Respiratory rate predicts outcome after acute ...

CLINICAL RESEARCH

European Heart Journal (2013) 34, 1644–1650 doi:10.1093/eurheartj/ehs420

Acute coronary syndromes

Respiratory rate predicts outcome after acute myocardial infarction: a prospective cohort study

1 Medizinische Klinik und Deutsches Herzzentrum Mu¨nchen der Technischen Universita¨t Mu¨nchen, Munich, Germany; 2International Centre for Circulatory Health, National Heart and Lung Institute, Imperial College, London, UK; 3Innere Medizin III, Eberhard-Karls-Universita¨t, Tu¨bingen, Germany; 4Institut fu¨r Medizinische Statistik und Epidemiologie der Technischen Universita¨t Mu¨nchen, Munich, Germany; and 5St Paul’s Cardiac Electrophysiology and St George’s, University of London, London, UK

Received 21 July 2012; revised 30 October 2012; accepted 13 November 2012; online publish-ahead-of-print 14 December 2012

Aims

Risk stratification after acute myocardial infarction (MI) remains imperfect and new indices are sought that might improve the post-MI risk assessment. In a contemporarily-treated cohort of acute MI patients, we tested whether the respiratory rate provides prognostic information and how this information compares to that of established risk assessment. ..................................................................................................................................................................................... Methods A total of 941 consecutive patients (mean age 61 years, 19% female) presenting with acute MI were enrolled between May 2000 and March 2005. The last follow-up was performed May 2010. Main outcome measure was total mortality and results during a follow-up period of 5 years. Patients underwent 10-min resting recordings of the respiratory rate within 2 weeks after MI in addition to the measurement of the left ventricular ejection fraction (LVEF) and standard clinical assessment including the GRACE score. During the follow-up, 72 patients died. The respiratory rate was a significant predictor of death in univariable analysis (hazard ratio 1.19 per 1/min, 95% confidence interval: 1.12–1.27) as was the GRACE score [1.04 (1.03– 1.05) per point], LVEF [0.96 (0.94– 0.97) per 1%], and the diagnosis of diabetes mellitus [2.78 (1.73 –4.47)], all P , 0.0001. On multivariate analysis, the GRACE score (P , 0.0001), respiratory rate (P , 0.0001), LVEF (P ¼ 0.013), and diabetes (P ¼ 0.016) were independent prognostic markers. ..................................................................................................................................................................................... Conclusion The respiratory rate provides powerful prognostic information which is independent and complementary to that of existing risk assessment. Simple and inexpensive assessment of the respiratory rate should be considered a complementary variable for the assessment of risk after acute MI.

----------------------------------------------------------------------------------------------------------------------------------------------------------Keywords

Myocardial infarction † Respiration rate † Risk assessment † GRACE score † Left ventricular ejection fraction

Introduction After acute myocardial infarction (MI), mortality risk of an individual patient can be estimated at the point of care by clinical risk scores that include variables showing evidence of MI, e.g. ST-segment deviation and elevated cardiac enzymes, co-morbidities, e.g. renal impairment, or abnormal physiology, e.g. systolic blood pressure and heart rate.1 – 3 The degree of left ventricular impairment by the measurement of the left ventricular ejection fraction (LVEF) is used to select patients suitable for ICD implantation.4,5 Although

LVEF measurement involves a substantial investment of specialist time and care, it is neither sensitive nor specific.6 – 8 Prognostic evaluation is far from perfect, and therefore a search is ongoing for risk markers complementary to currently applied risk scores and LVEF, preferably not imposing undue demands on equipment and/or specialist time.9 Changes in respiratory control due to cardiovascular dysfunction are known to indicate worsened prognosis. For example, chemoreflex hypersensitivity or enhanced ventilatory responses to exercise indicate failure of the auto-regulatory processes of

* Corresponding author: SPCE, Verulam Avenue, Purley, Surrey CR8 3NQ, UK. Tel: +44 20 8660 2112; Fax: +44 20 8660 6031, Email: [email protected] †

Equal contributions as first authors.

