Intern Emerg Med DOI 10.1007/s11739-013-1032-9
IM - ORIGINAL
The previous use of digoxin does not worsen early outcome of acute coronary syndromes: an analysis of the ARIAM Registry Juan Carlos Garcia-Rubira • Manuel Calvo-Taracido • Francisca Francisco-Aparicio • Manuel Almendro-Delia • Alejandro Recio-Mayoral • Antonio Reina Toral • Oscar Aramburu-Bodas • Pastora Gallego Garcı´a de Vinuesa • Jose´ Maria Cruz Ferna´ndez • Angel Garcia Alca´ntara • Rafael Hidalgo-Urbano
Received: 18 October 2013 / Accepted: 9 December 2013 Ó SIMI 2013
Abstract The aim of the study was to determine the influence of the previous use of digoxin on the hospital mortality and complications of patients admitted because of acute coronary syndrome (ACS). We analyzed the data of patients included in the ARIAM-Andalucia Registry, which involves 49 hospitals in Andalucia, Spain, from 2007 to 2012. Patients on digoxin treatment prior to their admission because of ACS constituted the digoxin group (DG), and were compared with the group of patients not on digoxin. Logistic regression and propensity score matching On behalf of the ARIAM Andalucı´a Study Group, the list of investigators is given in ESM Appendix.
Electronic supplementary material The online version of this article (doi:10.1007/s11739-013-1032-9) contains supplementary material, which is available to authorized users. J. C. Garcia-Rubira M. Calvo-Taracido F. Francisco-Aparicio M. Almendro-Delia A. Recio-Mayoral P. Gallego Garcı´a de Vinuesa J. M. Cruz Ferna´ndez R. Hidalgo-Urbano Cardiology Department, Virgen Macarena University Hospital, Seville, Spain J. C. Garcia-Rubira (&) Unidad Coronaria, Hospital Virgen Macarena, Avenida Dr Fedriani, sn, 41008 Seville, Spain e-mail: [email protected] A. Reina Toral Intensive Care Unit, Virgen de las Nieves Hospital, Granada, Spain O. Aramburu-Bodas Internal Medicine Department, Virgen Macarena University Hospital, Seville, Spain A. G. Alca´ntara Intensive Care Unit, Virgen de la Victoria Hospital, Ma´laga, Spain
were used to analyze the differences. We included 20,331 patients, of whom 244 (1.2 %) were on digoxin. DG patients were older (73.1 vs 63.7 years old), more often women, and had more diabetes, hypertension, previous myocardial infarction, heart failure, stroke, atrial fibrillation, peripheral vascular disease, obstructive pulmonary disease or kidney disease. On univariate analysis, DG patients had significantly higher hospital mortality (13.5 vs 5.3 % P \ 0.001), and more cardiogenic shock, but less ventricular fibrillation, and no differences in atrioventricular block, stroke or reinfarction. After the multivariate analysis, DG had no significant influence on hospital prognosis [odds ratio (OR) 1.21, 95 % confidence interval 0.79–1.86]. The analysis of a propensity-matched cohort of 464 patients (232 DG and 232 NoDG) did not find differences in hospital mortality (13.4 vs 13.4 %) nor other complications. In our cohort of ACS patients, the previous treatment with digoxin was not associated with an increase in dysrhythmic complications nor was an independent predictor of mortality during hospitalization. Keywords Acute coronary syndrome Atrial fibrillation Dysrhythmias Digoxin Heart failure Myocardial infarction List of abbreviations ACS Acute coronary syndrome AF Atrial fibrillation CI Confidence interval DG Patients on digoxin treatment previously to their admission MI Myocardial infarction NoDG Patients not on digoxin treatment previously to their admission NoSTE Acute coronary syndrome without ST elevation
123
Intern Emerg Med
OR STE
Odds ratio Acute coronary syndrome with ST elevation
mortality and complications of patients who are chronically taking digoxin, and who were subsequently admitted because of an acute coronary syndrome (ACS).
Introduction
Materials and methods
Years ago, digoxin along with diuretics, constituted the mainstay of treatment of patients with heart failure, and so remained for more than 200 years. Optimization of this therapeutic tool was encouraged, and relatively high doses, aiming to reach the higher half of the therapeutic range were usually recommended [1]. However, digoxin has a distinctive pharmacology, with an elimination half-life of 36–48 h when renal function is normal. The excretion is renal, and proportional to the glomerular filtration rate. It has interactions with other drugs and electrolyte disturbances easily enhance toxicity [2]. Early in the 70s, digitalis toxicity became an important issue, with the development of strategies for preventing and treating overdose of this drug [3]. Even today physicians must be aware of the potential toxicity in patients taking digoxin [4]. Nevertheless, digoxin is a very cheap drug that still has a role in the management of heart failure and atrial fibrillation, as it is recommended by some guidelines [5]. More recent statements in the management of symptomatic chronic heart failure have relegated the use of digoxin to the last position, only after the treatment with diuretics, betablockers, angiotensin-converting enzyme, angiotensin receptor blocker, mineralocorticoid receptor antagonist, and ivabradine, have failed. It is recommended earlier in patients with atrial fibrillation, but as a second-line drug in chronic heart failure, if heart rate control is not gained with either betablocker or calcium antagonists (in preserved systolic function) [6]. Digoxin is still one of the drugs recommended for controlling the heart rate in atrial fibrillation, although caution is warranted because of potential life threatening adverse effects. It is specially indicated by the European Task Force in patients with heart failure and left ventricular dysfunction [7]. In the scenario of ST elevation myocardial infarction, current guidelines of the European Society of Cardiology recommend the administration of digoxin with or without amiodarone, for heart rate control in patients who develop atrial fibrillation and acute heart failure or hypotension [8]. Together with these restrictive indications, a recent publication based on the data from the AFFIRM study calls attention to the finding of an increased mortality among patients taking digoxin [9]. This statement could be the final chapter in the history of digoxin, our old companion in heart failure! Our study attempts to shed a little more of light over the ‘‘digoxin question’’. We analyze the hospital
