© 2006 The Author(s)
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abnormal haemoglobin levels in acute coronary syndromes
From the Division of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow, UK
Address correspondence to Dr C. Berry, Division of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow G11 6NT. Scotland, UK. email: colin.berry{at}clinmed.gla.ac.uk
Received 26 February 2006 and in revised form 13 July 2006
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionAppendix 1. Reference ranges...References |
|---|
Background: Anaemia is an adverse prognostic marker in acute coronary syndromes (ACS), but the epidemiology of abnormal haemoglobin levels in such patients is uncertain.
Aims: To investigate the prevalence, nature and predictors of abnormal haemoglobin levels in ACS patients at admission.
Design: Observational study.
Methods: All emergency admissions from January to April 2005 were assessed within 24–48 h of hospital admission. ACS patients (unstable angina, non-ST-elevation or ST-elevation myocardial infarction) were enrolled (n = 320, 190 men). Clinical information was recorded.
Results: Overall, 71% had unstable angina; 18% non-ST-elevation myocardial infarction (MI), and 11% ST-elevation MI. Mean ± SD haemoglobin was 14.3 ± 1.7 g/dl in men and 13.2 ± 1.5 g/dl in women. Abnormal haemoglobin was more common in men (65, 34%) than in women (34, 22%) (p = 0.013). Anaemia (haemoglobin <13 g/dl in men, or <12.0 g/dl in women) was recorded in 35 (18%) men and 24 (18%) women. All had admission haemoglobin >8 g/dl, and anaemia was usually normocytic. Multivariate predictors of anaemia (OR, 95%CI) were age (1.07, 1.04–1.1) and serum albumin (0.90, 0.81–1.00). Elevated haemoglobin (>16 g/dl) was recorded in 30 (16%) men and 4 (3%) women (p < 0.01), and was more common in ST-elevation MI patients (26%) than in unstable angina or non-ST elevation MI patients (9%) (p = 0.005). In patients who underwent invasive management with a post-procedure haemoglobin the following day (n = 85), 15 (18%) new cases of anaemia were detected. Admission duration correlated with haemoglobin (p < 0.01), creatinine (p < 0.01), troponin I (p < 0.01) and C-reactive protein (p < 0.01). Anaemia was more common in those who died in hospital (3, 60%) than in those who survived (56, 18%) (adjusted p = 0.0135).
Discussion: Abnormal haemoglobin levels were common in our ACS admissions. Anaemia was generally mild. Increasing age and interventional management were associated with anaemia, which in turn was associated with adverse in-hospital outcomes. Interventions to prevent and detect anaemia in this setting merit prospective testing.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionAppendix 1. Reference ranges...References |
|---|
Unstable angina, ST-elevation and non-ST-elevation myocardial infarction, collectively known as the acute coronary syndromes, are leading causes of morbidity and mortality in developed countries.1–3 Notably, acute coronary syndromes are often complicated by the presence of co-existing health problems,4 such as anaemia.3,5–8 The adverse prognostic importance of haemoglobin abnormalities in coronary heart disease (CHD) has been extensively described elsewhere.3,5–7,9–12 However, much of this information is based on retrospective analyses of selected cohorts.3,5,7,9–13
Acute coronary syndromes are increasingly managed in an aggressive fashion. However, whether contemporary management strategies promote other prognostically important morbidity, such as anaemia, is uncertain. For example, could the wider prescription of adjunctive anti-thrombotic therapy, such as clopidogrel, or the use of interventional management of acute coronary syndromes, increase the risk of anaemia? While clinically important bleeding events in acute coronary syndrome clinical trial participants have been described,14,15 the frequency and nature of mild–moderate reductions in haemoglobin are not reported. Thus, the epidemiology of mild–moderate anaemia in acute coronary syndromes in contemporary practice is uncertain.
Although the cardiovascular consequences of anaemia are reasonably well understood, other questions remain. First, what is the prevalence of anaemia at the time of presentation to hospital in a contemporary acute coronary syndrome population? Second, what is the incidence of anaemia in hospital in-patients with a normal admission haemoglobin? Third, what are the aetiologies of anaemia (e.g. iron deficiency) in acute coronary syndrome patients? Fourth, what might the transfusion requirements be in these patients? Fifth, elevated haemoglobin is associated with an adverse prognosis in ischaemic heart disease patients,7,16 but the prevalence and associations of an increased haemoglobin level in acute coronary syndrome patients are uncertain.
We hypothesized that patients who present in normal contemporary practice differ markedly from those reported in clinical trials, and that reduced and elevated haemoglobin levels may commonly occur in an acute coronary syndrome population. Our aims were first, to characterize the range of haemoglobin levels in consecutive acute coronary syndrome admissions, and define the prevalence and associates of anaemia in this group; second, to characterize the nature of haemoglobin abnormalities in these patients; and third, by tracking these patients during admission, to define the incidence of anaemia and transfusion requirements in these patients.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionAppendix 1. Reference ranges...References |
|---|
Setting
We conducted a prospective observational study in Glasgow Royal Infirmary, an acute tertiary care hospital, from January to April 2005. Consecutive index admissions to the Acute Medical Receiving Unit or Coronary Care Units with a suspected acute coronary syndrome were included.
An acute coronary syndrome was diagnosed on the basis of the history, ECG and elevated cardiac biomarker concentration (troponin I concentration >0.2 µg/l).17 The initial diagnosis of a suspected ACS was made by the attending A&E and Acute Medicine physicians, in conjunction with the on-call cardiologist, where required. Patients who had been admitted to the Acute Receiving Unit were then re-assessed within 24 h by an experienced Specialist Chest Pain Admissions Nurse, who verified the diagnosis and completed a protocol-driven Chest Pain Triage form, if appropriate. The diagnosis was further verified by an investigator (KB), who checked both the Case Record and the Chest Pain Triage Form of individual patients.
