospective study of patients hospitalised with a final diagnosis of ACS, consecutively discharged from 70 hospitals in seven Western European countries (Finland, France, Germany, Greece, Italy, Portugal and Spain) between 2008–2010. The study has previously been described. 11 In brief, in each country, 8–10 centres selected to represent the distribution of university, regional and private hospitals within the country’s healthcare system were included. Each centre contributed to the study with at least 200 consecutive patients. The inclusion criteria were a discharge diagnosis of myocardial infarction, with or without ST-segment elevation, or unstable angina (International Classification of Diseases, 10th revision: I21.0–I21.9 and I20.0). The current analysis only considers data from six European countries (Finland, France, Germany, Greece, Portugal and Spain) including 58 hospitals. Italian patients were excluded due to the lack of information on heart rate at admission.
Patient and hospital data were collected by trained medical record extractors using a standardised data collection form. To ensure quality of data collection all investigators underwent specific training. The main source of information was the discharge letter, however information on emergency room records and laboratory information systems was also accessed, whenever available. For each case, the investigators collected demographic, clinical, biological and electrocardiographic data from the medical records. The study protocol, including a detailed description of variable collection and definitions, is available at http://www.eurhobop.eu/files/EURHOBOP%20Data%20Extraction%20Procedures%20FINAL.pdf.
Data analysis
The type of ACS was defined according to electrocardiogram (ECG) findings at admission in agreement with current European guidelines.12,13 Patients who were not classifiable into type of ACS or with missing data on that variable were excluded from the analysis. The outcome of interest was defined as death of any cause during hospitalisation. Heart rate at admission was extracted from the files as a continuous variable and categorised into 10-beat wide intervals. Patients with cardiogenic shock were excluded. 10
To compare socio-demographic characteristics, invasive procedures during hospitalisation and outcomes with respect to different heart rate categories, Kruskal-Wallis tests and Pearson chi-square tests were used for continuous variables and categorical variables, respectively.
Unconditional logistic regression was used to estimate the odds ratios (ORs) for the association between in-hospital mortality and heart rate. In addition to the crude OR, we computed OR adjusting for age, sex and prior clinical history of obesity, smoking, diabetes, coronary heart disease, known heart failure, stroke and renal failure. Please consult the Supplementary Material data for detailed description of statistics. A p-value<0.05 was considered statistically significant.
Ethics
The study was approved by the Internal Review Boards of the institutions enrolled.
Results
Sample characteristics
A total of 10,374 ACS patients without cardiogenic shock at admission and with information about heart rate at admission were included, 3209 (31%) STEMI patients, 6636 (64%) NSTE-ACS patients, and 529 with non-classifiable ACS or with missing data on the type of ACS (Figure 1). Overall, the median length of hospitalisation was six days (interquartile range (IQR) 3–8): six days (IQR 4–8) in STEMI patients and five days (IQR 3–8) in NSTE-ACS. A total of 199 (6.2%) of the STEMI patients and 175 (2.6%) of the NSTE-ACS patients died during the index hospitalisation.
Flowchart illustrating the sample selection for the present analysis, by type of acute coronary syndrome (ACS). NSTE-ACS: non-ST-segment elevation acute coronary syndrome; STEMI: ST-segment elevation myocardial infarction.
In the STEMI population, patients with the lowest and highest heart rates were slightly older and with a higher proportion of women compared to patients in the intermediate heart rate categories. The same pattern was observed for diabetes, known heart failure, previous stroke and renal failure. There was no association between admission heart rate and obesity, hypertension, or previous Coronary artery disease (CAD) in STEMI patients. Details for admission data, procedures, and outcomes in STEMI patients are available in Table 1.
Socio-demographic, prior clinical history, admission data, invasive procedures and outcomes during hospitalisation in the ST-segment elevation myocardial infarction (STEMI) patients of the European Hospital Benchmarking by Outcomes in ACS Processes (EURHOBOP) cohort by admission heart rate.
ACS: acute coronary syndrome; bpm: beats per min.
Mean±standard variation.
Previous history of myocardial infarction, percutaneous coronary intervention or coronary artery bypass grafting.
Median (interquartile range).