The ARIAM-Andalucia (Ana´lisis del Retraso en el Infarto Agudo de Miocardio en Andalucı´a) is a registry of patients with acute coronary syndrome admitted in the coronary units of 49 hospitals in Andalucia, Spain. This is an institutional project, aimed to improve the quality of treatment of ischemic heart disease, and include the patients admitted to the Intensive Care Units of these hospitals because of ACS. Each hospital introduced their patients through a Web interface, developed by Coresoft (http://www.cor esoft.es), which was accessed through the Internet (http:// www.ariam-andalucia.org). The data were subsequently reviewed and saved in a central unit. For this study, we performed a retrospective analysis with available data from the registry. The ARIAM-Andalucia virtual platform has become a global project, implemented in the Andalusian Public Health Service, partially financed by Health Counseling since 2005, which also includes other registries, such as adult cardiac surgery, cardiac arrests, or pediatric cardiac surgery. The basis of these records was published in an official gazette [10]. For the purpose of our present study, we have made a retrospective analysis of the data from January 2007 to January 2012. The analysis was retrospective, but all the data have been collected prospectively. This investigation has been approved by the local ethics committees. Other details on the methodology of introduction of the data have been previously published [11– 13]. Previous atrial fibrillation is not defined in the ARIAM, so that we took the combined variable ‘‘previous arrhythmia or atrial fibrillation at admission’’ (AF) as a surrogate. Both groups were compared in demographic and clinical characteristics at admission, and the clinical course of their hospitalization. ACS was classified as ACS with ST elevation (STE) or ACS without ST elevation (NoSTE). An initial comparison was performed according to the previous history of treatment with digoxin (DG) or without it (NoDG). We performed the Student’s t test for comparison of quantitative variables (age and length of hospital stay), and Pearson’s Chi-square test for qualitative variables. A multivariate model by logistic regression with the hospital mortality as dependent variable was made including all characteristics significantly different between DG and NoDG on the univariate analysis. To validate the results of this analysis, we calculated a propensity score of being DG. The propensity score model (PS) for the previous treatment with digoxin was obtained by logistic
123
Intern Emerg Med Table 1 Clinical characteristics of the patients, comparing DG with NoDG patients, both in the total cohort and in the propensity-matched patients Total cohort (N = 20,331) Global
Propensity-matched cohort (N = 464)
DG (N = 244)
NoDG (N = 20,087)
P value
DG (N = 232)
NoDG (N = 232)
P value
Age
63.8 ± 12.7
73.1 ± 8.9
63.7 ± 12.7
\0.001
73.1 ± 8.8
73.0 ± 9.0
0.921
Woman
5,150 (25.3)
98 (40.2)
5,052 (25.2)
\0.001
90 (38.8)
94 (40.5)
0.704
Smokers
7,101 (34.9)
29 (11.9)
7,072 (35.2)
\0.001
29 (12.5)
21 (9.1)
0.231
Diabetes
6,516 (32.0)
125 (51.2)
6,391 (31.8)
\0.001
119 (51.3)
121 (52.2)
0.853
Hypercholesteremia
8,494 (41.8)
105 (43.0)
8,389 (41.8)
0.689
99 (42.7)
107 (46.1)
0.455
Hypertension
10,699 (52.6)
187 (76.6)
10,512 (52.3)
178 (76.7)
182 (78.4)
0.656
Obesity
3,341 (16.4)
36 (14.8)
3,305 (16.5)
0.476
35 (15.1)
41 (17.7)
0.452
Familiar history
2,021 (9.9)
9 (3.7)
2,012 (10.0)
0.001
9 (3.9)
4 (1.7)
0.160
Previous myocardial infarction
3,013 (14.8)
57 (23.4)
2,956 (14.7)
\0.001
56 (24.1)
62 (26.7)
0.522
Previous heart failure
753 (3.7)
68 (27.9)
685 (3.4)
\0.001
63 (27.2)
56 (24.1)
0.457
Previous stroke
1,057 (5.2)
33 (13.5)
1,024 (5.1)
\0.001
30 (12.9)
41 (17.7)
0.156
AF
1,972 (9.7)
194 (79.5)
1,778 (8.9)
\0.001
182 (78.4)
186 (80.2)
0.647
Peripheral vascular disease
907 (4.5)
22 (9.0)
885 (4.4)
21 (9.1)
18 (7.8)
0.616
Obstructive pulmonary disease Renal disease
1,222 (6.0) 688 (3.4)
47 (19.3) 24 (9.8)
1,175 (5.8) 664 (3.3)
\0.001 \0.001
44 (19.0) 24 (10.3)
36 (15.5) 24 (10.3)
0.326 1.000
STE
11,904 (58.6)
85 (34.8)
11,819 (59.6)
\0.001
92 (39.7)
90 (38.8)
0.849
\0.001
0.001
Data are expressed in absolute numbers and (percentage), or mean ± standard deviation AF previous atrial fibrillation, STE ST elevation acute coronary syndrome
regression using baseline clinical characteristics as covariates [14]. Nearest neighbor propensity score matching was used to obtain almost identical PS pairs (DG and NoDG patients). Goodness of fit was achieved if standardized differences were lower or equal to 10 %, suggesting a perfect matching [15]. Using this method, a propensitymatched cohort of 464 patients was obtained, 232 DG and 232 NoDG, and we compared the hospital outcome variables between them. A value of P \ 0.05 (2 sided) was considered significant. The results are expressed as absolute number and percentage of the group for the qualitative variables, mean ± standard deviation for quantitative variables, and odds ratio (OR) with the 95 % confidence interval (CI) for the regression analysis. Data analysis was conducted with version 15.0 of SPSS software (Inc; Chicago, IL5).
Results A total of 20,331 patients were included in the registry during this period. Among them, 244 patients constituted the DG group. Clinical characteristics are analyzed in Table 1. The clinical presentation was STE in 12,247 patients (60.2 %) and NoSTE in 8084 (39.8 %). 5,150 patients were female (25.3 %), and the mean age was 63.81 ± 12.71 years. DG patients were older, and more
often women, diabetics, had hypertension, or previous myocardial infarction, heart failure, stroke, AF, peripheral vascular disease, obstructive pulmonary disease or kidney disease. NoDG patients were more often smokers, or had family history of early coronary disease. All the 16 variables shown in this table were included in logistic regression analysis to obtain the propensity score for DG. A near perfect matching was reached between DG and NoDG (Table 1, columns 5–7). Patients admitted because of STE were mainly treated with fibrinolytic therapy in this series: 54.6 % (6,505/ 11,904), while 20.7 % (2,407/11,904) patients were treated with primary angioplasty. DG patients were less often treated with fibrinolytics 31.8 % (27/85) vs 54.8 % (6,478/ 11,819), P \ 0.001. There were no differences in the rate of primary angioplasty: 23.5 % (20/85) in DG vs 20.7 % (2,447/11,819) in NoDG. The hospital outcome is presented in Table 2. There were no differences in the length of hospital stay. On univariate analysis, DG patients had significantly higher hospital mortality in the total cohort (13.5 vs 5.3 %, P \ 0.001), and more cardiogenic shock, but less ventricular fibrillation, and no differences in atrioventricular block, stroke or reinfarction. However, comparing DG with NoDG in the propensity-matched cohort, there were no differences either in hospital mortality or in other complications (Table 2, columns 5–7).