Data collection
We collected data on demographic and clinical attributes that were relevant to ACS patients (e.g. cardiac biomarker status, anti-thrombotic and anti-ischaemic drug therapy use, socio-economic status, and renal function). As anaemia is known to be more common in patients with heart disease, and to have prognostic importance in this setting, we sought to describe the potential associations between the clinical characteristics of ACS patients, and their haematological status. We therefore included a broad range of clinical attributes to ensure an unbiased approach in this study.
Haematology and blood chemistry were obtained from the first admission blood sample, which was obtained in either the Accident and Emergency Unit of the Medical Admissions ward. Patient triage documents, which incorporated the Thrombolysis In Myocardial Infarction (TIMI) risk score,18,19 were prospectively tracked for clinical information.
Invasive management
Patients identified by the cardiology team as having high-risk features, such as ongoing ischaemia, were triaged for additional investigations, which usually involved coronary angiography. A post-procedure haemoglobin (>24 h) was checked for all patients who had undergone interventional management in order to determine the incidence of iatrogenic anaemia.
Elective admissions
Patients with stable angina admitted for elective PCI during the study period were also included. This group served as a control population for comparison with the acute coronary syndrome patients.
Laboratory analysis and definitions of abnormal haemoglobin levels
Anaemia was defined according to WHO criteria (haemoglobin <13.0 g/dl in men and <12.0 g/dl in women).20 Anaemic status was recorded when a haemoglobin result beneath the cut-off was detected at presentation to hospital, or during hospital admission. An incident case of anaemia was recorded when anaemia was detected in a patient who previously had a normal haemoglobin on admission. The number of anaemic patients on admission to hospital divided by the total number of admissions provided the prevalence of anaemia at hospital presentation. The incidence of anaemia after invasive management was calculated by dividing the number of new cases of anaemia divided by the number of patients with a normal haemoglobin prior to PCI. Elevated haemoglobin was defined as >16 g/dl, as this cut-off has an adverse prognosis in ACS patients.7
Automated analysers were used for the biochemical (Advia 1650, Bayer) and haematological (Sysmex, SE9500) measurements made on admission (reference ranges in Appendix 1). Glomerular filtration rate was estimated using the Modification of Diet in Renal Disease (MDRD) equation.21 Renal dysfunction was defined as a reduction in the glomerular filtration rate (<90 ml/min/1.73 m2).
Socioeconomic status
Socioeconomic status was determined using the Deprivation Category score.22 This is an ordinal index of socioeconomic status in Scotland based on place of residence, as defined by postcode. The score includes contemporary, census-based information on the percentage of the population with attributes related to poverty (e.g. without a car, subject to overcrowding, male unemployment, and low social class). The scale ranges from 1 (low socioeconomic status/deprived) to 7 (high socioeconomic status/affluent).
Duration of hospital admission
We took duration of hospital admission as a measure of in-patient morbidity. We included admission duration (days) as a continuous variable in the regression analyses.
Statistical analyses
Clinical characteristics were tabulated and compared between patients with normal, low, and elevated haemoglobin levels, using either one-way analysis of variance (ANOVA) or the Kruskall-Wallis test (for normally or non-normally distributed continuous data, respectively). Clinical characteristics were also compared between anaemic patients and non-anaemic patients using a two-sample t-test. A 
2 test was used to assess frequencies of categorical data across levels of haemoglobin. ANOVA test for trend was used to assess for trends in continuous data according to haemoglobin level, and a 2 test was also used to assess for a linear trend in the proportions of categorical data across haemoglobin levels. A significance level of p < 0.05 was adopted for the definition of a covariate of haemoglobin category. Correlations between haemoglobin and haematocrit were explored using a Pearson's test.
Haemoglobin values were categorized as low (anaemia), normal or elevated, and potential associations with haemoglobin level were identified using descriptive analyses (Table 1). Regression analysis addressed anaemia, but not elevated haemoglobin, as too few subjects (n = 34) had an elevated haemoglobin.
|
Univariate logistic regression models for anaemia were formed using these covariates. A standard multivariate logistic regression model was then constructed using explanatory variables that were significant (p < 0.05) in the univariate analysis, and were unrelated (i.e. either admission creatinine or admission glomerular filtration rate; either previous myocardial infarction or TIMI risk category). Finally, a Hosmer-Lemeshow goodness-of-fit statistic was calculated, where a significant result (p < 0.05) indicates a lack of fit in the regression model. A multivariate association significant at p < 0.05 was taken as a predictor of anaemia.
Associations with admission duration were also studied. Variables significant (p < 0.05) in a simple univariate regression analysis were entered into a forward step-wise regression model, with entry and removal criteria of p < 0.2 and p < 0.4, respectively. A multivariate association significant at p < 0.05 was taken as a predictor of admission duration.
All analyses used STATA (version 7) software (Stata Corporation). No adjustments were made for multiple testing.
Ethical approval
The Chairman of the North Glasgow Hospitals Ethics Committee (Royal Infirmary) approved this survey. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by this hospital's human research committee.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionAppendix 1. Reference ranges...References |
|---|
Patient admission characteristics
Of 337 consecutive admissions with a suspected ACS, a haemoglobin result was available for 335 (99.4%). After exclusion of 15 re-admissions, 320 (190 men) index admissions were studied (Table 1): 299 (71%) had unstable angina, 57 (18%) a non-ST-elevation myocardial infarction and 34 (11%) an ST-elevation myocardial infarction.
