https://orcid.org/0000-0003-3986-9152
Abstract
To investigate the association between long-term β-blocker therapy and clinical outcomes in patients without heart failure (HF) after acute myocardial infarction (AMI).
Between 2010 and 2015, a total of 28 970 patients who underwent coronary revascularization for AMI with β-blocker prescription at hospital discharge and were event-free from death, recurrent myocardial infarction (MI), or HF for 1 year were enrolled from Korean nationwide medical insurance data. The primary outcome was all-cause death. The secondary outcomes were recurrent MI, hospitalization for new HF, and a composite of all-cause death, recurrent MI, or hospitalization for new HF. Outcomes were compared between β-blocker therapy for ≥1 year (N = 22 707) and β-blocker therapy for <1 year (N = 6263) using landmark analysis at 1 year after index MI. Compared with patients receiving β-blocker therapy for <1 year, those receiving β-blocker therapy for ≥1 year had significantly lower risks of all-cause death [adjusted hazard ratio (HR) 0.81; 95% confidence interval (CI) 0.72–0.91] and composite of all-cause death, recurrent MI, or hospitalization for new HF (adjusted HR 0.82; 95% CI 0.75–0.89), but not the risks of recurrent MI or hospitalization for new HF. The lower risk of all-cause death associated with persistent β-blocker therapy was observed beyond 2 years (adjusted HR 0.86; 95% CI 0.75–0.99) but not beyond 3 years (adjusted HR 0.87; 95% CI 0.73–1.03) after MI.
In this nationwide cohort, β-blocker therapy for ≥1 year after MI was associated with reduced all-cause death among patients with AMI without HF.
See page 3530 for the editorial comment on this article (doi: )
Introduction
In the pre-reperfusion era, β-blockers reduced mortality in patients with acute myocardial infarction (AMI).1 Unfortunately, evidence supporting routine β-blocker therapy after myocardial infarction (MI) is lacking in the era of primary percutaneous coronary intervention (PCI)2. The results of observational studies on the association between β-blocker therapy at discharge and 1-year clinical outcomes were inconsistent.3–5 However, β-blockers have anti-ischaemic, antiarrhythmic, and antiadrenergic properties.6 Early use of β-blockers reduces myocardial damage and is associated with improved clinical outcomes even in patients undergoing primary PCI.7 , 8 Current major guidelines recommend that β-blockers should be continued during and after hospitalization for all patients with ST-segment elevation myocardial infarction (STEMI) and with no contraindications to their use.9 , 10 In real-world practice, β-blockers are frequently prescribed in the early period (up to 1 year) after MI unless patients have contraindications. In observational studies, the proportion of patients receiving β-blockers at 1 year after MI ranged from 72% to 89%.3 , 11
An important but difficult decision on β-blocker therapy after MI is to determine the duration of β-blocker therapy after discharge in patients without specific indications such as heart failure (HF) or left ventricular systolic dysfunction. However, few observational studies3 , 11 , 12 and a randomized trial13 with relatively small sample size and potential selection bias are inadequate for definite conclusion for long-term β-blocker therapy. Therefore, in the present study, we sought to investigate the association between long-term (≥1 year) β-blocker therapy and mortality after acute MI in patients without HF using recent nationwide medical insurance data of Korean population.
Methods
Study population
Korea has a single-payer national health system. The National Health Insurance Service (NHIS) maintains records of all covered inpatient and outpatient visits, procedures, and prescriptions. This is a population-based retrospective cohort study built using the NHIS data.14–16
This study included all men and women over 18 years old who underwent revascularization [PCI or coronary artery bypass graft surgery (CABG)] for MI between 1 January 2010 and 31 December 2015 (n = 54 195). Patients were followed until 31 December 2016.
We excluded patients who had history of HF (I110, I130, I132, I255, I420, I425–I429, I43, I50, I971) (n = 5569), MI (I21–I23, I252), or revascularization before index MI (n = 5319); who had been prescribed β-blockers within 6 months before index MI (n = 11 300); who had had contraindications for β-blockers including asthma or chronic obstructive lung disease (J43–J46) within 6 months before index MI (n = 2994), or conduction abnormality (pacemaker, second- or third-degree atrioventricular block; I441, I442, I495, Z950) before index MI (n = 68). Since our objective was to compare clinical outcomes between long-term β-blocker therapy for ≥1 year and β-blocker therapy for <1 year after AMI, we excluded patients who had not been prescribed any β-blockers within 3 months after hospital discharge (n = 2453) to minimize difference in baseline characteristics between the groups. Finally, we excluded patients who had developed death, recurrent MI, or HF within 1 year after index MI (n = 8557) to control immortal time bias.17 The Institutional Review Board of Samsung Medical Center approved this study (2018-03-046) and informed consent was waived as we used de-identified administrative data.