123
Intern Emerg Med Table 2 Hospital outcome and complications comparing DG with NoDG patients, both in the total cohort and in the propensity-matched patients Total cohort (N = 20,331) Global
Propensity-matched cohort (N = 464)
DG (N = 244)
NoDG (N = 20,087)
P value
DG (N = 232)
NoDG (N = 232)
P value
Days of hospitalization
8.8 ± 24.0
8.7.1 ± 9.6
8.8 ± 24.20
0.967
8.8 ± 9.8
8.1 ± 5.3
0.463
Killip II or higher at admission
4,289 (21.3)
114 (46.7)
4,175 (21.0)
\0.001
107 (46.1)
106 (45.7)
0.988
Ventricular fibrillation
1,244 (6.1)
7 (2.9)
1,237 (6.2)
0.033
6 (2.6)
8 (3.4)
0.587
Ventricular tachycardia
820 (4.0)
9 (3.7)
811 (4.0)
0.783
9 (3.9)
11 (4.7)
0.648
High degree AV block
1,200 (5.9)
18 (7.4)
1,182 (5.9)
0.689
17 (7.3)
19 (8.2)
0.729
Reinfarction
377 (1.9)
4 (1.6)
373 (1.9)
0.802
4 (1.7)
6 (2.6)
0.523
Mechanical ventilation
1,082 (5.3)
16 (6.6)
1,066 (5.3)
0.476
15 (6.5)
27 (11.6)
0.052
Stroke
112 (0.6)
1 (0.4)
111 (0.6)
0.765
1 (0.4)
3 (1.3)
0.315
Cardiogenic shock
1,175 (5.8)
24 (9.8)
1,151 (5.7)
0.006
23 (9.9)
27 (11.6)
0.549
Exitus in CCU
889 (4.4)
27 (11.1)
862 (4.3)
\0.001
25 (10.8)
23 (9.8)
0.760
Exitus in hospital
1,088 (5.4)
33 (13.5)
1,055 (5.3)
\0.001
31 (13.4)
31 (13.4)
1.000
Data are expressed in absolute numbers and (percentage), or mean ± standard deviation AV atrioventricular, CCU critical care unit Table 3 Multivariate analysis of the determinants of hospital mortality, in the whole cohort
Discussion
Variable
Odds ratio
95 % CI
DG
1.2110
0.7886–1.8598
0.381
STE
2.1399
1.8568–2.4663
\0.001
We have found a significant increase of mortality related to the previous use of digoxin among patients admitted because of ACS, but this relationship vanished when corrected by a number of clinical variables. DG patients were 10 years older, more often woman, less often smokers, and had more diabetes, hypertension, previous myocardial infarction (MI), heart failure, stroke, AF, peripheral vascular disease, chronic obstructive pulmonary disease or renal disease. All these factors create a group of much higher risk when admitted because of ACS. As a consequence, it explains the differences in mortality. Such differences compensate the lesser proportion of STE in the DG group. Both the logistic regression analysis of determinants of mortality and the propensity score matching enable the analysis of the influence of previous digoxin on the prognosis, and the results are coincidental. Among patients with STE, fibrinolytic therapy was much less often administered in DG patients. The explanation is the high frequency of patients taking anticoagulants when admitted because of the myocardial infarction in the DG group. In a region like Andalucia, where distances to hospitals with angioplasty facilities can be very long, fibrinolysis still has an important role in STE management. This factor again would help to explain the poorer prognosis of the DG group. However, the second logistic regression analysis, which included fibrinolytic therapy, primary angioplasty, and invasive approach, did not find a significant influence of DG on hospital mortality. All these data do not sustain a dangerous role of digoxin that would overshadow the prognosis of patients if they
P value
Female
1.4513
1.2598–1.6720
\0.001
Age
1.0663
1.0587–1.0739
\0.001
Diabetes
1.3801
1.2075–1.5773
\0.001
Smoking habit
1.0052
0.8410–1.2014
0.954
Hypertension
0.8436
0.7345–0.9689
0.016
Family history of MI
0.6401
0.4677–0.8759
0.005
Previous MI
1.1936
1.0073- 1.4145
0.041
Previous HF
1.6186
1.2568–2.0845
\0.001
Previous stroke
1.7098
1.3852–2.1104
\0.001
Previous vasculopathy COPD
1.3261 1.4181
1.0298–1.7077 1.1351–1.7716
0.028 0.002
Previous kidney disease
1.5153
1.1599–1.9795
0.002
AF
1.2943
1.0781–1.5539
0.006
DG previous use of digoxin, STE ST elevation acute coronary syndrome, MI myocardial infarction, COPD chronic obstructive pulmonary disease, AF previous atrial fibrillation
After the multivariate analysis of determinants of hospital mortality (Table 3), DG had no significant influence on hospital prognosis (OR 1.2110, CI 0.7886–1.8598). A new model of logistic regression was then made, including fibrinolysis, primary angioplasty, invasive management, and the presence of Killip 2 or more. Again, the variable DG had no significant influence on mortality (OR 1.147, CI 0.745–1.767).
123
Intern Emerg Med
suffer an ACS while on treatment with it. Furthermore, our study did not found an adverse influence of previous digoxin on the appearance of dysrhythmias, the most commonly feared complications related to it. That contrasts with the findings of Whitbeck et al. [9]. In the AFFIRM trial, 4,060 patients with AF were randomized to rate control vs rhythm control over a 4-year period. Whitbeck et al. analyzed the impact on prognosis of the treatment with digoxin. They analyzed the use of digoxin at baseline and the consistency of it, and find an increase of long-term mortality of 66 %, which decreased to 41 % with adjustments for covariates and propensity scores. This contrasts with a more recently, new analysis of the same set of data from the AFFIRM trial, been presented by Gheorghiade et al. [16] with quite opposite conclusions. They made a propensity score analysis, obtaining 878 pairs of patients receiving and not receiving digoxin. Its model included 59 baseline characteristics, instead of the 27 variables of the model from Whitbeck et al. In this new analysis, they do not find any association of digoxin to increased mortality (hazard ratio 1.06, CI 0.83––1.37), allcause hospitalization (hazard ratio 0.96, CI 0.85–1.09) or incident non-fatal cardiac dysrhythmias (HR 0.90, CI 0.37–2.23). It is probably that the inclusion of 32 more variables had overcome the influence of confounding factors previously missed in the former analysis. Our results agree with this publication, as we did not find an increased mortality attributable to the previous use of digoxin, after correction for confounders. The controversy about digoxin is not new, with many old studies questioning the long-term safety of its use, while others show the benefits of its administration in chronic heart failure, and the association of withdrawal of digoxin with a significant clinical worsening [17–21]. All these polemics prompted a preeminent group of cardiologists to design a trial that could clarify the question of digoxin. A well designed and adequately powered randomized one, much stronger than all previous observational or post hoc analysis. This was the outstanding DIG trial, published 16 years ago. Six thousand and eight hundred patients in sinus rhythm and with left ventricular ejection fraction of 0.45 or less were randomly assigned to digoxin or placebo, and constituted the main study. An additional ancillary study was simultaneously made with 988 patients with left ventricular ejection fraction greater than 0.45. After a 3-year follow-up, they find no differences in mortality (34.8 % with digoxin vs 35.1 % with placebo, risk ratio 0.99, CI 0.91–1.07). Digoxin is associated with a significant decrease in hospitalization because of worsening heart failure (26.8 vs 34.7 %, risk ratio 0.72, CI 0.66–0.79, P \ 0.001). There were no differences in hospitalizations because of myocardial infarction, unstable
angina or coronary revascularization. These results are consistent whether the heart failures are ischemic or not ischemic, and also through the ancillary study of patients with ejection fraction greater than 0.45. No significant differences are found in hospitalization for ventricular dysrhythmias or atrioventricular block [22, 23]. The results of this landmark trial failed to resolve the old question. It gave arguments both to advocates and opponents of digoxin. But the absence of a significant benefit on mortality, together with the increase of the list of drugs that improve survival, would suggest that the use of digoxin should gradually diminish. The controversy continues on! Perhaps the key point in the ‘‘digoxin question’’ was the dose and the target of serum concentrations. In the PROMISE trial, designed to study the effect of milrinone on survival in heart failure patients, it was observed that high digoxin levels ([1.1 ng/mL) had higher mortality rate [2, 23]. Furthermore, Ahmed et al. performed an analysis of the data from de DIG trial, comparing the effects on mortality of low serum concentrations of digoxin with placebo. They matched 982 patients with serum digoxin of 0.5–0.9 ng/mL with 982 placebo patients, according to a propensity score, and compared outcomes. Low-digoxin results not only in reduction of heart failure, but also in a significant reduction of 3-year mortality, with a hazard ratio of 0.81. This reduction of mortality is also shown in patients with ischemic heart disease [24]. Consequently, in the Clinical Guidelines of the Heart Failure Society of America, it is recommended to use a digoxin dose of 0.125 mg daily, instead of the 0.250 commonly used in the 70s, and the serum level should be \1.0 ng/mL [5]. Our data strongly suggest that we should not fear a theoretical increase in ventricular dysrhythmias in our patients on digoxin therapy, nor a significant higher mortality rate, if a myocardial infarction eventually occurs. Low-dose digoxin is probably a safe medication, cheap and easy to administer, and it probably should regain a better position in the routine treatment of heart failure patients, whether they are ischemic or not.
Conclusions Patients who were taking digoxin when they present an ACS constitute a group of higher risk patients and with higher mortality rate during hospitalization. However, we have not found a significant increase in the risk of dysrhythmic complications. After adjustment for confounding variables, our study did not find an increase in mortality attributable to previous digoxin administration. Our results do not sustain any recommendation against the use of digoxin in chronic heart failure.
123
Intern Emerg Med
Limitations of the study This is a retrospective study of a registry not specifically designed to our purpose, so it has several limitations. First, we do not know the dose nor have serum concentrations of digoxin at admission, and such information would have been quite useful. Second, we do not have the exact indication for digoxin in our patients, although 69 % had a previous history consistent with AF, and 28 % had previous heart failure. Finally, we do not have long-term followup, but only hospitalization data.
9.
10.