Admission haematology
Mean ± SD haemoglobin in the first admission blood sample was 14.3 ± 1.7 in men and 13.2 ± 1.5 g/dl in women. Haemoglobin level was related to gender (Table 1). The frequency of anaemia was similar in men (n = 35, 18%) and women (n = 24, 18%), whereas elevated admission haemoglobin (>16 g/dl) was more common in men (n = 30, 16%) men than in women (n = 4, 3%) (p < 0.01). All patients had an admission haemoglobin >8 g/dl.
The characteristics of patients according to haemoglobin level are described in Table 1. Mean age decreased linearly with increasing haemoglobin level. Anaemia occurred with a similar frequency in women (n = 20, 17%) and men (32, n = 19%) with unstable angina or non-ST-elevation MI (p = 0.7). Anaemia was present in 4 women (31%) and 3 men (14%) with ST-elevation MI (p = 0.3).
The proportion of cigarette smokers increased, and the proportion of patients with diabetes decreased, with increasing haemoglobin level. The proportion of ST elevation myocardial infarction patients with an elevated haemoglobin was greater than in patients with either unstable angina or an non-ST-elevation MI (26% vs. 9%; p = 0.005).
Mean ± SD admission haematocrit was 0.42 ± 0.046 in men and 0.39 ± 0.043 in women. By WHO haematocrit criteria for anaemia (<0.39 in men and <0.36 in women), an even higher proportion of patients were anaemic: 41 men (22%) and 30 women (23%).
Microcytic, normocytic or macrocytic anaemia at the time of admission was seen in 11 (19%), 42 (71%) and 6 (10%) patients, respectively. Haematinic results from admission blood samples were available for 43 anaemic patients (73%). A low ferritin (<14 µg/l) concentration was found in 2 patients (5%), a low plasma vitamin B12 concentration (<150 ng/l) in 2 (5%), a low red-cell folate (<160 g/l) concentration in 7 (16%), and an increased plasma thyroid-stimulating-hormone concentration (>5.5 mU/l) in 4 (9%).
TIMI risk category on admission to hospital
Of 286 patients (89%) presenting with suspected unstable angina or a non-ST-elevation myocardial infarction, TIMI risk scores were available in 282 (92%). TIMI scores were categorized as low (0–2), intermediate (3–4) or high (5–7) in 163 (58%), 88 (31%) and 31 (11%) patients, respectively. Fewer anaemic patients had a low-risk TIMI score, compared to patients with a normal or elevated haemoglobin (Table 1).
Socio-economic status and anaemia at admission
The frequency of anaemia was similar in patients with a high (1–2), intermediate (3–4) or low (5–7) Deprivation Category score: 2 (17%) vs. 10 (15%) vs. 47 (20%); p = 0.724.
Drug therapy at hospital presentation
Clopidogrel therapy was recorded more often in anaemic patients than in patients with a normal or elevated haemoglobin (Table 1). Insulin-treated diabetes was more prevalent in anaemic patients than in those with a normal haemoglobin: 6 (10%) vs. 5 (2%); p = 0.011. Proton-pump-inhibitor therapy use was rare (2% of patients).
Drug therapy initiated during admission
Patients with an elevated haemoglobin were more often treated with thrombolytic therapy than were patients with a normal or reduced haemoglobin level (Table 1).
Blood chemistry at hospital admission
Renal function
Mild (<90 ml/min/1.73 m2), moderately (690 ml/min/1.73 m2), and severely (<30 ml/min/1.73 m2) impaired renal function (as assessed by glomerular filtration rate) was recorded in 297 (93%), 105 (33%), and 11 (3%) patients, respectively. Mean glomerular filtration rate increased with haemoglobin level (Table 1). Compared to non-anaemic patients, a higher proportion of anaemic patients had moderate (47% vs. 30%; p = 0.009) or severe (8% vs 2%; p = 0.036) renal impairment. Mean ± SD haemoglobin concentrations in men with glomerular filtration rates <90, 60 to <90, 30 to <60, and <30 ml/min/1.73 m2, were 14.7 ± 1.8, 14.5 ± 1.5, 13.9 ± 2.1, and 12.4 ± 2.2 g/dl, respectively (p = 0.0137). The corresponding mean haemoglobin concentrations in women were 13.5 ± 0.6, 13.5 ± 1.36, 13.0 ± 1.7, and 11.7 ± 1.5 g/dl, respectively (p = 0.0138). Admission haemoglobin was correlated with creatinine (r = –0.164, p = 0.003) and glomerular filtration rate (r = 0.338, p < 0.01).
Albumin
Mean serum albumin concentration increased with increasing haemoglobin level (Table 1). Low serum albumin (<36 g/l) was seen more often in anaemic patients (n = 10, 17%) than in non-anaemic patients (n = 7, 3%) (p = 0.001). Albumin and haemoglobin were weakly correlated (r = 0.27, p < 0.01).
Systemic inflammation
Plasma C-reactive protein concentration at admission was higher in anaemic patients than in non-anaemic patients (Table 1).
Associations with anaemia on admission to hospital
Univariate and multivariate associations with anaemia are shown in Tables 2a and 2b, respectively. On multivariate analysis, anaemia was associated with increasing age and albumin. Clopidogrel and smoking tended to be positive and negative predictors of anaemia, respectively.
|
|
Anaemia in ACS patients selected for invasive management
Ninety-seven (30%) patients with an acute coronary syndrome were referred for interventional management. Pre- and post-catheterization-lab haemoglobin values were available in 85 (88%) of these patients. Femoral access was used in 71 (84%) and radial access in 14 (16%) patients.