In the NSTE-ACS population, patients in the lower or higher heart rate categories tended to be older; there were relatively more women in the higher heart rate categories. In terms of comorbidities NSTE-ACS patients with CAD tended to have lower heart rates, whereas known heart failure was more frequent in patients with higher admission heart rates. Stroke and renal failure were more prevalent in the higher heart rate categories. There were no differences between admission heart rate and obesity, hypertension, and smoking. Details of admission data, procedures and outcomes in NSTE-ACS patients are available in Table 2.
Socio-demographic, prior clinical history, admission data, invasive procedures and outcomes during hospitalisation in the non-ST-segment elevation acute coronary syndrome (NSTE-ACS) patients of the European Hospital Benchmarking by Outcomes in ACS Processes (EURHOBOP) cohort by admission heart rate.
ACS: acute coronary syndrome; bpm: beats per min.
Mean±standard variation.
Previous history of myocardial infarction, percutaneous coronary intervention or coronary artery bypass grafting.
Median (interquartile range).
Admission heart rate and in-hospital mortality
STEMI
The association between admission heart rate and mortality in STEMI patients is displayed in Figure 2. As shown, there was a U-shaped relationship between admission heart rate and in-hospital mortality; in STEMI patients, heart rates between 70–79 bpm were associated with the lowest risk. Heart rate <40 bpm was associated with a 4.3-fold increased risk and admission heart rate categories >80 bpm with a 2.2 to 5.3-fold increased risk of mortality. In the adjusted model the U-shape persisted, although only heart rates >80 bpm remained significantly associated with increased risk compared to the reference group.
Association between admission heart rate and in-hospital mortality across 58 European Hospitals in patients with ST-segment elevation myocardial infarction (STEMI) and non-ST-segment elevation acute coronary syndrome (NSTE-ACS). *Adjusted for age, sex and prior clinical history of obesity, smoking, hypertension, diabetes, coronary heart disease, heart failure, stroke and renal failure. CI: confidence interval; OR: odds ratio.
NSTE-ACS
For NSTE-ACS, an admission heart rate of 60–69 was associated with the lowest risk (Figure 2); a heart rate <40 bpm was associated with an 11-fold increased risk and heart rate categories >80 bpm with a 3.8 to 8.8-fold increased risk of mortality. As with the STEMI population, the U-shaped relationship persisted but was attenuated for the low heart rates in the multivariable models. Again, heart rates >80 bpm were significantly associated with mortality after multivariable adjustments.
Admission heart rate, in-hospital mortality, by sex, in the elderly and in patients with diabetes
Additional analyses were performed and are shown in the data in the Supplementary Material, Figures 1–4. When stratifying into men and women, the U-shaped relationship was observed in men for both STEMI and NSTE-ACS, and for women for NSTE-ACS, but not in women with STEMI; here, a positive association with admission heart rate was observed (Supplementary Material, Figure 1) with a low heart rate being associated with the lowest risk, however, there were no patients in the <40 bpm heart rate category.
To examine if the relationship was observed also in the elderly patients, the population was stratified into two groups, above or below 75 years (Supplementary Material, Figure 2); as shown, the U-shaped relationship between admission heart rate and in-hospital mortality was observed in the elderly and younger patients alike.
Also, in both patients with diabetes and without diabetes the U-shape persisted, see the results in the Supplementary Material, Figure 3.
Admission heart rate, in-hospital mortality, atrial fibrillation and use of beta-blockers
Information about presence of atrial fibrillation during hospitalisation was available only in patients admitted in Portugal encompassing a total of 3009 patients (1021 STEMI, 1765 NSTE-ACS, 223 with non-classifiable or missing type of ACS). The U-shape observed across all countries was also observed in the Portuguese data alone. The effect of atrial fibrillation was examined by excluding patients with atrial fibrillation (n=280), resulting in no change in the overall association between admission heart rate and in-hospital mortality; if anything, the association was stronger. Further adjustment for the use of beta-blockers (n=1713 (57%)) was included, however, no change in the association between heart rate and mortality was observed. The analyses are shown in the data in the Supplementary Material, Figure 4.
Discussion
In the present study of 10,000 patients admitted with an ACS across 58 Western European hospitals we found that heart rates at admission in patients with both STEMI and NSTE-ACS are highly associated with in-hospital mortality.