11. Acknowledgments We are indebt with the Health Council of the Junta de Andalucia and the ‘‘Plan Integral de Cardiopatı´as en Andalucı´a’’ (PICA) for the support of the registry. We are also grateful to the general secretariat of the virtual platform ARIAM-Andalucia, and to the Software Company Coresoft Clı´nico (http://www.coresoft.es). Without their help, this work would not have been possible. Conflict of interest
12.
None. 13.
References 1. Doherty JE (1973) Digitalis glycosides. Pharmacokinetics and their clinical implications. Ann Intern Med 79:229–238 2. Hauptman PJ, Kelly RA (1999) Digitalis. Circulation 99:1265–1270 3. Butler VP Jr (1970) Digoxin: immunologic approaches to measurement and reversal of toxicity. N Engl J Med 283:1150–1156 4. Garcia-Rubira JC, Suarez A, Iban˜ez B (2012) Ventricular fibrillation after correct pacing in digoxin intoxication. J Emerg Med 42:e93–e94 5. Lindenfeld J, Albert NM, Boehmer JP, Collins SP, Ezekowitz JA, Givertz MM, Katz SD, Klapholz M, Moser DK, Rogers JG, Starling RC, Stevenson WG, Tang WH, Teerlink JR, Walsh MN (2010) HFSA 2010 Comprehensive Heart Failure Practice Guideline. J Card Fail 16:e1–e194 6. McMurray JJ, Adamopoulos S, Anker SD, Auricchio A, Bo¨hm M, Dickstein K, Falk V, Filippatos G, Fonseca C, Gomez-Sanchez MA, Jaarsma T, Køber L, Lip GY, Maggioni AP, Parkhomenko A, Pieske BM, Popescu BA, Rønnevik PK, Rutten FH, Schwitter J, Seferovic P, Stepinska J, Trindade PT, Voors AA, Zannad F, Zeiher A; ESC Committee for Practice Guidelines (2012). ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure 2012: The Task Force for the Diagnosis and Treatment of Acute and Chronic Heart Failure 2012 of the European Society of Cardiology. Developed in collaboration with the Heart Failure Association (HFA) of the ESC. Eur Heart J 33:1787–1847 7. Camm AJ, Kirchhof P, Lip GY, Schotten U, Savelieva I, Ernst S, Van Gelder IC, Al-Attar N, Hindricks G, Prendergast B, Heidbuchel H, Alfieri O, Angelini A, Atar D, Colonna P, De Caterina R, De Sutter J, Goette A, Gorenek B, Heldal M, Hohloser SH, Kolh P, Le Heuzey JY, Ponikowski P, Rutten FH (2010) Guidelines for the management of atrial fibrillation: the Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC). Eur Heart J 31:2369–2429 8. Steg PG, James SK, Atar D, Badano LP, Blo¨mstrom-Lundqvist C, Borger MA, Di Mario C, Dickstein K, Ducrocq G, FernandezAviles F, Gershlick AH, Giannuzzi P, Halvorsen S, Huber K, Juni P, Kastrati A, Knuuti J, Lenzen MJ, Mahaffey KW, Valgimigli
123
14. 15.
16.
17.
18.
19.
20.
21.
22.
M, van’t Hof A, Widimsky P, Zahger D (2012) ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation. Eur Heart J 33:2569–2619 Whitbeck MG, Charnigo RJ, Khairy P, Ziada K, Bailey AL, Zegarra MM, Shah J, Morales G, Macaulay T, Sorrell VL, Campbell CL, Gurley J, Anaya P, Nasr H, Bai R, Di Biase L, Booth DC, Jondeau G, Natale A, Roy D, Smyth S, Moliterno DJ, Elayi CS (2013) Increased mortality among patients taking digoxin—analysis from the AFFIRM study. Eur Heart J 34:1481–1488 Consejeria de Salud de la Junta de Andalucia (2009). Orden de 3 de Septiembre de 2009 por la que se crea el fichero con datos de caracter personal denominado Registro de Cardiopatı´as ARIAM. Boletı´n Oficial de la Junta de Andalucı´a 192:59–60. http://junta deandalucia.es/boja/2009/192/38 Saturno PJ, Felices F, Segura J, Vera A, Rodriguez JJ (2000) Reducing time delay in the thrombolysis of myocardial infarction: an internal quality improvement project. ARIAM Project Group. Analisis del Retraso en Infarto Agudo de Miocardio. Am J Med Qual 15:85–93 Ruiz-Baile´n M, Aguayo de Hoyos E, Serrano-Co´rcoles MC, Dia´z-Castellanos MA, Ramos-Cuadra JA, Reina-Toral A (2001) Efficacy of thrombolysis in patients with acute myocardial infarction requiring cardiopulmonary resuscitation. Intensive Care Med 27:1050–1057 Ruiz-Baile´n M, Aguayo de Hoyos E, Ramos-Cuadra JA, Dı´azCastellanos MA, Issa-Khozouz Z, Reina-Toral A, Lo´pez-Martı´nez A, Calatrava-Lo´pez J, Laynez-Bretones F, Castillo-Parra JC, De La Torre-Prados MV, ARIAM Group (2002) Influence of age on clinical course, management and mortality of acute myocardial infarction in the Spanish population. Int J Cardiol 85:285–296 Rubin DB (1997) Estimating causal effects from large data sets using propensity scores. Ann Intern Med 127:757–763 D’Agostino RB Jr (1998) Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group. Stat Med 17:2265–2281 Gheorghiade M, Fonarow GC, van Veldhuisen DJ, Cleland JG, Butler J, Epstein AE, Patel K, Aban IB, Aronow WS, Anker SD, Ahmed A (2013) Lack of evidence of increased mortality among patients with atrial fibrillation taking digoxin: findings from post hoc propensity-matched analysis of the AFFIRM trial. Eur Heart J 34:1489–1497 Jaeschke R, Oxman AD, Guyatt GH (1990) To what extent do congestive heart failure patients in sinus rhythm benefit from digoxin therapy? A systematic overview and meta-analysis. Am J Med 88:279–286 Køber L, Torp-Pedersen C, Gadsbøll N, Hildebrandt P, HøilundCarlsen PF (1994) Is digoxin an independent risk factor for longterm mortality after acute myocardial infarction? Eur Heart J 15:382–388 Packer M, Gheorghiade M, Young JB, Costantini PJ, Adams KF, Cody RJ, Smith LK, Van Voorhees L, Gourley LA, Jolly MK (1993) Withdrawal of digoxin from patients with chronic heart failure treated with angiotensin-converting-enzyme inhibitors. N Engl J Med 329:1–7 Uretsky BF, Young JB, Shahidi FE, Yellen LG, Harrison MC, Jolly MK (1993) Randomized study assessing the effect of digoxin withdrawal in patients with mild to moderate chronic congestive heart failure: results of the PROVED Trial. J Am Coll Cardiol 22:955–962 The Digitalis Investigation Group (1997) The effect of digoxin on mortality and morbidity in patients with heart failure. N Engl J Med 336:525–533 Packer M (1997) End of the oldest controversy in medicine. Are we ready to conclude the debate on digitalis? N Engl J Med 336:575–576
Intern Emerg Med 23. Packer M, Carver JR, Rodeheffer RJ, Ivanhoe RJ, DiBianco R, Zeldis SM, Hendrix GH, Bommer WJ, Elkayam U, Kukin ML, Mallis GI, Sollano JA, Shannon J, Tandon PK, DeMets DL (1991) Effect of oral milrinone on mortality in severe chronic heart failure. The PROMISE Study Research Group. N Engl J Med 325:1468–1475
24. Ahmed A, Pitt B, Rahimtoola SH, Waagstein F, White M, Love TE, Braunwald E (2008) Effects of digoxin at low serum concentrations on mortality and hospitalization in heart failure: a propensity-matched study of the DIG trial. Int J Cardiol 123:138–146
123
IM - ORIGINAL
The previous use of digoxin does not worsen early outcome of acute coronary syndromes: an analysis of the ARIAM Registry Juan Carlos Garcia-Rubira • Manuel Calvo-Taracido • Francisca Francisco-Aparicio • Manuel Almendro-Delia • Alejandro Recio-Mayoral • Antonio Reina Toral • Oscar Aramburu-Bodas • Pastora Gallego Garcı´a de Vinuesa • Jose´ Maria Cruz Ferna´ndez • Angel Garcia Alca´ntara • Rafael Hidalgo-Urbano
Received: 18 October 2013 / Accepted: 9 December 2013 Ó SIMI 2013
Abstract The aim of the study was to determine the influence of the previous use of digoxin on the hospital mortality and complications of patients admitted because of acute coronary syndrome (ACS). We analyzed the data of patients included in the ARIAM-Andalucia Registry, which involves 49 hospitals in Andalucia, Spain, from 2007 to 2012. Patients on digoxin treatment prior to their admission because of ACS constituted the digoxin group (DG), and were compared with the group of patients not on digoxin. Logistic regression and propensity score matching On behalf of the ARIAM Andalucı´a Study Group, the list of investigators is given in ESM Appendix.