Measurements
NHIS claims for inpatient and outpatient visits, procedures, and prescriptions were coded using the International Classification of Diseases, 10th Revision.18 As the NHIS routinely audits the claims, such data are considered reliable and used in numerous peer-reviewed publications.14 , 15 In regard to diagnosis of MI, the validation study in 2013 showed the value of 92%.19
β-Blocker use was identified by prescription data with Korean Drug and Anatomical Therapeutic Chemical Codes. If β-blocker prescription had been continued for 1 year from hospital discharge, we considered that the patient was a long-term β-blocker user. If a new prescription had been redeemed within 180 days after expiration of prior prescription, β-blocker therapy was considered to be continued.
Outcomes and definitions
The primary outcome was all-cause death. Vital status and cause of death were obtained from death certification collected by Statistics Korea at the Ministry of Strategy and Finance of South Korea.14 The secondary outcomes were recurrent MI, hospitalization for new HF, and a composite of each secondary outcome and all-cause death.
Comorbidities were summarized using Charlson index. In addition, we included hypertension (I10–I13, I15), atrial fibrillation or flutter (I48), hyperlipidaemia (E78), gout (M10), depression (F32, F33), and valve disorder (I05–I08, I34–I37, Z952–Z954). We identified medications of angiotensin-converting enzyme inhibitor, angiotensin receptor blocker, calcium channel blocker, statin, aspirin, clopidogrel, ticagrelor or prasugrel, and anticoagulant (Supplementary material online, Table S1).
Statistical analysis
Clinical outcomes were compared between patients receiving β-blocker for ≥1 year and those receiving β-blocker for <1 year using landmark analysis at 1 year after index MI. The Kaplan–Meier method was used to analyse the cumulative incidence of events over time, whereas log-rank tests were applied to evaluate differences between the groups. Cox proportional-hazards regression analyses were used to compare the clinical outcomes. We adjusted for age, sex, hypertension, atrial fibrillation of flutter, hyperlipidaemia, gout, depression, valve disorder, Charlson comorbidities, and medications at 1 year after index MI. We used multivariable mixed-effects Cox regression analysis including an admitted hospital as a random intercept to adjust hospital effect. Clinical outcomes according to type of β-blocker were also compared.
To identify long-term beneficial effects of persistent β-blocker therapy, we conducted same analysis at 2 years and 3 years after index MI. The analyses were also conducted based on landmark analysis in patients who were event-free from death, recurrent MI, or hospitalization for new HF at 2 years (n = 22 769) or 3 years (n = 17 009) after index MI. These landmark analyses were adjusted for comorbidities and medications at 2 years or 3 years after index MI, respectively.
Subgroup analyses were conducted by age (age <65 vs. ≥65), sex, type of revascularization (PCI vs. CABG), hypertension, diabetes, Charlson comorbidity index (<3 vs. ≥3), use of angiotensin-converting enzyme inhibitor, angiotensin receptor blocker, or statin, and type of hospital (tertiary hospital vs. others). All P-values were two-sided, and a P-value of <0.05 was considered as significant. Analyses were performed with the use of SAS® Visual Analytics (SAS Institute Inc., USA) and STATA version 14 (StataCorp LP, College Station, TX, USA).
Results
Clinical characteristics
Among 28 970 patients who underwent revascularization for AMI, 22 707 (78.4%) were prescribed β-blocker for ≥1 year (median 1146 days) (Figure 1). The annual rates of patients receiving β-blocker therapy for ≥1 year are shown in Supplementary material online, Figure S1. The most frequently prescribed β-blocker was carvedilol (52.9%), followed by bisoprolol (33.0%) and nebivolol (5.8%) (Supplementary material online, Table S2). At the index MI, patients receiving β-blocker for ≥1 year were younger (60.2 vs. 62.4 years; P < 0.001) and more likely to be men (79.6% vs. 78.3%; P = 0.03), be treated at tertiary hospital (51.2% vs. 48.1%; P < 0.001) and have hyperlipidaemia (5.6% vs. 4.6%; P = 0.002) compared with those receiving β-blocker for <1 year (Table 1).