‡

Equal contributions as senior authors.

Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2012. For permissions please email: [email protected]

Downloaded from http://eurheartj.oxfordjournals.org/ at Bibliothek der Technischen Universitaet Muenchen on June 19, 2013

Petra Barthel 1†, Roland Wensel 2†, Axel Bauer 1,3, Alexander Mu¨ller 1, Petra Wolf4, Kurt Ulm 4, Katharina M. Huster 1, Darrel P. Francis 2, Marek Malik 5*‡, and Georg Schmidt 1‡

1645

Respiratory rate after acute myocardial infarction

Methods Study cohort This study was conducted at two University hospitals of the Technische Universita¨t Mu¨nchen, the German Heart Centre, and the Klinikum Rechts der Isar, both in Munich, Germany. Subjects were prospectively enrolled between March 2000 and May 2005, the last follow-up was performed on May 2010. Eligible acute MI patients were aged ≤80 years, presented in sinus rhythm, survived till hospital discharge, and did not meet criteria for immediate implantation of a cardioverter-defibrillator.1 Acute myocardial infarction was diagnosed if a patient had at least two of the following findings: chest pain for ≥20 min, creatine kinase.twice the upper limit of normal, and ST-elevation ≥0.1 mV in two or more limb leads or ≥0.2 mV in two or more contiguous precordial leads on admission.14 The study protocol was approved by the local ethics committee and written consent was obtained.

Measurements Respiratory rate Within the first 2 weeks after index MI, all patients underwent three series of 10-min measurement of the respiratory rate under standardized conditions. Patients were studied in the morning in supine position without interruption of their normal medication regimen. A piezoelectric thoracic sensor (Pro-Tech) was attached, and the signal acquired at 1.6 kHz was analysed automatically (Porti, TMS International, The Netherlands). To avoid disturbing the spontaneous pattern of breathing, patients were not instructed that the sensor was attached to monitor respiration (although they initially gave informed consent to monitoring of ECG, arterial pressure, and respiratory rate as part of the protocol). The last of the repeated 10-min recordings of the respiration rate were used for the measurements used in the study.

Left ventricular ejection fraction Left ventricular ejection fraction was assessed by angiography (n ¼ 445; 47.3%) or biplane echocardiography according to Simpson’s method (n ¼ 496; 52.7%; Sonos 5500, Hewlett Packard) within the first 2 weeks [median 7 days, inter-quartile range (IQR) 5 – 9 days] after the index MI.

Clinical variables In addition to the respiratory rate and the LVEF, we also assessed the diagnosis of diabetes mellitus and a point-of-care clinical evaluation. Diabetes was considered present if a patient was already diagnosed and was receiving treatment (diet, tablets, or insulin) or if fasting blood glucose concentration repeatedly exceeded 11 mmol/L.15 Of the previously suggested clinical assessment scores, we selected the GRACE score proposed for the prediction of long-term prognosis.3 This includes nine elements (age of the patient, a history of past heart failure, a history of past MI, serum creatinine at admission, the cardiac biomarker status at admission, systolic blood pressure at admission, pulse at admission, ST-deviation at admission, and in-hospital percutaneous coronary intervention). The presence of chronic obstructive pulmonary disease was defined as the necessity of chronic treatment with bronchodilators and/or inhalation glucocorticoids.

Endpoints Patients were followed up with clinical appointments every 6 months. If patients did not attend, they were contacted by letter, by telephone or through their General Practitioner. If this was not successful, the local population registry office was contacted to trace their new address or to identify that they had died. The endpoint of the study was total mortality during 5 years of the index MI.

Evaluation of risk predictors In the primary analysis, the respiratory rate, LVEF, and GRACE score were treated as continuous variables. In secondary analysis, the respiratory rate was prospectively dichotomized at 20 cycles per minute,17 and the LVEF was prospectively dichotomized at 35%.6 The GRACE score’s dichotomy was set at 120, optimizing the separation between high-and low-risk cases (log-rank optimization) as no independent guidance for prospective dichotomy is available.

