Summary
Background
Owing to the heterogeneous population of patients with acute coronary syndromes (ACS), risk stratification with tools such as the GRACE risk score is recommended to guide therapeutic management and improve outcome.
Aim
To evaluate whether anaemia refines the value of the GRACE risk model to predict midterm outcome after an ACS.
Methods
A prospective registry of 1064 ACS patients (63 ± 14 years; 73% men; 57% ST-segment elevation myocardial infarction [MI]) was studied. Anaemia was defined as haemoglobin less than 13 mg/dL in men or less than 12 mg/dL in women. The primary endpoint was 6-month death or rehospitalization for MI.
Results
The primary endpoint was reached in 132 patients, including 68 deaths. Anaemia was associated with adverse clinical outcomes (hazard ratio 3.008, 95% confidence interval 2.137–4.234; P < 0.0001) in univariate analysis and remained independently associated with outcome after adjustment for the Global Registry of Acute Coronary Events (GRACE) risk score (hazard ratio 2.870, 95% confidence interval 1.815–4.538; P < 0.0001). Anaemia provided additional prognostic information to the GRACE score as demonstrated by a systematic improvement in global model fit and discrimination (c-statistic increasing from 0.633 [0.571;0.696] to 0.697 [0.638;0.755]). Subsequently, adding anaemia to the GRACE score led to reclassification of 595 patients into different risk categories; 16.5% patients at low risk (≤ 5% risk of death or rehospitalization for MI) were upgraded to intermediate (> 5–10%) or high risk (> 10%); 79.5% patients at intermediate risk were reclassified as low (55%) or high risk (24%); and 45.5% patients at high risk were downgraded to intermediate risk. Overall, 174 patients were reclassified into a higher risk category (17.3%) and 421 into a lower risk category (41.9%).
Conclusion
Anaemia provides independent additional prognostic information to the GRACE score. Combining anaemia with the GRACE score refines its predictive value, which often overestimates the risk.
Résumé
Contexte
En raison de l’extrême hétérogénéité des patients hospitalisés pour un syndrome coronaire aigu, une stratification du risque est recommandée avec des outils tels que la score GRACE pour guider la prise en charge thérapeutique et améliorer le pronostic.
Objectif
Évaluer si l’anémie affine la valeur pronostique du score de prédication GRACE pour prédire les évènements cliniques à moyen terme après un syndrome coronaire aigu (SCA).
Méthodes
Un registre prospectif de 1064 patients admis pour un SCA a été étudié (63 ± 14 ans ; 73 % d’hommes ; 57 % de SCA ST sus). L’anémie était définie par un taux d’hémoglobine inférieur à 13 mg/dL chez les hommes et inférieur à 12 mg/dL chez les femmes. Le critère d’évaluation principal était la survenue de décès ou ré-hospitalisation pour infarctus du myocarde à six mois.
Résultats
Le critère d’évaluation était observé chez 132 patients incluant 68 décès. L’anémie était associée à un mauvais pronostic (HR 3,008, CI 95 % 2,137–4,234 ; p < 0,0001) en analyse univariée et restait indépendamment associée au pronostic après ajustement sur le score de risque GRACE (HR 2,870, CI 95 % 1,815–4,538 ; p < 0,0001). L’anémie ajoutait une information pronostique additionnelle à celle du modèle GRACE, cela était démontré par l’amélioration consistante du modèle global d’ajustement et discrimination (augmentation de la statistique c de 0,633 [0,571 ; 0,696] à 0,697 [0,638 ; 0,755]). L’incorporation de l’anémie dans le score GRACE permettait de reclasser 595 patients dans des catégories différentes de risque ; 16,5 % des patients à bas risque (≤ 5 % de décès pour ré-hospitalisation pour IDM) étaient reclassés à risque intermédiaire (> 5–10 %) ou élevé (> 10 %) ; 79,5 % des patients à risque intermédiaire étaient reclassés à bas risque (55 %) ou risque élevé (24 %) et 45,5 % des patients à risque élevé étaient reclassés à risque intermédiaire. Globalement, 174 patients étaient reclassés à risque plus élevé (17,3 %) et 421 à risque moins élevé (41,9 %).
Conclusion
L’anémie apporte une information pronostique additionnelle indépendante à celle du score de risque GRACE. Combiner l’anémie au score GRACE affine la valeur prédictive du score GRACE qui isolé surestime souvent le risque d’évènements.
Background
Owing to the broad and heterogeneous population of patients with acute coronary syndromes (ACS), early risk stratification is highly recommended to guide therapeutic management and improve outcome (European Society of Cardiology [ESC] guidelines) . Age, Killip class, heart rate and systolic blood pressure have been identified as powerful indicators of risk . When combined with electrocardiography, creatinine and troponin concentrations, these variables can be modelled into an easy-to-use score to further improve risk stratification. The Global Registry of Acute Coronary Events (GRACE) risk score has been recommended as a decision-making algorithm for the management of ACS patients (ESC guidelines) .