Study population. CABG, coronary artery bypass grafting; MI, myocardial infarction; PCI, percutaneous coronary intervention.
Baseline characteristics at index MI and at 1 year after index MI
| At MI | At 1 year after MI | |||||
|---|---|---|---|---|---|---|
| β-blocker ≥1 year (N = 22 707) | β-blocker <1 year (N = 6263) | P-value | β-blocker ≥1 year (N = 22 707) | β-blocker <1 year (N = 6263) | P-value | |
| Age (years), mean (SD) | 60.2 (12.1) | 62.4 (13.1) | <0.001 | |||
| Sex | 0.03 | |||||
| Male | 18 074 (79.6) | 4904 (78.3) | ||||
| Female | 4633 (20.4) | 1359 (21.7) | ||||
| Tertiary hospital | 11 628 (51.2) | 3010 (48.1) | <0.001 | |||
| Hypertension | 6866 (30.2) | 1595 (25.5) | <0.001 | 8091 (35.6) | 1996 (31.9) | <0.001 |
| Atrial fibrillation or flutter | 67 (0.3) | 19 (0.3) | 0.92 | 138 (0.6) | 57 (0.9) | 0.01 |
| Charlson index, mean (SD) | 0.7 (1.0) | 0.7 (1.1) | <0.001 | 1.9 (1.2) | 2.0 (1.2) | <0.001 |
| Charlson comorbiditya | ||||||
| Peripheral vascular disease | 298 (1.3) | 112 (1.8) | 0.005 | 458 (2.0) | 163 (2.6) | 0.005 |
| Cerebrovascular disease | 1177 (5.2) | 451 (7.2) | <0.001 | 1394 (6.1) | 577 (9.2) | <0.001 |
| Dementia | 226 (1.0) | 149 (2.4) | <0.001 | 403 (1.8) | 266 (4.2) | <0.001 |
| Chronic pulmonary disease | 2147 (9.5) | 600 (9.6) | 0.77 | 2934 (12.9) | 812 (13) | 0.63 |
| Peptic ulcer disease | 1741 (7.7) | 517 (8.3) | 0.13 | 1372 (6.0) | 392 (6.3) | 0.53 |
| Liver disease | 780 (3.4) | 223 (3.6) | 0.63 | 804 (3.5) | 230 (3.7) | 0.62 |
| Diabetes | 4246 (18.7) | 1169 (18.7) | 0.95 | 4702 (20.7) | 1280 (20.4) | 0.64 |
| Paraplegia and hemiplegia | 47 (0.2) | 27 (0.4) | 0.002 | 114 (0.5) | 75 (1.2) | <0.001 |
| Renal disease | 244 (1.1) | 119 (1.9) | <0.001 | 484 (2.1) | 196 (3.1) | <0.001 |
| Cancer | 634 (2.8) | 188 (3.0) | 0.38 | 697 (3.1) | 234 (3.7) | 0.008 |
| AIDS/HIV | 7 (0.0) | 3 (0.0) | 0.46 | 7 (0.0) | 3 (0.0) | 0.46 |
| Hyperlipidaemia | 1264 (5.6) | 287 (4.6) | 0.002 | 2149 (9.5) | 524 (8.4) | 0.008 |
| Gout | 458 (2.0) | 120 (1.9) | 0.613 | 686 (3.0) | 175 (2.8) | 0.349 |
| Depression | 288 (1.3) | 93 (1.5) | 0.183 | 490 (2.2) | 166 (2.7) | 0.02 |
| Valve disorder | 5 (0) | 3 (0) | 0.382 | 290 (1.3) | 9 (0.1) | 0.218 |
| Medications | ||||||
| Angiotensin-converting enzyme inhibitor | 12 788 (56.3) | 3126 (49.9) | <0.001 | 6456 (28.4) | 882 (14.1) | <0.001 |
| Angiotensin receptor blocker | 6782 (29.9) | 2076 (33.1) | <0.001 | 10 505 (46.3) | 2287 (36.5) | <0.001 |
| Calcium channel blocker | 5526 (24.3) | 1812 (28.9) | <0.001 | 4536 (20.0) | 1787 (28.5) | <0.001 |
| Statin | 21 006 (92.5) | 5720 (91.3) | 0.002 | 21 623 (95.2) | 4298 (68.6) | <0.001 |