Statistics The study was planned and powered to evaluate the influence of clinical variables on mortality. Applying the recommendation by Peduzzi et al.16 that at least 10 events are needed per variable, the study was continued until 70 events were observed. This also allowed considering some elements of the GRACE score separately (analysis not presented here). Continuous variables are presented as median and IQR. Categorical data are expressed as absolute frequencies and percentages. A Cox proportional hazard model was used that included all variables at one time (enter procedure) to assess independence of prognostic capability of mortality predictors. The proportional hazards assumption was investigated using Schoenfeld residuals.18 Survival curves were estimated by the Kaplan– Meier method and compared using the log-rank test. Continuous net reclassification improvement (NRI) and integrated discrimination improvement (IDI) were calculated19 together with the change of the area under the receiver operator characteristic (AUC ROC) for the addition of the respiratory rate to the other variables considered. Differences in the AUC were tested using bootstrap methods. Differences were considered statistically significant if P , 0.05. Reproducibility of measurements of the respiratory rate was assessed by calculating the intra-class correlation coefficient of the three measurements. The regression analysis results were validated by internal bootstrap.20 IBM SPSS Statistics 19.0, SPSS, Inc. statistical package and R 2.15.1 (R Foundation for Statistical Computing, Vienna, Austria) were used in the calculations.


the cardiopulmonary system and may serve diagnosis in chronic heart failure.10 – 12 However, the quantification of these reflex abnormalities requires complex equipment, trained staff, and patient co-operation. It is thus unlikely suitable for widespread application. The most elementary measure of respiratory reaction to cardiovascular disturbance is the elevation in the respiratory rate, which is very easy to measure. An association between respiratory rate and post-MI prognosis has been described well before the contemporary treatment of acute MI.13 However, this observation is not in routinely used in clinical practice and its validity in modern practice is unknown. In this study, we examined whether an increased respiratory rate, measured under standardized conditions before hospital discharge, still contains clinically useful prognostic information in MI patients. The study was a prospective observational evaluation of the Autonomic Regulation Trial (ART).

1646

Results

Comparison of continuous variables On both, univariable and multivariable analyses, the respiratory rate, the GRACE score, LVEF, chronic obstructive pulmonary disease, and diabetic status were all statistically associated with the outcome (Table 2 and Supplementary material online). The internal bootstrap validation provided statistically significant results (P , 0.001). An increased respiratory rate was continuously related to worsening prognosis, with a doubling of mortality for every four-breath increment in the respiratory rate (Figure 2). Intra-class correlation coefficient of the three serial measurements of the respiratory rate was 0.9, indicating a high degree of reproducibility.

Figure 1 Flow chart of patient selection.

Analyses by ROC AUC, NRI, and IDI showed increases in the predictive ability of the respiratory rate; IDI increased by 2.12% (P ¼ 0.004) for non-surviving patients, and by 0.24% (P ¼ 0.081) for survivors. Net reclassification improvement improved in 50.8% of the cases: in 25.1% of patients who died, and in 25.7% of survivors. The ROC AUC increased from 0.79 to 0.82 (P ¼ 0.12).

Subgroup analysis Of the total population, 335 patients (36%) had the GRACE score .120, 233 (25%) patients had the respiration rate .20 cycles per minute, and 87 patients (9%) had the LVEF ,35%. The left panel of Figure 3 shows that in patients with GRACE ,120, no meaningful risk stratification is provided by the LVEF, respiratory rate, or their combination. All the survival curves in this panel of Figure 3 show 5-year mortality ,5% and are not statistically different from each other (P ¼ 0.936). On the contrary, the right panel of Figure 3 shows that in patients with the GRACE score ≥120; both the LVEF and the respiratory rate provide further meaningful risk prediction. The differences between the survival curves shown in the panel are highly statistically significant (P , 0.0001). Among these patients, those with both, LVEF ≤35% and respiration rate ≥20 breaths/ min, are at the greatest risk with 5-year mortality exceeding 50%. Patients with either LVEF ≤35% or respiration rate ≥ 20 breaths/min have practically the same intermediate risk of 5-year mortality 25% although the survival profile over the observation period is different. Interestingly, patients with the GRACE score ≥120 but both LVEF .35% and respiration rate ,20 breaths/min (who constitute 65% of all patients with the