Mean ± SD post-procedure haemoglobin was lower than pre-procedure haemoglobin (14.1 ± 1.7 vs. 13.3 ± 1.9 g/dl; p < 0.01). Of the 19 patients (22%) who experienced a haemoglobin fall >10%, 8 (42%) had undergone angiography alone, and 11 (58%) had undergone follow-on PCI (p = 0.32). Incident cases of anaemia (n = 15) occurred more commonly in patients with a history of myocardial infarction (38% vs. 10%; p = 0.008), and in those with a glomerular filtration rate <60 ml/min/1.73 m2 (36% vs. 10% in those with glomerular filtration rate 
60 ml/min/1.73 m2; p = 0.006). Incident cases of anaemia tended to occur in women (27% vs. 8%; p = 0.15), in patients aged >65 years (29% vs. 12%; p = 0.07), and in those with a higher TIMI risk score (low 11%, intermediate 13%, high 37%; p = 0.08) but these trends were not statistically significant.
Post-catheterization-lab haemoglobin and haematocrit values were highly correlated (r = 0.991, p < 0.01), indicating that haemodilution was unlikely to be a cause of anaemia after interventional management.
Anaemia in stable patients undergoing elective percutaneous coronary intervention
Thirty patients who had elective percutaneous coronary intervention during the study period and who had haemoglobin values available were included as a reference group. The mean ± SD haemoglobin values pre- and 24 h post-procedure were 13.5 ± 1.3 g/dl and 13.4 ± 1.3 g/dl, respectively.
The prevalence rates of anaemia in patients with stable (reference group) and unstable (study group) coronary artery disease selected for PCI were 6/30 (20%) and 15/97 (15%), respectively (p = 0.34). One patient in the stable group and 12 patients in the unstable group did not have a post-procedure haemoglobin result. The incidence rates of anaemia in stable and unstable patients who underwent interventional management groups were similar: 7/29 (24%) vs. 15/85 (18%); p = 0.44.
Blood transfusions during hospital admission
Seventeen (5%) acute coronary syndrome patients were transfused, of whom 6 (35%) were anaemic on admission to hospital. Of the patients who had undergone interventional management (n = 85), 3 (4%) and 8 (9%) patients were transfused pre- and post-catheterization-lab, respectively.
Duration of hospital admission
Compared with patients who had a normal admission haemoglobin, anaemic patients had a longer median (IQR) duration of admission: 7 (3–15) vs. 3 (1–7) days; p = 0.0002. Admission duration was similar in those with haemoglobin >16 g/dl and those with haemoglobin <16 g/dl: 4 (2–6) vs. 4 (2–9) days; p = 0.737.
Admission characteristics that were univariate and multivariate predictors of admission duration (days) are shown in Tables 3a and 3b. Haemoglobin, creatinine, troponin, and C-reactive protein were multivariate predictors of admission duration.
|
|
Discharge outcomes
Five patients died (2%), four of whom had undergone non-invasive management. Of the in-hospital deaths, three (60%) were anaemic on admission to hospital (anaemia vs. normal haemoglobin: p = 0.038; after adjustment for baseline differences between anaemic and non-anaemic patients (Table 1): p = 0.0135). Mean±SD haemoglobin was higher in patients who survived to discharge vs. those who died: 13.9 ± 1.7 vs. 12.6 ± 2.1 g/dl; p = 0.115.
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionAppendix 1. Reference ranges...References |
|---|
In this group of consecutive acute coronary syndrome admissions, nearly one third had an abnormal admission haemoglobin. One fifth of these patients were anaemic and a further 10% had an elevated haemoglobin. All cases of anaemia were relatively mild and usually normocytic. Haemoglobin level was negatively related to patient age and diabetes, and positively related to glomerular filtration rate and albumin concentration. Anaemia tended to be more common in older patients and in those treated with clopidogrel, whereas elevated haemoglobin levels tended to be more common in men and smokers. The type of acute coronary syndrome presentation was also related to haemoglobin level, and haemoglobin elevations were most often observed in patients with an ST-elevation myocardial infarction. The incidence of anaemia in patients undergoing invasive management was 15%. Anaemia predicted duration of admission, and was more common in patients who died.
Prognostically important abnormal haemoglobin levels were common in our acute coronary syndrome patients, with a higher prevalence than has been previously reported. For example, one retrospective analysis of acute MI admissions to non-Federal hospitals in New Jersey in 1986 (prethrombolytic era, n = 15 584) and 1996 (thrombolytic era, n = 14 757) detected anaemia in 6.4% of patients in 1986 and 10.2% (p < 0.0001) of patients in 1996.5 The reasons for the increased prevalence of anaemia in our study may include definition of anaemia, changing prevalence with time, different populations, or method of data collection. The frequency of haemoglobin elevation in prospectively documented acute coronary syndrome admissions has not been previously reported.