Electronic supplementary material The online version of this article (doi:10.1007/s11739-013-1032-9) contains supplementary material, which is available to authorized users. J. C. Garcia-Rubira M. Calvo-Taracido F. Francisco-Aparicio M. Almendro-Delia A. Recio-Mayoral P. Gallego Garcı´a de Vinuesa J. M. Cruz Ferna´ndez R. Hidalgo-Urbano Cardiology Department, Virgen Macarena University Hospital, Seville, Spain J. C. Garcia-Rubira (&) Unidad Coronaria, Hospital Virgen Macarena, Avenida Dr Fedriani, sn, 41008 Seville, Spain e-mail: [email protected] A. Reina Toral Intensive Care Unit, Virgen de las Nieves Hospital, Granada, Spain O. Aramburu-Bodas Internal Medicine Department, Virgen Macarena University Hospital, Seville, Spain A. G. Alca´ntara Intensive Care Unit, Virgen de la Victoria Hospital, Ma´laga, Spain
were used to analyze the differences. We included 20,331 patients, of whom 244 (1.2 %) were on digoxin. DG patients were older (73.1 vs 63.7 years old), more often women, and had more diabetes, hypertension, previous myocardial infarction, heart failure, stroke, atrial fibrillation, peripheral vascular disease, obstructive pulmonary disease or kidney disease. On univariate analysis, DG patients had significantly higher hospital mortality (13.5 vs 5.3 % P \ 0.001), and more cardiogenic shock, but less ventricular fibrillation, and no differences in atrioventricular block, stroke or reinfarction. After the multivariate analysis, DG had no significant influence on hospital prognosis [odds ratio (OR) 1.21, 95 % confidence interval 0.79–1.86]. The analysis of a propensity-matched cohort of 464 patients (232 DG and 232 NoDG) did not find differences in hospital mortality (13.4 vs 13.4 %) nor other complications. In our cohort of ACS patients, the previous treatment with digoxin was not associated with an increase in dysrhythmic complications nor was an independent predictor of mortality during hospitalization. Keywords Acute coronary syndrome Atrial fibrillation Dysrhythmias Digoxin Heart failure Myocardial infarction List of abbreviations ACS Acute coronary syndrome AF Atrial fibrillation CI Confidence interval DG Patients on digoxin treatment previously to their admission MI Myocardial infarction NoDG Patients not on digoxin treatment previously to their admission NoSTE Acute coronary syndrome without ST elevation
123
Intern Emerg Med
OR STE
Odds ratio Acute coronary syndrome with ST elevation
mortality and complications of patients who are chronically taking digoxin, and who were subsequently admitted because of an acute coronary syndrome (ACS).
Introduction
Materials and methods
Years ago, digoxin along with diuretics, constituted the mainstay of treatment of patients with heart failure, and so remained for more than 200 years. Optimization of this therapeutic tool was encouraged, and relatively high doses, aiming to reach the higher half of the therapeutic range were usually recommended [1]. However, digoxin has a distinctive pharmacology, with an elimination half-life of 36–48 h when renal function is normal. The excretion is renal, and proportional to the glomerular filtration rate. It has interactions with other drugs and electrolyte disturbances easily enhance toxicity [2]. Early in the 70s, digitalis toxicity became an important issue, with the development of strategies for preventing and treating overdose of this drug [3]. Even today physicians must be aware of the potential toxicity in patients taking digoxin [4]. Nevertheless, digoxin is a very cheap drug that still has a role in the management of heart failure and atrial fibrillation, as it is recommended by some guidelines [5]. More recent statements in the management of symptomatic chronic heart failure have relegated the use of digoxin to the last position, only after the treatment with diuretics, betablockers, angiotensin-converting enzyme, angiotensin receptor blocker, mineralocorticoid receptor antagonist, and ivabradine, have failed. It is recommended earlier in patients with atrial fibrillation, but as a second-line drug in chronic heart failure, if heart rate control is not gained with either betablocker or calcium antagonists (in preserved systolic function) [6]. Digoxin is still one of the drugs recommended for controlling the heart rate in atrial fibrillation, although caution is warranted because of potential life threatening adverse effects. It is specially indicated by the European Task Force in patients with heart failure and left ventricular dysfunction [7]. In the scenario of ST elevation myocardial infarction, current guidelines of the European Society of Cardiology recommend the administration of digoxin with or without amiodarone, for heart rate control in patients who develop atrial fibrillation and acute heart failure or hypotension [8]. Together with these restrictive indications, a recent publication based on the data from the AFFIRM study calls attention to the finding of an increased mortality among patients taking digoxin [9]. This statement could be the final chapter in the history of digoxin, our old companion in heart failure! Our study attempts to shed a little more of light over the ‘‘digoxin question’’. We analyze the hospital
The ARIAM-Andalucia (Ana´lisis del Retraso en el Infarto Agudo de Miocardio en Andalucı´a) is a registry of patients with acute coronary syndrome admitted in the coronary units of 49 hospitals in Andalucia, Spain. This is an institutional project, aimed to improve the quality of treatment of ischemic heart disease, and include the patients admitted to the Intensive Care Units of these hospitals because of ACS. Each hospital introduced their patients through a Web interface, developed by Coresoft (http://www.cor esoft.es), which was accessed through the Internet (http:// www.ariam-andalucia.org). The data were subsequently reviewed and saved in a central unit. For this study, we performed a retrospective analysis with available data from the registry. The ARIAM-Andalucia virtual platform has become a global project, implemented in the Andalusian Public Health Service, partially financed by Health Counseling since 2005, which also includes other registries, such as adult cardiac surgery, cardiac arrests, or pediatric cardiac surgery. The basis of these records was published in an official gazette [10]. For the purpose of our present study, we have made a retrospective analysis of the data from January 2007 to January 2012. The analysis was retrospective, but all the data have been collected prospectively. This investigation has been approved by the local ethics committees. Other details on the methodology of introduction of the data have been previously published [11– 13]. Previous atrial fibrillation is not defined in the ARIAM, so that we took the combined variable ‘‘previous arrhythmia or atrial fibrillation at admission’’ (AF) as a surrogate. Both groups were compared in demographic and clinical characteristics at admission, and the clinical course of their hospitalization. ACS was classified as ACS with ST elevation (STE) or ACS without ST elevation (NoSTE). An initial comparison was performed according to the previous history of treatment with digoxin (DG) or without it (NoDG). We performed the Student’s t test for comparison of quantitative variables (age and length of hospital stay), and Pearson’s Chi-square test for qualitative variables. A multivariate model by logistic regression with the hospital mortality as dependent variable was made including all characteristics significantly different between DG and NoDG on the univariate analysis. To validate the results of this analysis, we calculated a propensity score of being DG. The propensity score model (PS) for the previous treatment with digoxin was obtained by logistic