During the recruitment period, a total of 1937 patients hospitalized for acute MI were screened. The inclusion criteria were met by 1511 patients. Out of these, 542 patients declined to participate. In 28 patients who initially consented, the protocol was not carried out for various reasons (e.g. withdrawal of consent, technical defects, early discharge, or transfer to another hospital). The remaining 941 patients constituted the study cohort (Figure 1). Their clinical and demographic characteristics are shown in Table 1 (for further details, see Supplementary material online). Most patients received percutaneous coronary intervention as acute MI therapy and were treated with beta-blockers, angiotensin-converting enzyme inhibitors, statins, and aspirin. Eleven of the 941 patients (1.2%) were lost to follow-up and were censored at the time of last contact. At 5 years of follow-up, 72 (7.7%) patients had died.

P. Barthel et al.

1647


Table 1

Clinical characteristics of the study cohort

Variable

Study population (n 5 941)

................................................................................ Age (years), median (IQR)

61 (52–69) 182 (19.3)

Diabetes mellitus, n (%) History of previous MI, n (%)

184 (19.6) 90 (9.6)

Hypertension, n (%)

682 (72.5)

Smoking, n (%) Chronic obstructive pulmonary disease, n (%) CK max (U/l), median (IQR)

488 (51.9) 39 (4.1)

LVEF (%), median (IQR)

1302 (647–2465) 53 (45–60)

................................................................................ MI localization Anterior, n (%) Posterior, n (%)

391 (41.6) 435 (46.2)

Lateral, n (%)

102 (10.8)

Unclassified, n (%)

12 (1.3)

................................................................................ BMI (kg/m2), median (IQR) Serum creatinine (mg/dL), median (IQR)

27 (24–29) 1.1 (0.9– 1.3)

Cardiogenic shock/CPR, n (%)

41 (4.4)

................................................................................ Intervention PCI, n (%)

878 (93.3)

Thrombolysis, n (%)

14 (1.5)

CABG, n (%) No revascularization possible, n (%)

6 (0.6) 43 (4.6)

................................................................................ Aspirin, n (%)

913 (97.0)

Clopidogrel, n (%)

920 (97.8)

Beta-blockers, n (%) ACE-inhibitors, n (%)

897 (95.3) 885 (94.0)

Statins, n (%)

879 (93.4)

Diuretics, n (%)

415 (44.1)

MI, myocardial infarction; CK, serum creatine kinase; IQR, inter-quartile range; LVEF, left ventricular ejection fraction; BMI, body mass index; CPR, cardiopulmonary resuscitation; PCI percutaneous coronary intervention; CABG, coronary artery bypass graft; ACE, angiotensin-converting enzyme.

GRACE score ≥120) are only at relatively low risk not very different from those with the GRACE score ,120, with 5-year mortality ,10%. The tests for interaction in the Cox model were both significant (interaction respiration rate and LVEF: P ¼ 0.013, respiration rate and GRACE: P ¼ 0.014). Kaplan–Meier curves of patients with a GRACE score ≥120 stratified by a combination of the left ventricular function ≤35% and the respiratory rate ≥20 breaths/min are shown in the Supplementary material online. Analyses by ROC AUC, NRI, and IDI showed all significant increases in the predictive ability of the respiratory rate: IDI increased by 4.26% (P ¼ 0.007) for non-surviving patients and by 1.25% (P ¼ 0.012) for survivors. Net reclassification improvement improved in 58.4% of the cases: in 25.9% of patients who died, and in 32.4% of survivors. The ROC AUC increased from 0.71 to 0.78 (P ¼ 0.004).

Figure 2 The effect of the respiratory rate on mortality. Mortality risk is shown together with 95% confidence intervals.