It is relevant for clinicians to know whether patient outcome can be improved by implementing additional easy-to-capture features. Anaemia, an independent correlate of survival , is present in up to 43% of the elderly population presenting with an ACS but is not considered in the GRACE score.
Our aim was to assess whether anaemia can improve the predictability of the GRACE score in a prospective cohort of ‘real-life’ patients hospitalized with an ACS.
Methods
Study population and follow-up
Data from 1064 consecutive patients with ST-segment elevation or non-ST-segment elevation ACS admitted between 2009 and 2010 to two tertiary care centres (CHRU Lille and CHU La Pitié) were prospectively collected. ACS was defined using the ESC/American College of Cardiology guidelines . Exclusion criteria were mechanical ventilation or initial cardiac arrest, significant valvular heart disease, bleeding events before ACS and postoperative ACS. Blood samples were drawn for serum creatinine and haemoglobin concentration determinations on admission. Anaemia was defined according to the World Health Organization definition (haemoglobin < 13 mg/dL in men or < 12 mg/dL in women) . Serum haemoglobin concentrations were obtained using a commercially available device (Beckmann Coulter Automated CBC Analyzer; Beckman Coulter, Inc., Fullerton, CA, USA). The glomerular filtration rate (GFR) was estimated using the four-component Modification of Diet in Renal Disease equation . The GRACE risk score model includes the following weighted variables: age, heart rate and systolic blood pressure on presentation, congestive heart failure Killip class, ST-segment deviation, initial serum creatinine concentration, elevated cardiac enzymes and cardiac arrest at admission .
The status of the patients was monitored by telephone calls to referring cardiologists and primary care physicians, and by review of the medical records; data were collected as part of their routine evaluation at baseline and at clinical follow-up. Informed consent for access to protected health information was obtained. The registry was declared to the Commission Nationale de l’Informatique et des Libertés (CNIL).
Endpoint
The primary endpoint, as in the GRACE report, was a composite of death and rehospitalization for myocardial infarction (MI), including ST-segment elevation MI or non-ST-segment elevation MI, 6 months after admission.
Statistical analysis
Continuous variables are expressed as mean ± standard deviation or median [25th;75th percentiles], as appropriate. Categorical variables are presented as absolute numbers and percentages. Baseline characteristics of subjects with anaemia were compared with those of subjects without anaemia using the Chi-square test or Fisher’s test for categorical variables and Student’s t -test for continuous variables.
Event-free survival curves were estimated using the Kaplan-Meier method and compared using the log-rank test. Median follow-up time was estimated with the reverse Kaplan-Meier method. Univariate followed by multivariable Cox analyses were performed to identify independent predictors of MI or death. The log-linearity assumption for continuous variables and the proportional hazard assumption were tested by supremum tests. In case of violation of the former assumption, the continuous variable was dichotomized, the cut-off value being visually established; in case of violation of the latter assumption, a piecewise model was used to model the hazard ratio (HR) as a step function of time, the cut-off value again being visually established. Owing to non-proportional hazards for the GRACE score, prognostic factors of event-free survival were identified beyond 15 days of admission for patients who remained alive and/or without any new MI (leading to a sample of 1003 subjects).
To assess the incremental value of anaemia over the GRACE score on event-free survival beyond a 15-day follow-up period, improvement in model performance was assessed by: Schwarz’s Bayesian criterion with the likelihood ratio test to evaluate the contribution of anaemia to the model combining anaemia with GRACE; change in Harrell’s c-statistics; change in Kent and O’Quigley’s ρ 2 ; and, following Chambless et al. , integrated discrimination improvement (IDI) and the continuous version of net reclassification improvement (cNRI) for censored data. These analyses were carried out in the sample of 1003 subjects (training sample).
For illustrational purposes, a category-based NRI was computed, with a risk of event defined as low, intermediate and high for an estimated cumulative incidence of ≤ 5%, > 5–10% and > 10%, respectively. For each category of risk based on the GRACE score alone, Kaplan-Meier event-free survival curves were plotted according to the reclassification with anaemia.
Internal validity of the model including anaemia was estimated by bootstrapping techniques. Two thousand bootstrap samples of 1003 subjects were generated and Harrell’s c-statistics, Kent and O’Quigley’s ρ 2 , the IDI and the cNRI were computed in each of the 2000 samples. For each performance index, given numbers are the mean of the 2000 estimates and 95% bootstrap confidence intervals, where the lower and upper bounds are the 2.5% and 97.5% percentiles of the resampling distribution, respectively.