Few prospective studies of haemoglobin levels in ischaemic heart disease have been undertaken. One investigation in women with suspected ischaemic chest pain referred for coronary angiography, which used the same cut-off for anaemia as in our study, detected anaemia in 21% of patients.23 In one other recent single-centre report of consecutive patients undergoing primary angioplasty, anaemia was detected in 24.6% of patients.11 Taken together with our own findings, the prevalence of anaemia appears consistently higher in contemporary populations than in those of earlier investigations.5 One reason for the high frequency of anaemia in acute coronary syndrome patients may be the significant risk of bleeding associated with contemporary interventions. In the Fifth Organization to Assess Strategies in Acute Ischemic Syndromes Investigators (OASIS-5), a trial of enoxaprin versus the anti-thrombin inhibitor fondaparinux, in 20 078 acute coronary syndrome patients, the 9-day rates of major and minor bleeding in the enoxaparin group were 4.1% and 3.2%, respectively.24 Fondaparinux reduced the risk of major bleeding by nearly half (hazard ratio 0.52; p < 0.01). Our findings add weight to the known prognostic importance of blood loss and anaemia in acute coronary syndrome patients.25 Treatment with the anti-platelet drug, clopidogrel, and the use of invasive management, were both associated with anaemia, presumably due to blood loss in these patients, whether occult or clinically apparent.
Although anaemia is generally associated with an adverse prognosis, elevated haemoglobin concentrations may also be harmful. In their analyses of the TIMI acute coronary syndrome populations, Sabatine et al.7 described a J-shaped relationship between haemoglobin and mortality, with both reduced (haemoglobin <11 g/dl) and elevated (>16 g/dl) haemoglobin concentrations being associated with an adverse prognosis. In their retrospective analysis of patients undergoing percutaneous coronary intervention, Reinecke et al.9 found that patients with a haemoglobin >15.3 g/dl were also at increased risk of death, with an adjusted hazard ratio (95%CI) of 2.26 (0.79–6.48). Haemoglobin >15.3 g/dl was seen in 22% of our patients, and haemoglobin >16 g/dl was also relatively common, particularly in men and smokers. Notably, there was a linear trend with smoking status across haemoglobin level (Table 1). Haemoglobin elevation in smokers26 may reflect a compensatory response to reduced oxygen delivery associated with elevated carboxyhaemoglobin.27
Overall, haematinic abnormalities were uncommon. The most frequent haematinic abnormality was a reduced plasma folate concentration, observed in 16% of the anaemic patients, supporting the possibility of malnutrition in some of our patients. Serum albumin was lower in anaemic patients, compared to in non-anaemic patients, suggesting nutritional deficiency may be an aetiological factor in some anaemic patients. Although we cannot discount a dilutional effect on albumin (and haemoglobin) concentrations, the very strong correlation between haemoglobin and haematocrit on admission (r = 0.98, p < 0.0001) suggests that haemodilution was not an important cause of anaemia in our patients. A reduced ferritin concentration (<14 µg/l) tended to be less common in anaemic (5%) patients, compared with non-anaemic patients (10%). Ferritin is an acute-phase protein, and a reduction in ferritin because of iron deficiency may have been masked by an acute inflammatory response in some patients. When the cut-off of the reference range for ferritin is taken as <40 µg/l, one third of anaemic patients had a low ferritin concentration, suggesting that iron deficiency (and possibly blood loss) was associated with anaemia in these patients.
Although mean troponin concentrations were similar in anaemic and non-anaemic patients, anaemic patients had a higher TIMI risk score, tended to have ST-segment depression more frequently (21 vs. 11%; p = 0.161) and received IV nitrate therapy more often. Taken together, these findings suggest acute coronary syndrome patients with anaemia may experience more pronounced ischaemia. This observation is supported by the recent report by Sabatine et al.,7 who found that in patients with an ST elevation myocardial infarction, heart rate, likelihood of hypotension and Killip class were related to haemoglobin level. In patients with an non-ST-elevation myocardial infarction, a J-shape relationship existed between haemoglobin and ST segment depression.8
Haemoglobin level was negatively related to diabetic status, and anaemia was most common in insulin-treated diabetic patients. The inverse association between haemoglobin level and diabetes may be explained by the influence of other clinical factors. After adjustment for a wide range of other univariate predictors of anaemia in our population, diabetes was no longer a predictor of anaemia. Notably, renal dysfunction,28,29 associated with reduced erythropoeitin production and an impaired haemopoeitic response to this hormone, is one likely explanation for the association between diabetes and anaemia.30 The prevalence of diabetes increases with advancing age, which may help explain the association between anaemia and diabetes, as both are more common in the elderly.
Haemoglobin level was also related to renal function. In the present study, over 95% of anaemic patients had a reduced glomerular filtration rate, and the prevalence and severity of anaemia was related to the level of renal function. Anaemia is a common associate of advanced renal disease, but may also occur in lesser degrees of renal impairment. Renal dysfunction is an adverse prognostic marker in ischaemic heart disease patients,21 particularly those who are anaemic.9,12,31 Furthermore, the combination of anaemia and renal dysfunction confers a particularly poor prognosis in more severe forms of ischaemic heart disease, such as heart failure.31 We also observed that systemic inflammation, reflected by CRP concentration, was more pronounced in anaemic patients. This may have aetiological significance, as cytokines inhibit erythropoietin production, leading to reduced red blood cell production.32
We were surprised to find that statin therapy was more common in anaemic subjects (Table 1). Statin use tended to be a univariate predictor of anaemia (OR 1.85, 95%CI 0.98–3.5; p = 0.058), but was not a multivariate predictor of anaemia. Statin use was most likely a marker for chronic heart disease, which was more common in elderly subjects. As the most robust associate of anaemia was increasing age, we suspect statin use was associated with anaemia through the associations between statin use, comorbidity (notably chronic heart disease) and increasing age. ACE-inhibitor therapy was also more common in anaemic patients. The association between ACE inhibitors and anaemia is well described,33 and the mechanism for this effect is probably due to an ACE-inhibitor-mediated increase in the metabolism of erythropoeitin.