Discussion The study shows that the measurement of resting respiratory rate under standardized conditions provides important prognostic information in patients with contemporary post-MI treatment. This prognostic information is independent of and complementary to that of established risk assessment including left ventricular ejection fraction and point-of-care-risk assessment. The respiratory rate is easy and inexpensive to measure and—if performed under controlled conditions—highly reproducible. There are no technical barriers to its incorporation as part of immediate point-of-care-risk assessment. The study measured the respiratory rate shortly prior to hospital discharge. The protocol was designed in this way since earlier measurement (e.g. after admission) might have interfered with clinical care in the acute MI phase. In designing the protocol, we also hypothesized that the measurement of the respiratory rate under controlled conditions just before hospital discharge would be most predictive of future risk.

Respiratory responses in heart disease There is a body of evidence that an abnormal respiratory response in heart disease is an indicator of adverse prognosis.11,12 The respiratory response can be quantified using a variety of stimuli. Patients with the most abnormal ventilatory responses during exercise stress testing have a poor prognosis, even when limiting the analysis to those with preserved exercise capacity.12 Abnormal responses can also be elicited without physical exercise by the stimulation of the chemoreflexes that raise ventilation when carbon dioxide levels rise or oxygen levels fall. However, despite clear evidence from these reflex measurements, these tests are unlikely to become clinical routine in post-MI patients because they require significant specialist involvement and equipment, as well as substantial patient co-operation. Our prospective data collection protocol did not include serum creatinine assessment on the same day as the respiratory frequency was measured (this was available in slightly less than two-thirds of the patients). When using serum creatinine data


Females, n (%)

1648

Table 2

P. Barthel et al.

Univariable and multivariable Cox regression analysis

Variable

Univariable

.................................................... HR

x

2

P-value

Multivariable

....................................................

HR

x2

P-value

............................................................................................................................................................................... Respiratory rate (per 1/min)

1.19 (1.12– 1.27)

30.4

,0.0001

1.14 (1.07–1.22)

16.5

,0.0001

LVEF (per 1%)

0.96 (0.94– 0.97)

24.2

,0.0001

0.98 (0.96–1.00)

5.4

0.020

Diabetes mellitus (n/y) GRACE score (per 1 point)

2.78 (1.73– 4.47) 1.04 (1.03– 1.05)

17.9 65.0

,0.0001 ,0.0001

1.67 (1.03–2.72) 1.03 (1.02–1.04)

4.3 35.9

0.038 ,0.0001

Chronic obstructive pulmonary disease (n/y)

4.31 (2.21– 8.41)

18.4

,0.0001

2.67 (1.35–5.29)

8.0

0.005

Figure 3 Cumulative rates of deaths, in patients with a GRACE score of ,120 (left panel) and ≥120 (right panel) stratified according to left ventricular function (LVEF ≤35 vs. LVEF .35%), and respiratory rate (≥20 breaths/min vs. ,20 breaths/min). The numbers of patients of the individual groups involved in the analysis at 0, 1, 2, 3, 4, and 5 years are shown below each graph; the order of the rows corresponds to the order of the mortality curves. Test by log-rank statistics.

from up to +3 days of respiration measurement, we were able to add the estimated glomerular filtration rate21 to the Cox regression model. This weakened the multivariable contribution of the GRACE score (x2 statistics was lowered to 22.7) but the contribution of the respiratory rate remained unchanged. Adding medication to the model had no further effect (for details, see supplementary material online).