A two-tailed type I error rate less than 0.05 was considered for statistical significance. Analyses were conducted using SAS software (SAS version 9.1, SAS Institute Inc., Cary, NC, USA).
Results
Study population
A total of 1064 patients with ACS were included in the present prospective observational registry. Most patients were men, 54% were hypertensive, 24% had diabetes mellitus, 17% had Killip class greater or equal to 2 and anaemia was diagnosed in 29% of patients on admission ( Table 1 ). Fig. 1 shows the distribution of serum haemoglobin at admission in the study population. Fifty-seven per cent had ST-segment elevation MI, 38% received glycoprotein IIb/IIIa inhibitors and 12% received fibrinolytic therapy. Coronary angiography was performed in most patients; primary percutaneous coronary intervention was the first reperfusion strategy in 48% and coronary artery bypass graft surgery was performed in 7% of the population.
| Variable | All | No anaemia | Anaemia | P |
|---|---|---|---|---|
| ( n = 1064) | ( n = 759) | ( n = 305) | ||
| Age (years) | 63 ± 14 | 60 ± 14 | 71 ± 13 | < 0.0001 |
| Male sex | 781 (73) | 584 (77) | 197 (65) | < 0.0001 |
| History of CAD | 277 (26) | 161 (21) | 116 (38) | < 0.0001 |
| Hypertension | 577 (54) | 376 (50) | 201 (67) | < 0.0001 |
| Diabetes mellitus | 253 (24) | 149 (20) | 104 (34) | < 0.0001 |
| Smoker | 470 (44) | 377 (50) | 93 (31) | < 0.0001 |
| ST-segment elevation MI | 608 (57) | 459 (60) | 149 (49) | 0.0005 |
| Anterior ST-segment elevation MI | 279 (27) | 224 (31) | 55 (19) | < 0.0001 |
| Killip class | < 0.0001 | |||
| 1 | 875 (83) | 658 (87) | 217 (72) | |
| 2 | 105 (10) | 59 (8) | 46 (15) | |
| 3 | 53 (5) | 23 (3) | 30 (10) | |
| 4 | 23 (2) | 13 (2) | 10 (3) | |
| Heart rate (bpm) | 81 ± 19 | 80 ± 18 | 83 ± 20 | 0.04 |
| Systolic BP (mmHg) | 140 ± 27 | 141 ± 27 | 138 ± 28 | 0.25 |
| Creatinine (mg/dL) | 0.9 [0.8;1.1] | 0.9 [0.8;1.1] | 1.0 [0.8;1.4] | < 0.0001 |
| eGFR (mL/min/1.73 m 2 ) | 83 ± 31 | 88 ± 29 | 69 ± 31 | < 0.0001 |
| Haemoglobin (g/dL) | 13 ± 2 | 14 ± 1 | 11 ± 1 | < 0.0001 |
| Anaemia | 305 (29) | |||
| Peak cardiac troponin I (ng/mL) | 8 [1;47] | 9 [1;49] | 6 [1;40] | 0.004 |
| Prior use of aspirin | 281 (27) | 165 (22) | 116 (38) | < 0.0001 |
| Prior use of clopidogrel | 157 (15) | 80 (11) | 77 (26) | < 0.0001 |
| Coronary angiography | 972 (92) | 728 (96) | 244 (80) | < 0.0001 |
| Multivessel CAD | 196 (20) | 131 (18) | 65 (25) | 0.01 |
| Left main CAD | 60 (6) | 38 (5) | 22 (9) | 0.055 |
| Primary angioplasty | 482 (48) | 362 (50) | 120 (45) | 0.23 |
| Fibrinolytic therapy | 122 (12) | 101 (13) | 21 (7) | 0.003 |
| Glycoprotein IIb/IIIa blocker | 400 (38) | 316 (42) | 84 (28) | < 0.0001 |
| CABG surgery | 66 (7) | 46 (6) | 20 (7) | 0.52 |
| Aspirin at discharge | 1027 (98) | 740 (99) | 287 (96) | 0.002 |
| Clopidogrel at discharge | 950 (91) | 694 (93) | 256 (86) | 0.0006 |
| GRACE score – death or MI | 25 [17;34] | 23 [16;31] | 32 [22;40] | < 0.0001 |
| Events | 132 (12.4) | 62 (8.2) | 70 (22.9) | |
| 6-month event-free survival (%) | 87.6 [85.5;89.4] | 91.8 [89.6;93.6] | 77.0 [71.9;81.4] | < 0.0001 |
| 30-day event-free survival (%) | 93.2 [91.6;94.6] | 95.0 [93.2;96.3] | 88.9 [84.8;91.9] |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