Non-steroidal anti-inflammatory drugs and aspirin, which can cause gastrointestinal bleeding, were used with similar frequency in anaemic and non-anaemic patients. However, the thienopyridine anti-platelet agent, clopidogrel, was used much more often in anaemic patients, and clopidogrel use was a multivariate associate of anaemia that approached conventional statistical significance. Although we did not assess for bleeding complications in this population, clopidogrel is a predictor of bleeding in ischaemic heart disease patients undergoing invasive management.34 While this drug has been reported as a cause of marrow suppression, this is a very rare side-effect. One alternative potential explanation of the association between clopidogrel with anaemia is that this drug may have been prescribed in higher-risk patients in whom the prevalence of anaemia was greater. There was a negative association between serum albumin concentration and anaemia, which may reflect impairments in anabolic and haemopoeitic function.
Compared with non-anaemic patients, anaemic patients undergoing coronary angioplasty have a worse long-term prognosis.9–11 However, the incidence of anaemia occurring after coronary angioplasty has not been previously described. We found that a very high proportion of patients (68%) experienced a fall in haemoglobin after either angiography or PCI, and a fall in haemoglobin >10% occurred in 22% of patients. The incidence of anaemia overall was 15%. The majority of procedures used femoral artery access. Although the causes of anaemia were not evaluated, periprocedural blood loss was a likely factor in many of these patients.34 We were surprised to find that the prevalences and incidences of anaemia in stable versus unstable PCI patients were similar. One might have expected anaemia to have been uncommon in elective cases who normally undergo pre-PCI assessment. Furthermore, the incidence of anaemia post-PCI may be expected to be higher in unstable CAD patients, who may have been treated with aggressive anti-thrombotic strategies leading to a greater risk of bleeding, than in stable elective PCI patients. The relatively high rates of anaemia in both groups of PCI patients underline the need for careful screening and prevention measures pre-PCI.
Duration of admission was assessed as a surrogate measure of hospital outcome. In a multivariate analysis, anaemia and haemoglobin level were independent predictors of admission duration, and anaemia was also associated with in-hospital death. These findings are supported by earlier studies.3,7,9,11–13 In the CADILLAC trial,13 which enrolled acute MI patients undergoing primary PCI, and in a single-centre registry of primary angioplasty performed between 1994–1999, anaemic patients had a longer duration of admission,13 and anaemia was an independent predictor of in-hospital and 1-year mortality.11,13 In the TIMI acute coronary syndrome trial populations, anaemia predicted cardiovascular mortality.7
We believe that this report provides clinically relevant information about haemoglobin abnormalities in unstable coronary artery disease, adding to what is already known in this area. The characteristics of the patients included in this study are comparable to those in one other UK acute coronary syndrome registry (the Prospective Registry of Acute Ischaemic Syndromes in the UK; PRAIS-UK);35 however, PRAIS has not provided information on haemoglobin. Compared to earlier acute coronary syndrome investigations which provided limited haematological information, our report provides contemporary information on a broader range of haemoglobin abnormalities (including elevated haemoglobin), and describes the relationships between haemoglobin and contemporary anti-thrombotic therapies (e.g. thienopyridine and Gp IIb/IIIa inhibitor therapies),3,5,7,23 in a population of patients that is not restricted by age,3,7 diagnosis,3,5,11,13 or concomitant health problems.7,13,23
One key implication of our study is the importance of detecting and preventing anaemia in patients with unstable coronary artery disease. The very low rate of proton-pump-inhibitor use in patients at the time of hospital presentation suggests under-recognition of mild anaemia in the community. The role of drug therapy to prevent anaemia in at-risk patients, such as those taking chronic anti-platelet therapies, is uncertain, and correction of anaemia in ACS patients by blood transfusion may be harmful.36 Prevention and earlier detection of anaemia may lead to improved outcomes in ACS patients. Potential beneficial interventions which merit prospective testing are the prophylactic use of proton-pump antagonists in at-risk patients (e.g. elderly ischaemic heart disease patients taking clopidogrel), or a screening program with interval measurement of haemoglobin in community-based patients
Strengths and limitations
We have provided a comprehensive haematological assessment of a reasonably large number of consecutive, acute coronary syndrome patients. Due to the complexities of managing patients in a high-throughput medical admissions unit, several anaemic patients did not have haematinic information obtained, indicating under-recognition of this problem by clinicians. Our findings are based on the analysis of a UK urban population, and may not be reflective of populations in other settings. Due to the limited sample size and absence of follow-up data the a priori purpose of this study was not to provide prognostic information, which has been well described elsewhere, but rather to provide a descriptive report of the frequency and associates of haemoglobin problems in contemporary acute coronary syndrome patients.
Conclusions
Haemoglobin abnormalities are common and potentially modifiable predictors of poor outcome in acute coronary syndrome patients. Anaemia was mostly mild, and was associated with co-morbidity, adjunctive anti-platelet therapy, and invasive management. Blood transfusions were also relatively common (10% of patients), and anaemia was related to in-hospital outcomes. Although these findings are exploratory in nature, they indicate that haemoglobin abnormalities are very common in contemporary acute coronary syndrome patients.
|
Appendix 1. Reference ranges in SI units |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionAppendix 1. Reference ranges...References |
|---|
(i) Haematology
|
(ii) Blood chemistry
|
|
Acknowledgments |
|---|
Funding to pay the Open Access publication charges for this article was provided by Royal Infirmary Departmental funds.