The association between admission heart rate and in-hospital mortality is U-shaped with the lowest and highest heart rates conferring the greatest risk. In NSTE-ACS, the risk of death in patients with bradycardia is increased by up to 11-fold and heart rates >80 bpm are associated with up to almost nine-fold increased risk of mortality compared to the reference heart rate category of 60–69 bpm. In STEMI patients, the risk of mortality is increased up to five-fold in patients with admission heart rates below or above the reference heart rate category of 70–79 bpm. When adjusting for confounding factors, a heart rate above 80 bpm remains associated with increased risk of in-hospital mortality in both STEMI and NSTE-ACS patients whereas bradycardia is no longer significantly associated with risk of mortality. Altogether, these findings suggest that ACS patients with high heart rates at admission are at increased risk of in-hospital mortality.
Heart rate is one of the first and most readily available physiological parameters in ACS patients and is obtainable within the first few minutes of a patient being admitted to the emergency room. In the acute setting, information about heart rate is therefore available much earlier than any other clinical and paraclinical information, such as troponin, and often even before other ECG changes occur. Furthermore, heart rate is very easy to obtain and there are no difficulties in interpretation or analysis. Heart rate at admission is therefore a fundamental and obvious biomarker to assess in the clinical setting, and to study prognostically.
Heart rate in both the acute and non-acute setting has been studied previously. Heart rate at rest has been shown to be predictive of ischaemic heart disease and mortality in several prospective studies of general populations.1–3,13,14 In patients from the Coronary Artery Surgery Study (CASS) registry 16 with suspected or proven coronary artery disease, elevated resting heart rate was found to be predictive of adverse events during 14 years of follow-up. In a more recent study of 3300 patients undergoing coronary angiography, 17 heart rate at rest was associated with both total mortality and cardiovascular outcome, and in the Effects of Ivabradine in Patients With Stable Coronary Artery Disease and Left Ventricular Systolic Dysfunction (BEAUTIFUL) trial a heart rate at rest >70 bpm was associated with increased risk. 6 Furthermore, discharge heart rate of patients admitted with ACS has been studied in recent populations following percutaneous coronary intervention for STEMI 7 and for both NSTE-ACS and STEMI, 8 and in both populations an increased discharge heart rate was associated with long-term mortality.
Admission heart rate in ACS has been studied in previous populations in both pre-thrombolytic and thrombolytic eras; Madsen et al. 18 and Hjalmarson et al. 19 found high heart rate to be associated with increased risk, and Designi et al. 20 showed similar findings, however treatment of ACS has changed dramatically since these studies were conducted. Admission heart rate has also been studied in a relatively more recent cohort; in the CRUSADE cohort of NSTE-ACS patients only included in the USA from 2001–2005, a U-shaped relationship was found between admission heart rate and in-hospital mortality. 10 In the present study of a contemporary population of both STEMI and NSTE-ACS patients included across 58 European hospitals, we also found a U-shaped relationship in both NSTE-ACS and STEMI patients. In the multivariable analyses, the overall relationship persisted but was significant only for admission heart rates above 80 bpm. Furthermore, we found the relationship to be present in both the elderly population and the diabetes population, as well as in both men and women. The relationship between admission heart rate and mortality therefore seems to be a robust and clinically relevant marker of prognosis, also in a contemporary population of patients treated at European hospitals.
In light of the present findings, we suggest that a heart rate >80 bpm in patients admitted with ACS, irrespective of type of ACS, should be considered a marker of adverse events during hospitalisation. In the GRACE score model, 9 increasing admission heart rates from a baseline of ⩽50 bpm is considered to be associated with increased risk; in our present study, however, the crude results suggest that an admission heart rate ⩽50 bpm should not be considered to be a low-risk heart rate, and secondly, the cut-off value for a high-risk heart rate should be considered to be 80 bpm and above.