| Aspirin | 22 237 (97.9) | 6102 (97.4) | 0.02 | 21 542 (94.9) | 4364 (69.7) | <0.001 |
| Clopidogrel | 16 192 (71.3) | 4397 (70.2) | 0.09 | 15 730 (69.3) | 3175 (50.7) | <0.001 |
| Ticagrelor or prasugrel | 5650 (24.9) | 1597 (25.5) | 0.32 | 3282 (14.5) | 555 (8.9) | <0.001 |
| DAPT | 21 391 (94.2) | 5839 (93.2) | 0.004 | |||
| Duration of DAPT, median (IQR) | 434 (351–778) | 344 (93–550) | <0.001 | |||
| Anticoagulant | 420 (1.8) | 144 (2.3) | 0.02 | 323 (1.4) | 98 (1.6) | 0.41 |
| Duration of β-blocker therapy, median (IQR) | 1146 (722–1723) | 72 (12–194) | <0.001 | |||
| At MI | At 1 year after MI | |||||
|---|---|---|---|---|---|---|
| β-blocker ≥1 year (N = 22 707) | β-blocker <1 year (N = 6263) | P-value | β-blocker ≥1 year (N = 22 707) | β-blocker <1 year (N = 6263) | P-value | |
| Age (years), mean (SD) | 60.2 (12.1) | 62.4 (13.1) | <0.001 | |||
| Sex | 0.03 | |||||
| Male | 18 074 (79.6) | 4904 (78.3) | ||||
| Female | 4633 (20.4) | 1359 (21.7) | ||||
| Tertiary hospital | 11 628 (51.2) | 3010 (48.1) | <0.001 | |||
| Hypertension | 6866 (30.2) | 1595 (25.5) | <0.001 | 8091 (35.6) | 1996 (31.9) | <0.001 |
| Atrial fibrillation or flutter | 67 (0.3) | 19 (0.3) | 0.92 | 138 (0.6) | 57 (0.9) | 0.01 |
| Charlson index, mean (SD) | 0.7 (1.0) | 0.7 (1.1) | <0.001 | 1.9 (1.2) | 2.0 (1.2) | <0.001 |
| Charlson comorbiditya | ||||||
| Peripheral vascular disease | 298 (1.3) | 112 (1.8) | 0.005 | 458 (2.0) | 163 (2.6) | 0.005 |
| Cerebrovascular disease | 1177 (5.2) | 451 (7.2) | <0.001 | 1394 (6.1) | 577 (9.2) | <0.001 |
| Dementia | 226 (1.0) | 149 (2.4) | <0.001 | 403 (1.8) | 266 (4.2) | <0.001 |
| Chronic pulmonary disease | 2147 (9.5) | 600 (9.6) | 0.77 | 2934 (12.9) | 812 (13) | 0.63 |
| Peptic ulcer disease | 1741 (7.7) | 517 (8.3) | 0.13 | 1372 (6.0) | 392 (6.3) | 0.53 |
| Liver disease | 780 (3.4) | 223 (3.6) | 0.63 | 804 (3.5) | 230 (3.7) | 0.62 |
| Diabetes | 4246 (18.7) | 1169 (18.7) | 0.95 | 4702 (20.7) | 1280 (20.4) | 0.64 |
| Paraplegia and hemiplegia | 47 (0.2) | 27 (0.4) | 0.002 | 114 (0.5) | 75 (1.2) | <0.001 |
| Renal disease | 244 (1.1) | 119 (1.9) | <0.001 | 484 (2.1) | 196 (3.1) | <0.001 |
| Cancer | 634 (2.8) | 188 (3.0) | 0.38 | 697 (3.1) | 234 (3.7) | 0.008 |
| AIDS/HIV | 7 (0.0) | 3 (0.0) | 0.46 | 7 (0.0) | 3 (0.0) | 0.46 |
| Hyperlipidaemia | 1264 (5.6) | 287 (4.6) | 0.002 | 2149 (9.5) | 524 (8.4) | 0.008 |
| Gout | 458 (2.0) | 120 (1.9) | 0.613 | 686 (3.0) | 175 (2.8) | 0.349 |
| Depression | 288 (1.3) | 93 (1.5) | 0.183 | 490 (2.2) | 166 (2.7) | 0.02 |
| Valve disorder | 5 (0) | 3 (0) | 0.382 | 290 (1.3) | 9 (0.1) | 0.218 |
| Medications | ||||||