We thank Susan Bateman of the Department of Transfusion Medicine, Glasgow Royal Infirmary, and Dr Lilian Murray, Department of Medicine and Therapeutics, University of Glasgow. This project was funded by the University of Glasgow. Dr Berry and Professor Cobbe are supported by the British Heart Foundation.
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionAppendix 1. Reference ranges...References |
|---|
1. Goldberg RJ, Currie K, White K, Brieger D, Steg PG, Goodman SG, Dabbous O, Fox KAA, Gore JM. (2004) Six-month outcomes in a multinational registry of patients hospitalized with an acute coronary syndrome (The Global Registry of Acute Coronary Events [GRACE]). Am J Cardiol 93 288–93.[CrossRef][Web of Science][Medline]
2. Murphy N, MacIntyre K, Capewell S, Stewart S, Pell J, Chalmers J, Redpath A, Frame S, Boyd J, McMurray JJV. (2003) The chest pain epidemic: A new clinical challenge for the NHS. Heart 89 A3.
3. Wu WC, Rathore SS, Wang Y, Radford MJ, Krumholz HM. (2001) Blood transfusion in elderly patients with acute myocardial infarction. N Engl J Med 345 1230–6.
4. Sachdev M, Sun JLN, Tsiatis AA, Nelson CL, Mark DB, Jollis JG. (2004) The prognostic importance of comorbidity for mortality in patients with stable coronary artery disease. J Am Coll Cardiol 43 576–82.
5. Al Falluji N, Lawrence-Nelson J, Kostis JB, Lacy CR, Ranjan R, Wilson AC. (2002) Effect of anemia on 1-year mortality in patients with acute myocardial infarction. Am Heart J 144 636–41.[Web of Science][Medline]
6. Goncalves PA, Ferreira J, Aguiar C, Trabulo M, Silva JA, Seabra-Gomes R. (2002) Prognostic value of baseline hemoglobin in acute coronary syndromes. Circulation 106 402.
7. Sabatine MS, Morrow DA, Giugliano R, Burton PBJ, Murphy SA, McCabe C, Gibson CM, Braunwald E. (2005) Association of hemoglobin levels with clinical outcomes in acute coronary syndromes. Circulation 111 2042–9.
8. Zeidman A, Fradin Z, Blecher A, Oster HS, Avrahami Y, Mittelman M. (2004) Anemia as a risk factor for ischemic heart disease. Isr Med Assoc J 6 16–18.[Web of Science][Medline]
9. Reinecke H, Trey T, Wellmann J, Heidrich J, Fobker M, Wichter T, Walter M, Breithardt G, Schaefer RM. (2003) Haemoglobin-related mortality in patients undergoing percutaneous coronary interventions. Eur Heart J 24 2142–50.
10. McKechnie RS, Smith D, Montoye C, et al. (2004) Prognostic implication of anemia on in-hospital outcomes after percutaneous coronary intervention. Circulation 110 271–7.
11. Nikolsky E, Mehran R, Aymong ED, Mintz GS, Lansky AJ, Lasic Z, Negoita M, Fahy M, Pocock SJ, Na YB, Krieger S, Moses JW, Stone GW, Leon MB, Dangas G. (2004) Impact of anemia on outcomes of patients undergoing percutaneous coronary interventions. Am J Cardiol 94 1023–7.[CrossRef][Web of Science][Medline]
12. Lee PC, Kini AS, Ahsan C, Fisher E, Sharma SK. (2004) Anemia is an independent predictor of mortality after percutaneous coronary intervention. J Am Coll Cardiol 44 541–6.
13. Nikolsky E, Aymong ED, Halkin A, et al. (2004) Impact of anemia in patients with acute myocardial infarction undergoing primary percutaneous coronary intervention – Analysis from the Controlled Abciximab and Device Investigation to Lower Late Angioplasty Complications (CADILLAC) Trial. J Am Coll Cardiol 44 547–53.
14. Yusuf S, Fox KAA, Tognoni G, et al. (2001) Effects of clopidogrel in addition to aspirin in patients with acute coronary syndromes without ST-segment elevation. N Engl J Med 345 494–502.
15. Mehta SR, Yusuf S, Peters RJG, Bertrand ME, Lewis BS, Natarajan MK, Maimberg K, Rupprecht HJ, Zhao F, Chrolavicius S, Copland I, Fox KAA. (2001) Effects of pretreatment with clopidogrel and aspirin followed by long-term therapy in patients undergoing percutaneous coronary intervention: the PCI-CURE study. Lancet 358 527–33.[CrossRef][Web of Science][Medline]
16. Sabatine MS, McCabe CH, Morrow DA, Giugliano RP, de Lemos JA, Cohen M, Antman EM, Braunwald E. (2002) Identification of patients at high risk for death and cardiac ischemic events after hospital discharge. Am Heart J 143 966–70.[CrossRef][Web of Science][Medline]
17. Bertrand ME, Simoons ML, Fox KAA, Wallentin LC, Hamm CW, McFadden E, De Feyter PJ, Specchia G, Ruzyllo W. (2002) Management of acute coronary syndromes in patients presenting without persistent ST-segment elevation. Eur Heart J 23 1809–40.
18. Antman EM, Cohen M, Bernink PJLM, McCabe CH, Horacek T, Papuchis G, Mautner B, Corbalan R, Radley D, Braunwald E. (2000) The TIMI risk score for unstable angina/non-ST elevation MI—A method for prognostication and therapeutic decision making. JAMA 284 835–42.