123
Intern Emerg Med Table 1 Clinical characteristics of the patients, comparing DG with NoDG patients, both in the total cohort and in the propensity-matched patients Total cohort (N = 20,331) Global
Propensity-matched cohort (N = 464)
DG (N = 244)
NoDG (N = 20,087)
P value
DG (N = 232)
NoDG (N = 232)
P value
Age
63.8 ± 12.7
73.1 ± 8.9
63.7 ± 12.7
\0.001
73.1 ± 8.8
73.0 ± 9.0
0.921
Woman
5,150 (25.3)
98 (40.2)
5,052 (25.2)
\0.001
90 (38.8)
94 (40.5)
0.704
Smokers
7,101 (34.9)
29 (11.9)
7,072 (35.2)
\0.001
29 (12.5)
21 (9.1)
0.231
Diabetes
6,516 (32.0)
125 (51.2)
6,391 (31.8)
\0.001
119 (51.3)
121 (52.2)
0.853
Hypercholesteremia
8,494 (41.8)
105 (43.0)
8,389 (41.8)
0.689
99 (42.7)
107 (46.1)
0.455
Hypertension
10,699 (52.6)
187 (76.6)
10,512 (52.3)
178 (76.7)
182 (78.4)
0.656
Obesity
3,341 (16.4)
36 (14.8)
3,305 (16.5)
0.476
35 (15.1)
41 (17.7)
0.452
Familiar history
2,021 (9.9)
9 (3.7)
2,012 (10.0)
0.001
9 (3.9)
4 (1.7)
0.160
Previous myocardial infarction
3,013 (14.8)
57 (23.4)
2,956 (14.7)
\0.001
56 (24.1)
62 (26.7)
0.522
Previous heart failure
753 (3.7)
68 (27.9)
685 (3.4)
\0.001
63 (27.2)
56 (24.1)
0.457
Previous stroke
1,057 (5.2)
33 (13.5)
1,024 (5.1)
\0.001
30 (12.9)
41 (17.7)
0.156
AF
1,972 (9.7)
194 (79.5)
1,778 (8.9)
\0.001
182 (78.4)
186 (80.2)
0.647
Peripheral vascular disease
907 (4.5)
22 (9.0)
885 (4.4)
21 (9.1)
18 (7.8)
0.616
Obstructive pulmonary disease Renal disease
1,222 (6.0) 688 (3.4)
47 (19.3) 24 (9.8)
1,175 (5.8) 664 (3.3)
\0.001 \0.001
44 (19.0) 24 (10.3)
36 (15.5) 24 (10.3)
0.326 1.000
STE
11,904 (58.6)
85 (34.8)
11,819 (59.6)
\0.001
92 (39.7)
90 (38.8)
0.849
\0.001
0.001
Data are expressed in absolute numbers and (percentage), or mean ± standard deviation AF previous atrial fibrillation, STE ST elevation acute coronary syndrome
regression using baseline clinical characteristics as covariates [14]. Nearest neighbor propensity score matching was used to obtain almost identical PS pairs (DG and NoDG patients). Goodness of fit was achieved if standardized differences were lower or equal to 10 %, suggesting a perfect matching [15]. Using this method, a propensitymatched cohort of 464 patients was obtained, 232 DG and 232 NoDG, and we compared the hospital outcome variables between them. A value of P \ 0.05 (2 sided) was considered significant. The results are expressed as absolute number and percentage of the group for the qualitative variables, mean ± standard deviation for quantitative variables, and odds ratio (OR) with the 95 % confidence interval (CI) for the regression analysis. Data analysis was conducted with version 15.0 of SPSS software (Inc; Chicago, IL5).
Results A total of 20,331 patients were included in the registry during this period. Among them, 244 patients constituted the DG group. Clinical characteristics are analyzed in Table 1. The clinical presentation was STE in 12,247 patients (60.2 %) and NoSTE in 8084 (39.8 %). 5,150 patients were female (25.3 %), and the mean age was 63.81 ± 12.71 years. DG patients were older, and more
often women, diabetics, had hypertension, or previous myocardial infarction, heart failure, stroke, AF, peripheral vascular disease, obstructive pulmonary disease or kidney disease. NoDG patients were more often smokers, or had family history of early coronary disease. All the 16 variables shown in this table were included in logistic regression analysis to obtain the propensity score for DG. A near perfect matching was reached between DG and NoDG (Table 1, columns 5–7). Patients admitted because of STE were mainly treated with fibrinolytic therapy in this series: 54.6 % (6,505/ 11,904), while 20.7 % (2,407/11,904) patients were treated with primary angioplasty. DG patients were less often treated with fibrinolytics 31.8 % (27/85) vs 54.8 % (6,478/ 11,819), P \ 0.001. There were no differences in the rate of primary angioplasty: 23.5 % (20/85) in DG vs 20.7 % (2,447/11,819) in NoDG. The hospital outcome is presented in Table 2. There were no differences in the length of hospital stay. On univariate analysis, DG patients had significantly higher hospital mortality in the total cohort (13.5 vs 5.3 %, P \ 0.001), and more cardiogenic shock, but less ventricular fibrillation, and no differences in atrioventricular block, stroke or reinfarction. However, comparing DG with NoDG in the propensity-matched cohort, there were no differences either in hospital mortality or in other complications (Table 2, columns 5–7).
123
Intern Emerg Med Table 2 Hospital outcome and complications comparing DG with NoDG patients, both in the total cohort and in the propensity-matched patients Total cohort (N = 20,331) Global
Propensity-matched cohort (N = 464)
DG (N = 244)
NoDG (N = 20,087)
P value
DG (N = 232)
NoDG (N = 232)
P value
Days of hospitalization
8.8 ± 24.0
8.7.1 ± 9.6
8.8 ± 24.20
0.967
8.8 ± 9.8
8.1 ± 5.3
0.463
Killip II or higher at admission
4,289 (21.3)
114 (46.7)
4,175 (21.0)
\0.001
107 (46.1)
106 (45.7)
0.988
Ventricular fibrillation
1,244 (6.1)
7 (2.9)
1,237 (6.2)
0.033
6 (2.6)
8 (3.4)
0.587
Ventricular tachycardia
820 (4.0)
9 (3.7)
811 (4.0)
0.783
9 (3.9)
11 (4.7)
0.648
High degree AV block
1,200 (5.9)
18 (7.4)
1,182 (5.9)
0.689
17 (7.3)
19 (8.2)
0.729
Reinfarction
377 (1.9)
4 (1.6)
373 (1.9)
0.802
4 (1.7)
6 (2.6)
0.523
Mechanical ventilation
1,082 (5.3)
16 (6.6)
1,066 (5.3)
0.476
15 (6.5)
27 (11.6)
0.052
Stroke
112 (0.6)
1 (0.4)
111 (0.6)
0.765
1 (0.4)
3 (1.3)
0.315
Cardiogenic shock
1,175 (5.8)
24 (9.8)
1,151 (5.7)
0.006
23 (9.9)
27 (11.6)
0.549
Exitus in CCU
889 (4.4)
27 (11.1)
862 (4.3)
\0.001
25 (10.8)
23 (9.8)
0.760
Exitus in hospital
1,088 (5.4)
33 (13.5)
1,055 (5.3)
\0.001
31 (13.4)
31 (13.4)
1.000
Data are expressed in absolute numbers and (percentage), or mean ± standard deviation AV atrioventricular, CCU critical care unit Table 3 Multivariate analysis of the determinants of hospital mortality, in the whole cohort
Discussion
Variable
Odds ratio
95 % CI
DG
1.2110
0.7886–1.8598
0.381
STE
2.1399
1.8568–2.4663
\0.001
We have found a significant increase of mortality related to the previous use of digoxin among patients admitted because of ACS, but this relationship vanished when corrected by a number of clinical variables. DG patients were 10 years older, more often woman, less often smokers, and had more diabetes, hypertension, previous myocardial infarction (MI), heart failure, stroke, AF, peripheral vascular disease, chronic obstructive pulmonary disease or renal disease. All these factors create a group of much higher risk when admitted because of ACS. As a consequence, it explains the differences in mortality. Such differences compensate the lesser proportion of STE in the DG group. Both the logistic regression analysis of determinants of mortality and the propensity score matching enable the analysis of the influence of previous digoxin on the prognosis, and the results are coincidental. Among patients with STE, fibrinolytic therapy was much less often administered in DG patients. The explanation is the high frequency of patients taking anticoagulants when admitted because of the myocardial infarction in the DG group. In a region like Andalucia, where distances to hospitals with angioplasty facilities can be very long, fibrinolysis still has an important role in STE management. This factor again would help to explain the poorer prognosis of the DG group. However, the second logistic regression analysis, which included fibrinolytic therapy, primary angioplasty, and invasive approach, did not find a significant influence of DG on hospital mortality. All these data do not sustain a dangerous role of digoxin that would overshadow the prognosis of patients if they