In the present cohort of patients with ACS, heart rate may possibly be both a marker of risk and potentially also related to the pathophysiological mechanism. Patients with ACS admitted with a very low heart rate may have bradycardia due to conduction defects such as high-grade atrioventricular block although we have no information about this; we found that both patients presenting with NSTE-ACS or STEMI and heart rate <40 bpm were at high risk of in-hospital mortality (OR 11.0 and 4.3 – relative to their respective reference heart rate category), 21 however, the data are difficult to interpret due to the low number of patients and events in the lowest heart rate group. Also, the relationship with in-hospital mortality did not remain significant after multivariable adjustments, indicating that the patients in the <40 bpm group were more comorbid compared to the others heart rate groups which is supported by their lower blood pressure, haemoglobin and worse kidney function at admission. These patients may therefore have been in a state of ‘pre-cardiogenic shock’ explaining the relationship seen in the crude model. Conventionally, tachycardia is defined as >100 bpm, although this has been debated,22,23 and the present study supports a lower cut-off in patients admitted with ACS. We found admission heart rates >80 bpm to be associated with a highly increased risk of in-hospital mortality, and these findings were robust to multivariable adjustments. Furthermore, we show that this relationship is similar when stratified according to, age, diabetes and presence of atrial fibrillation or use of beta-blockers. Here, it is important to note that cardiogenic shock patients were excluded, so other mechanisms must be in play. These may include decreased diastolic filling time, increased myocardial oxygen consumption, and increased susceptibility to development of malignant arrhythmias with increase in heart rate.24,25
Potential study limitations should be considered. Firstly, although all 58 participating centres adhered to European guidelines, the available means for therapeutic intervention were not uniform across participating countries, and in-hospital mortality therefore varied between countries. This is likely explained by regional differences in the organisation of healthcare systems and geographical distances to invasive centres. This has been discussed in detail elsewhere. 11 However, inclusion of country or participant centre in the multivariable analyses did not change the association. Finally, only in-hospital events are available in the EURHOBOP cohort, and long-term follow-up could therefore not be assessed.
In conclusion, we examined the relationship between admission heart rate and in-hospital mortality in 10,374 consecutive ACS patients admitted from 2008–2010 across 58 European hospitals in six participant countries of the EURHOBOP collaboration. There was a U-shaped relationship between mortality and admission heart rate in both NSTE-ACS and STEMI patients in that heart rates above 80 bpm and below 40 bpm were associated with increased risk. In STEMI, patients with admission heart rate between 70–79 bpm were at lowest risk, and in NSTE-ACS patients in the 60–69 bpm range were at lowest risk. After multivariable adjustments the association with bradycardia was attenuated. An admission heart rate above 80 bpm was associated with a greatly increased risk of in-hospital mortality in both patients with STEMI and NSTE-ACS, and the relationship was present irrespective of beta-blocker use, presence of atrial fibrillation, in both men and women, in patients with and without diabetes, and in the younger and elderly patient alike.
Supplemental Material
Heart_rate_at_admission_SupplData25February16_EHJ – Supplemental material for Heart rate at admission is a predictor of in-hospital mortality in patients with acute coronary syndromes: Results from 58 European hospitals: The European Hospital Benchmarking by Outcomes in acute coronary syndrome Processes study
Supplemental material, Heart_rate_at_admission_SupplData25February16_EHJ for Heart rate at admission is a predictor of in-hospital mortality in patients with acute coronary syndromes: Results from 58 European hospitals: The European Hospital Benchmarking by Outcomes in acute coronary syndrome Processes study by Magnus T Jensen, Marta Pereira, Carla Araujo, Anti Malmivaara, Jean Ferrieres, Irene R Degano, Inge Kirchberger, Dimitrios Farmakis, Pascal Garel, Marina Torre, Jaume Marrugat and Ana Azevedo in European Heart Journal: Acute Cardiovascular Care
Footnotes
Acknowledgements
The authors gratefully acknowledge the collaboration of the hospitals and local researchers who participated in the EURHOBOP study, and Mascia Masciocchi who was responsible for editing EURHOBOP’s website.
Conflict of interest
D Farmakis reports personal fees from Servier, outside the submitted work; J Ferrières reports grants from Amgen and Merck, educational activities with Astra Zeneca and Sanofi, also outside the submitted work. All other authors have nothing to declare.
Funding
The EURHOBOP study was supported by the Executive Agency for Health and Consumers (2008 13 12 EURHOBOP).
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