Pathophysiology of abnormal respiratory responses Circulatory failure can affect the physiology of respiration by altering both ventilatory control and pulmonary perfusion. An increase in left ventricular filling pressure (and hence in pulmonary capillary pressure) affects—among others—pulmonary compliance and alveolocapillary membrane conductance. These alterations raise the respiratory rate (rather than tidal volume),22 which can be

detected using any of a wide spectrum of monitoring systems. This simple count of the respiratory rate was first identified to be a prognostic marker after MI before the advent of routine echocardiography, and of modern medical and interventional therapy.13 Our data show that the prognostic value of the respiratory rate remains undiminished in the era of modern acute treatment. It is well established that frank pulmonary oedema is a poor prognostic sign in acute MI.23 In this study, however, only eight patients had clinically overt pulmonary congestion. Repeating the analyses after exclusion of these patients revealed essentially identical results (data not shown). Subclinical pulmonary congestion, and abnormally enhanced physiological responses to it, may well be a forme fruste of pulmonary oedema. There may also be a contribution from the profound disturbance of cardiac autonomic control seen in acute MI, with increased sympathetic activity and vagal withdrawal which could further interfere with ventilatory control. Observations from


All x2 values are distributed with 1 degree of freedom. LVEF, left ventricular ejection fraction; GRACE score, composite of age of the patient, serum creatinine, history of a past MI, history of past congestive heart failure, in-hospital percutaneous coronary intervention, pulse, systolic blood pressure, ST-segment deviation and positive enzymes; COPD, chronic obstructive pulmonary disease.

1649


patients with chronic heart failure show increased ventilation as a result of increased chemoreflex gain and peripheral stimuli,24,25 which signify poor prognosis.11,12 These changes are linked to cardiac autonomic control and it is conceivable that similar changes might occur acutely after MI. Increase in the respiratory rate might therefore also be a reflection of disturbed autonomic control. Nevertheless, regardless of the precise mechanism, the respiratory rate is arguably the simplest physiological variable of all to measure, requiring only standard medical equipment.

The prognostic strengths combined with simplicity of measurement make the respiratory rate a strong candidate to include in risk-assessment algorithms. The subgroup analysis of our data suggests that the respiratory rate can complement the LVEF, for prognostic assessment especially in patients with abnormal clinical scores. In our data, the patients with an abnormal GRACE score represented 35% of all post-MI patients of whom 30% presented with an increased respiratory rate. On the contrary, in our patients with a low to moderate GRACE score (65% of the population), the mortality risk was very low and no meaningful improvement in risk prediction was achieved by considering respiration rate, LVEF, or their combination. Results of single risk stratification studies might serve as a starting point of randomized interventional studies which are needed to prove the appropriateness of any general proposal of changes in clinical practice. At the same time, risk assessment on its own can and should contribute to individualized medicine. Thus, while the standard clinical procedures should presently be used in patients with an elevated respiratory rate, their scheduling, combination, and intensity should be tailored to the increased risk. These patients should receive closer clinical attention with more densely scheduled diagnostic procedures, more frequent follow-up assessments, and increased intensity of therapy control.

Study limitations This study only evaluated the respiratory rate and not ventilation or tidal volume. The rationale for this restriction was three-fold. First, ventilation it primarily altered by the respiratory rate and not tidal volume in heart disease.22 Secondly, the measurement of tidal volume requires some element of specialized equipment which may be difficult to provide and therefore limit universal applicability in routine clinical settings. Thirdly, although our patients were generally aware that their physiology including respiration was monitored, they were put to no significant inconvenience by that measurement and did not have specific attention drawn to the time window when the quantification of the respiratory rate was made. This would not be possible with tidal volume assessment, possibly influencing the natural character of measured data. As detailed lung function measurements were not prospectively collected, we were unable to classify lung disease beyond the distinction of patients with chronic obstructive pulmonary disease. Also, we have not collected comparable data of the respiratory rate on admission since the detailed protocol on measuring the respiration rate during undisturbed supine rest would be difficult if not impossible to include with the admission tests. We are

Conclusion After MI, patients with an elevated respiratory rate have a significantly worse survival. A four-breath-per-minute increase in the respiratory rate indicates a doubling of risk. The prognostic information of the respiratory rate is independent to those of established risk assessments models. In patients with high-pointof-care-risk scores (e.g. such as the GRACE score), the respiratory rate allows further risk stratification. As one of the simplest, quickest and most inexpensive of physiological variables to measure, the independent prognostic power of the respiratory rate commends it for consideration for incorporation within established risk-assessment strategies for patients after MI even in the era of modern investigations and treatment strategies. In conjunction with clinical assessment and LVEF, it allows identifying patients with truly benign prognosis as well as those at substantial risk of death.