19. Soiza RL, Leslie SJ, Williamson S, Wai S, Harrild K, Peden NR, Hargreaves AR. (2006) Risk stratification in acute coronary syndromes-does the TIMI risk score work in unselected cases? Q J Med 99 69–79.
20. The World Health Organisation. The International Classification of Diseases : ninth revision (ICD-9) [http://www.who.int/whosis/icd10/].
21. Gibson CM, Dumaine RL, Gelfand EV, Murphy SA, Morrow DA, Wiviott SD, Giugliano RP, Cannon CP, Antman EM, Braunwald E. (2004) Association of glomerular filtration rate on presentation with subsequent mortality in non-ST-segment elevation acute coronary syndrome; observations in 13307 patients in five TIMI trials. Eur Heart J 25 1998–2005.
22. MRC Social and Public Health Services Unit. (2001) Executive Summary—Carstairs scores for Scottish postcode sectors from the 2001 Census.
23. Arant CB, Wessel TR, Olson MB, Merz CNB, Sopko G, Rogers WJ, Sharaf BL, Reis SE, Smith KM, Johnson BD, Handberg E, Mankad S, Pepine CJ. (2004) Hemoglobin level is an independent predictor for adverse cardiovascular outcomes in women undergoing evaluation for chest pain—Results from the national heart, lung, and blood institute women's ischemia syndrome evaluation study. J Am Coll Cardiol 43 2009–14.
24. Yusuf S, Mehta SR, Bassand JP, Budaj A, Chrolavicius S, Fox KAA, et al. (2006) Comparison of fondaparinux and enoxaparin in acute coronary syndromes. N Engl J Med 354 1464–76.
25. Rao SV, O’Grady K, Pieper KS, Granger CB, Newby LK, Van de Werf F, Mahaffey KW, Califf RM, Harrington RA. (2005) Impact of bleeding severity on clinical outcomes among patients with acute coronary syndromes. Am J Cardiol 96 1200–6.[CrossRef][Web of Science][Medline]
26. Shimakawa T and Bild DE. (1993) Relationship between hemoglobin and cardiovascular risk-factors in young-adults. J Clin Epidemiol 46 1257–66.[CrossRef][Web of Science][Medline]
27. Rietbrock N, Kunkel S, Worner W, Eyer P. (1992) Oxygen-dissociation kinetics in the blood of smokers and non-smokers—interaction between oxygen and carbon-monoxideat the hemoglobin molecule. Naunyn-Schmiedebergs Arch Pharmacol 345 123–8.[Web of Science][Medline]
28. Silverberg DS, Wexler D, Blum M, Iaina A. (2003) The Cardio Renal Anemia Syndrome: correcting anemia in patients with resistant congestive heart failure can improve both cardiac and renal function and reduce hospitalizations. Clin Nephrol 60 S93–102.
29. Yun YS, Lee HC, Yoo NC, Song YD, Lim SK, Kim KR, Huh KB. (1999) Reduced erythropoietin responsiveness to anemia in diabetic patients before advanced diabetic nephropathy. Diabetes Res Clin Prac 46 223–9.[CrossRef][Web of Science][Medline]
30. Thomas MC, Cooper ME, Tsalamandris C, MacIsaac R, Jerums G. (2005) Anemia with impaired erythropoietin response in diabetic patients. Arch Int Med 165 466–9.
31. Berry C, Norrie JN, Hogg K, Brett M, Stevenson K, McMurray JJV. (2006) The prevalene, nature and importance of hematologic abnormalities in heart failure. Am Heart J 151 1513–21.
32. Means RT. (1999) Advances in the anemia of chronic disease. Int J Hematol 70 7–12.[Web of Science][Medline]
33. Ishani A, Weinhandl E, Zhao ZH, Gilbertson DT, Collins AJ, Yusuf S, Herzog CA. (2005) Angiotensin-converting enzyme inhibitor as a risk factor for the development of anemia, and the impact of incident anemia on mortality in patients with left ventricular dysfunction. J Am Coll Cardiol 45 391–9.
34. Berry C, Kelly J, Cobbe SM, Eteiba H. (2004) Comparison of femoral bleeding complications after coronary anglography versus percutaneous coronary intervention. Am J Cardiol 94 361–3.[CrossRef][Web of Science][Medline]
35. Collinson J, Flather MD, Fox KAA, Findlay I, Rodrigues E, Dooley P, Ludman P, Adgey J, Bowker TJ, Mattu P. (2000) Clinical outcomes, risk stratification and practice patterns of unstable angina and myocardial infarction without ST elevation: Prospective Registry of Acute Ischaemic Syndromes in the UK (PRAIS-UK). Eur Heart J 21 1450–7.
36. Rao SV, Jollis JG, Harrington RA, Granger CB, Newby LK, Armstrong PW, Moliterno DJ, Lindblad L, Pieper K, Topol EJ, Stamler JS, Califf RM. (2004) Relationship of blood transfusion and clinical outcomes in patients with acute coronary syndromes. JAMA 292 1555–62.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. H. Shishehbor, S. Madhwal, V. Rajagopal, A. Hsu, P. Kelly, H. S. Gurm, S. R. Kapadia, M. S. Lauer, and E. J. Topol Impact of Blood Transfusion on Short- and Long-Term Mortality in Patients With ST-Segment Elevation Myocardial Infarction J. Am. Coll. Cardiol. Intv., January 1, 2009; 2(1): 46 - 53. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Agarwal, J. L. Davis, and L. Smith Serum Albumin Is Strongly Associated with Erythropoietin Sensitivity in Hemodialysis Patients Clin. J. Am. Soc. Nephrol., January 1, 2008; 3(1): 98 - 104. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||