P value
Female
1.4513
1.2598–1.6720
\0.001
Age
1.0663
1.0587–1.0739
\0.001
Diabetes
1.3801
1.2075–1.5773
\0.001
Smoking habit
1.0052
0.8410–1.2014
0.954
Hypertension
0.8436
0.7345–0.9689
0.016
Family history of MI
0.6401
0.4677–0.8759
0.005
Previous MI
1.1936
1.0073- 1.4145
0.041
Previous HF
1.6186
1.2568–2.0845
\0.001
Previous stroke
1.7098
1.3852–2.1104
\0.001
Previous vasculopathy COPD
1.3261 1.4181
1.0298–1.7077 1.1351–1.7716
0.028 0.002
Previous kidney disease
1.5153
1.1599–1.9795
0.002
AF
1.2943
1.0781–1.5539
0.006
DG previous use of digoxin, STE ST elevation acute coronary syndrome, MI myocardial infarction, COPD chronic obstructive pulmonary disease, AF previous atrial fibrillation
After the multivariate analysis of determinants of hospital mortality (Table 3), DG had no significant influence on hospital prognosis (OR 1.2110, CI 0.7886–1.8598). A new model of logistic regression was then made, including fibrinolysis, primary angioplasty, invasive management, and the presence of Killip 2 or more. Again, the variable DG had no significant influence on mortality (OR 1.147, CI 0.745–1.767).
123
Intern Emerg Med
suffer an ACS while on treatment with it. Furthermore, our study did not found an adverse influence of previous digoxin on the appearance of dysrhythmias, the most commonly feared complications related to it. That contrasts with the findings of Whitbeck et al. [9]. In the AFFIRM trial, 4,060 patients with AF were randomized to rate control vs rhythm control over a 4-year period. Whitbeck et al. analyzed the impact on prognosis of the treatment with digoxin. They analyzed the use of digoxin at baseline and the consistency of it, and find an increase of long-term mortality of 66 %, which decreased to 41 % with adjustments for covariates and propensity scores. This contrasts with a more recently, new analysis of the same set of data from the AFFIRM trial, been presented by Gheorghiade et al. [16] with quite opposite conclusions. They made a propensity score analysis, obtaining 878 pairs of patients receiving and not receiving digoxin. Its model included 59 baseline characteristics, instead of the 27 variables of the model from Whitbeck et al. In this new analysis, they do not find any association of digoxin to increased mortality (hazard ratio 1.06, CI 0.83––1.37), allcause hospitalization (hazard ratio 0.96, CI 0.85–1.09) or incident non-fatal cardiac dysrhythmias (HR 0.90, CI 0.37–2.23). It is probably that the inclusion of 32 more variables had overcome the influence of confounding factors previously missed in the former analysis. Our results agree with this publication, as we did not find an increased mortality attributable to the previous use of digoxin, after correction for confounders. The controversy about digoxin is not new, with many old studies questioning the long-term safety of its use, while others show the benefits of its administration in chronic heart failure, and the association of withdrawal of digoxin with a significant clinical worsening [17–21]. All these polemics prompted a preeminent group of cardiologists to design a trial that could clarify the question of digoxin. A well designed and adequately powered randomized one, much stronger than all previous observational or post hoc analysis. This was the outstanding DIG trial, published 16 years ago. Six thousand and eight hundred patients in sinus rhythm and with left ventricular ejection fraction of 0.45 or less were randomly assigned to digoxin or placebo, and constituted the main study. An additional ancillary study was simultaneously made with 988 patients with left ventricular ejection fraction greater than 0.45. After a 3-year follow-up, they find no differences in mortality (34.8 % with digoxin vs 35.1 % with placebo, risk ratio 0.99, CI 0.91–1.07). Digoxin is associated with a significant decrease in hospitalization because of worsening heart failure (26.8 vs 34.7 %, risk ratio 0.72, CI 0.66–0.79, P \ 0.001). There were no differences in hospitalizations because of myocardial infarction, unstable
angina or coronary revascularization. These results are consistent whether the heart failures are ischemic or not ischemic, and also through the ancillary study of patients with ejection fraction greater than 0.45. No significant differences are found in hospitalization for ventricular dysrhythmias or atrioventricular block [22, 23]. The results of this landmark trial failed to resolve the old question. It gave arguments both to advocates and opponents of digoxin. But the absence of a significant benefit on mortality, together with the increase of the list of drugs that improve survival, would suggest that the use of digoxin should gradually diminish. The controversy continues on! Perhaps the key point in the ‘‘digoxin question’’ was the dose and the target of serum concentrations. In the PROMISE trial, designed to study the effect of milrinone on survival in heart failure patients, it was observed that high digoxin levels ([1.1 ng/mL) had higher mortality rate [2, 23]. Furthermore, Ahmed et al. performed an analysis of the data from de DIG trial, comparing the effects on mortality of low serum concentrations of digoxin with placebo. They matched 982 patients with serum digoxin of 0.5–0.9 ng/mL with 982 placebo patients, according to a propensity score, and compared outcomes. Low-digoxin results not only in reduction of heart failure, but also in a significant reduction of 3-year mortality, with a hazard ratio of 0.81. This reduction of mortality is also shown in patients with ischemic heart disease [24]. Consequently, in the Clinical Guidelines of the Heart Failure Society of America, it is recommended to use a digoxin dose of 0.125 mg daily, instead of the 0.250 commonly used in the 70s, and the serum level should be \1.0 ng/mL [5]. Our data strongly suggest that we should not fear a theoretical increase in ventricular dysrhythmias in our patients on digoxin therapy, nor a significant higher mortality rate, if a myocardial infarction eventually occurs. Low-dose digoxin is probably a safe medication, cheap and easy to administer, and it probably should regain a better position in the routine treatment of heart failure patients, whether they are ischemic or not.
Conclusions Patients who were taking digoxin when they present an ACS constitute a group of higher risk patients and with higher mortality rate during hospitalization. However, we have not found a significant increase in the risk of dysrhythmic complications. After adjustment for confounding variables, our study did not find an increase in mortality attributable to previous digoxin administration. Our results do not sustain any recommendation against the use of digoxin in chronic heart failure.
123
Intern Emerg Med
Limitations of the study This is a retrospective study of a registry not specifically designed to our purpose, so it has several limitations. First, we do not know the dose nor have serum concentrations of digoxin at admission, and such information would have been quite useful. Second, we do not have the exact indication for digoxin in our patients, although 69 % had a previous history consistent with AF, and 28 % had previous heart failure. Finally, we do not have long-term followup, but only hospitalization data.
9.
10.