Supplementary material Supplementary material is available at European Heart Journal online.

Funding This work was supported by the Bundesministerium fu¨r Bildung, Wissenschaft, Forschung und Technologie (13N/7073/7), the Kommission fu¨r Klinische Forschung, the Deutsche Forschungsgemeinschaft (SFB 386), the British Heart Foundation (FS/10/038), and the NIHR Biomedical Research Centre Scheme. These sponsors had no role in study design, data collection, data analysis, data interpretation, or writing of the report. Conflict of interest: none declared.

References 1. Fox KAA, Dabbous OH, Goldberg RJ, Pieper KS, Eagle KA, Van de Werf F, Avezum A, Goodman SG, Flather MD, Anderson FA Jr,, Granger CB. Prediction of risk of death and myocardial infarction in the six months after presentation with acute coronary syndrome: prospective multinational observational study (GRACE). BMJ 2006;333:1091 –1094. 2. Morrow DA, Antman EM, Charlesworth A, Cairns R, Murphy SA, de Lemos JA, Giugliano RP, McCabe CH, Braunwald E. TIMI risk score for ST-elevation myocardial infarction: a convenient, bedside, clinical score for risk assessment at presentation. An intravenous nPA for treatment of infarcting myocardium early II trial substudy. Circulation 2000;102:2031 –2037. 3. Eagle KA, Lim MJ, Dabbous OH, Pieper KS, Goldberg RJ, Van de Werf F, Goodman SG, Granger CB, Steg PG, Gore JM, Budaj A, Avezum A, Flather MD, Fox KAA, for the GRACE Investigators. A validated prediction model for all forms of acute coronary syndrome: estimating the risk of


Clinical implications

thus unable to comment on the development of the respiration rate during hospital stay and of its prognostic strength. Our study cannot confirm the precise mechanism by which the increase in mortality is associated with higher respiratory rates. Nevertheless the effect is large: indeed comparing well in magnitude to the effects of the GRACE score and of low ejection fraction. Finally, per study protocol, the age of investigated patients was not exceeding 80 years. We are unable to comment on any extrapolation of the results to older patients especially since pulmonary disease might be more prevalent in this age group.

1650

4.

5.

6.

8. 9. 10.

11.

12.

13. Helmers C. Assessment of 3-year prognosis in survivors of acute myocardial infarction. Br Heart J 1975;37:593 –597. 14. Nomenclature and criteria for diagnosis of ischemic heart disease. Report of the joint international society and federation of cardiology/world health organization task force on standardization of clinical nomenclature. Circulation 1979; 59:607 – 609. 15. Barthel P, Bauer A, Mu¨ller A, Junk N, Huster KM, Ulm K, Malik M, Schmidt G. Reflex and tonic autonomic markers for risk stratification in patients with type 2 diabetes surviving acute myocardial infarction. Diabetes Care 2011; 34:1833 –1837. 16. Peduzzi P, Concato J, Feinstein AR, Holford TR. Importance of events per independent variable in proportional hazards regression analysis II. Accuracy and precision of regression estimates. J Clin Epidemiol 1995;48:1503 –1510. 17. Walker HK, Hall WD, Hurst JW, eds. Clinical Methods: The History, Physical, and Laboratory Examinations. 3rd ed., Boston: Butterworths; 1990. 18. Schoenfeld D. Residuals for the proportional hazards regression model. Biometrika 1982;69:239 –241. 19. Pencina MJ, D’Agostino RB Sr, Steyerberg EW. Extensions of net reclassification improvement calculations to measure usefulness of new biomarkers. Stat Med 2011;30:11– 21. 20. Steyerberg EW, Harrell FE Jr, Borsboom GJ, Eijkemans MJ, Vergouwe Y, Habbema JD. Internal validation of predictive models: efficiency of some procedures for logistic regression analysis. J Clin Epidemiol 2001;54:774 –781. 21. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. Modification of Diet in Renal Disease Study Group. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Ann Intern Med 1999;130:461 –470. 22. Agostoni P, Cattadori G, Bussotti M, Apostolo A. Cardiopulmonary interaction in heart failure. Pulm Pharmacol Ther 2007;20:130 –134. 23. Killip T III, Kimball JT. Treatment of myocardial infarction in a coronary care unit. A two year experience with 250 patients. Am J Cardiol 1967;20:457–464. 24. Chua TP, Clark AL, Amadi AA, Coats AJ. Relation between chemosensitivity and the ventilatory response to exercise in chronic heart failure. J Am Coll Cardiol 1996;27:650 –657. 25. Piepoli M, Clark AL, Volterrani M, Adamopoulos S, Sleight P, Coats AJ. Contribution of muscle afferents to the hemodynamic, autonomic, and ventilatory responses to exercise in patients with chronic heart failure: effects of physical training. Circulation 1996;93:940 –952.