11. Acknowledgments We are indebt with the Health Council of the Junta de Andalucia and the ‘‘Plan Integral de Cardiopatı´as en Andalucı´a’’ (PICA) for the support of the registry. We are also grateful to the general secretariat of the virtual platform ARIAM-Andalucia, and to the Software Company Coresoft Clı´nico (http://www.coresoft.es). Without their help, this work would not have been possible. Conflict of interest
12.
None. 13.
References 1. Doherty JE (1973) Digitalis glycosides. Pharmacokinetics and their clinical implications. Ann Intern Med 79:229–238 2. Hauptman PJ, Kelly RA (1999) Digitalis. Circulation 99:1265–1270 3. Butler VP Jr (1970) Digoxin: immunologic approaches to measurement and reversal of toxicity. N Engl J Med 283:1150–1156 4. Garcia-Rubira JC, Suarez A, Iban˜ez B (2012) Ventricular fibrillation after correct pacing in digoxin intoxication. J Emerg Med 42:e93–e94 5. Lindenfeld J, Albert NM, Boehmer JP, Collins SP, Ezekowitz JA, Givertz MM, Katz SD, Klapholz M, Moser DK, Rogers JG, Starling RC, Stevenson WG, Tang WH, Teerlink JR, Walsh MN (2010) HFSA 2010 Comprehensive Heart Failure Practice Guideline. J Card Fail 16:e1–e194 6. McMurray JJ, Adamopoulos S, Anker SD, Auricchio A, Bo¨hm M, Dickstein K, Falk V, Filippatos G, Fonseca C, Gomez-Sanchez MA, Jaarsma T, Køber L, Lip GY, Maggioni AP, Parkhomenko A, Pieske BM, Popescu BA, Rønnevik PK, Rutten FH, Schwitter J, Seferovic P, Stepinska J, Trindade PT, Voors AA, Zannad F, Zeiher A; ESC Committee for Practice Guidelines (2012). ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure 2012: The Task Force for the Diagnosis and Treatment of Acute and Chronic Heart Failure 2012 of the European Society of Cardiology. Developed in collaboration with the Heart Failure Association (HFA) of the ESC. Eur Heart J 33:1787–1847 7. Camm AJ, Kirchhof P, Lip GY, Schotten U, Savelieva I, Ernst S, Van Gelder IC, Al-Attar N, Hindricks G, Prendergast B, Heidbuchel H, Alfieri O, Angelini A, Atar D, Colonna P, De Caterina R, De Sutter J, Goette A, Gorenek B, Heldal M, Hohloser SH, Kolh P, Le Heuzey JY, Ponikowski P, Rutten FH (2010) Guidelines for the management of atrial fibrillation: the Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC). Eur Heart J 31:2369–2429 8. Steg PG, James SK, Atar D, Badano LP, Blo¨mstrom-Lundqvist C, Borger MA, Di Mario C, Dickstein K, Ducrocq G, FernandezAviles F, Gershlick AH, Giannuzzi P, Halvorsen S, Huber K, Juni P, Kastrati A, Knuuti J, Lenzen MJ, Mahaffey KW, Valgimigli
123
14. 15.
16.
17.
18.
19.
20.
21.
22.
M, van’t Hof A, Widimsky P, Zahger D (2012) ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation. Eur Heart J 33:2569–2619 Whitbeck MG, Charnigo RJ, Khairy P, Ziada K, Bailey AL, Zegarra MM, Shah J, Morales G, Macaulay T, Sorrell VL, Campbell CL, Gurley J, Anaya P, Nasr H, Bai R, Di Biase L, Booth DC, Jondeau G, Natale A, Roy D, Smyth S, Moliterno DJ, Elayi CS (2013) Increased mortality among patients taking digoxin—analysis from the AFFIRM study. Eur Heart J 34:1481–1488 Consejeria de Salud de la Junta de Andalucia (2009). Orden de 3 de Septiembre de 2009 por la que se crea el fichero con datos de caracter personal denominado Registro de Cardiopatı´as ARIAM. Boletı´n Oficial de la Junta de Andalucı´a 192:59–60. http://junta deandalucia.es/boja/2009/192/38 Saturno PJ, Felices F, Segura J, Vera A, Rodriguez JJ (2000) Reducing time delay in the thrombolysis of myocardial infarction: an internal quality improvement project. ARIAM Project Group. Analisis del Retraso en Infarto Agudo de Miocardio. Am J Med Qual 15:85–93 Ruiz-Baile´n M, Aguayo de Hoyos E, Serrano-Co´rcoles MC, Dia´z-Castellanos MA, Ramos-Cuadra JA, Reina-Toral A (2001) Efficacy of thrombolysis in patients with acute myocardial infarction requiring cardiopulmonary resuscitation. Intensive Care Med 27:1050–1057 Ruiz-Baile´n M, Aguayo de Hoyos E, Ramos-Cuadra JA, Dı´azCastellanos MA, Issa-Khozouz Z, Reina-Toral A, Lo´pez-Martı´nez A, Calatrava-Lo´pez J, Laynez-Bretones F, Castillo-Parra JC, De La Torre-Prados MV, ARIAM Group (2002) Influence of age on clinical course, management and mortality of acute myocardial infarction in the Spanish population. Int J Cardiol 85:285–296 Rubin DB (1997) Estimating causal effects from large data sets using propensity scores. Ann Intern Med 127:757–763 D’Agostino RB Jr (1998) Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group. Stat Med 17:2265–2281 Gheorghiade M, Fonarow GC, van Veldhuisen DJ, Cleland JG, Butler J, Epstein AE, Patel K, Aban IB, Aronow WS, Anker SD, Ahmed A (2013) Lack of evidence of increased mortality among patients with atrial fibrillation taking digoxin: findings from post hoc propensity-matched analysis of the AFFIRM trial. Eur Heart J 34:1489–1497 Jaeschke R, Oxman AD, Guyatt GH (1990) To what extent do congestive heart failure patients in sinus rhythm benefit from digoxin therapy? A systematic overview and meta-analysis. Am J Med 88:279–286 Køber L, Torp-Pedersen C, Gadsbøll N, Hildebrandt P, HøilundCarlsen PF (1994) Is digoxin an independent risk factor for longterm mortality after acute myocardial infarction? Eur Heart J 15:382–388 Packer M, Gheorghiade M, Young JB, Costantini PJ, Adams KF, Cody RJ, Smith LK, Van Voorhees L, Gourley LA, Jolly MK (1993) Withdrawal of digoxin from patients with chronic heart failure treated with angiotensin-converting-enzyme inhibitors. N Engl J Med 329:1–7 Uretsky BF, Young JB, Shahidi FE, Yellen LG, Harrison MC, Jolly MK (1993) Randomized study assessing the effect of digoxin withdrawal in patients with mild to moderate chronic congestive heart failure: results of the PROVED Trial. J Am Coll Cardiol 22:955–962 The Digitalis Investigation Group (1997) The effect of digoxin on mortality and morbidity in patients with heart failure. N Engl J Med 336:525–533 Packer M (1997) End of the oldest controversy in medicine. Are we ready to conclude the debate on digitalis? N Engl J Med 336:575–576
Intern Emerg Med 23. Packer M, Carver JR, Rodeheffer RJ, Ivanhoe RJ, DiBianco R, Zeldis SM, Hendrix GH, Bommer WJ, Elkayam U, Kukin ML, Mallis GI, Sollano JA, Shannon J, Tandon PK, DeMets DL (1991) Effect of oral milrinone on mortality in severe chronic heart failure. The PROMISE Study Research Group. N Engl J Med 325:1468–1475
24. Ahmed A, Pitt B, Rahimtoola SH, Waagstein F, White M, Love TE, Braunwald E (2008) Effects of digoxin at low serum concentrations on mortality and hospitalization in heart failure: a propensity-matched study of the DIG trial. Int J Cardiol 123:138–146
123