7.

6-month postdischarge death in an international registry. J Am Med Assoc 2004; 291:2727 –2733. Bardy GH, Lee KL, Mark DB, Poole JE, Packer DL, Boineau R, Domanski M, Troutman C, Anderson J, Johnson G, McNulty SE, Clapp-Channing N, Davidson-Ray LD, Fraulo ES, Fishbein DP, Luceri RM, Ip JH. Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) Investigators. Amiodarone or an implantable cardioverter-defibrillator for congestive heart failure. N Engl J Med 2005; 352:225–237. Moss AJ, Zareba W, Hall WJ, Klein H, Wilber DJ, Cannom DS, Daubert JP, Higgins SL, Brown MW, Andrews ML; Multicenter Automatic Defibrillator Implantation Trial II Investigators. Prophylactic implantation of a defibrillator in patients with myocardial infarction and reduced ejection fraction. N Engl J Med 2002;346:877 –883. Gregoratos G, Abrams J, Epstein AE, Freedman RA, Hayes DL, Hlatky MA, Kerber RE, Naccarelli GV, Schoenfeld MH, Silka MJ, Winters SL, Gibbons RI, Antman EM, Alpert JS, Hiratzka LF, Faxon DP, Jacobs AK, Fuster V, Smith SC Jr. ACC/AHA/NASPE 2002 Guideline update for implantation of cardiac pacemakers and antiarrhythmia devices—summary article: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (ACC/AHA/NASPE Committee to Update the 1998 Pacemaker Guidelines). J Am Coll Cardiol 2002;40:1703 –1719. Buxton AE. Risk stratification for sudden death: do we need anything more than ejection fraction? Card Electrophysiol Rev 2003;7:434 –437. Huikuri HV, Castellanos A, Myerburg RJ. Sudden death due to cardiac arrhythmias. N Engl J Med 2001;345:1473 –1482. Malik M. Autonomic tests to detect cardiac risk and their clinical practicality. J Cardiovasc Electrophysiol 2011;22:128 –130. Ponikowski PP, Chua TP, Francis DP, Capucci A, Coats AJ, Piepoli MF. Muscle ergoreceptor overactivity reflects deterioration in clinical status and cardiorespiratory reflex control in chronic heart failure. Circulation 2001;104:2324 –2330. Ponikowski P, Chua TP, Anker SD, Francis DP, Doehner W, Banasiak W, Poole-Wilson PA, Piepoli MF, Coats AJ. Peripheral chemoreceptor hypersensitivity: an ominous sign in patients with chronic heart failure. Circulation 2001;104: 544 –549. Ponikowski P, Francis DP, Piepoli MF, Davies LC, Chua TP, Davos CH, Florea V, Banasiak W, Poole-Wilson PA, Coats AJ, Anker SD. Enhanced ventilatory response to exercise in patients with chronic heart failure and preserved exercise tolerance: marker of abnormal cardiorespiratory reflex control and predictor of poor prognosis. Circulation 2001;103:967 –972.

P. Barthel et al.