| Angiotensin-converting enzyme inhibitor | 12 788 (56.3) | 3126 (49.9) | <0.001 | 6456 (28.4) | 882 (14.1) | <0.001 |
| Angiotensin receptor blocker | 6782 (29.9) | 2076 (33.1) | <0.001 | 10 505 (46.3) | 2287 (36.5) | <0.001 |
| Calcium channel blocker | 5526 (24.3) | 1812 (28.9) | <0.001 | 4536 (20.0) | 1787 (28.5) | <0.001 |
| Statin | 21 006 (92.5) | 5720 (91.3) | 0.002 | 21 623 (95.2) | 4298 (68.6) | <0.001 |
| Aspirin | 22 237 (97.9) | 6102 (97.4) | 0.02 | 21 542 (94.9) | 4364 (69.7) | <0.001 |
| Clopidogrel | 16 192 (71.3) | 4397 (70.2) | 0.09 | 15 730 (69.3) | 3175 (50.7) | <0.001 |
| Ticagrelor or prasugrel | 5650 (24.9) | 1597 (25.5) | 0.32 | 3282 (14.5) | 555 (8.9) | <0.001 |
| DAPT | 21 391 (94.2) | 5839 (93.2) | 0.004 | |||
| Duration of DAPT, median (IQR) | 434 (351–778) | 344 (93–550) | <0.001 | |||
| Anticoagulant | 420 (1.8) | 144 (2.3) | 0.02 | 323 (1.4) | 98 (1.6) | 0.41 |
| Duration of β-blocker therapy, median (IQR) | 1146 (722–1723) | 72 (12–194) | <0.001 | |||
AIDS, acquired immune deficiency syndrome; DAPT, dual antiplatelet therapy; HIV, human immunodeficiency virus; IQR, interquartile range; MI, myocardial infarction, SD, standard deviation.
Charlson comorbidity were defined as presence of disease within 1 year prior to index MI.
Baseline characteristics at index MI and at 1 year after index MI
| At MI | At 1 year after MI | |||||
|---|---|---|---|---|---|---|
| β-blocker ≥1 year (N = 22 707) | β-blocker <1 year (N = 6263) | P-value | β-blocker ≥1 year (N = 22 707) | β-blocker <1 year (N = 6263) | P-value | |
| Age (years), mean (SD) | 60.2 (12.1) | 62.4 (13.1) | <0.001 | |||
| Sex | 0.03 | |||||
| Male | 18 074 (79.6) | 4904 (78.3) | ||||
| Female | 4633 (20.4) | 1359 (21.7) | ||||
| Tertiary hospital | 11 628 (51.2) | 3010 (48.1) | <0.001 | |||
| Hypertension | 6866 (30.2) | 1595 (25.5) | <0.001 | 8091 (35.6) | 1996 (31.9) | <0.001 |
| Atrial fibrillation or flutter | 67 (0.3) | 19 (0.3) | 0.92 | 138 (0.6) | 57 (0.9) | 0.01 |
| Charlson index, mean (SD) | 0.7 (1.0) | 0.7 (1.1) | <0.001 | 1.9 (1.2) | 2.0 (1.2) | <0.001 |
| Charlson comorbiditya | ||||||
| Peripheral vascular disease | 298 (1.3) | 112 (1.8) | 0.005 | 458 (2.0) | 163 (2.6) | 0.005 |
| Cerebrovascular disease | 1177 (5.2) | 451 (7.2) | <0.001 | 1394 (6.1) | 577 (9.2) | <0.001 |
| Dementia | 226 (1.0) | 149 (2.4) | <0.001 | 403 (1.8) | 266 (4.2) | <0.001 |
| Chronic pulmonary disease | 2147 (9.5) | 600 (9.6) | 0.77 | 2934 (12.9) | 812 (13) | 0.63 |
| Peptic ulcer disease | 1741 (7.7) | 517 (8.3) | 0.13 | 1372 (6.0) | 392 (6.3) | 0.53 |
| Liver disease | 780 (3.4) | 223 (3.6) | 0.63 | 804 (3.5) | 230 (3.7) | 0.62 |
| Diabetes | 4246 (18.7) | 1169 (18.7) | 0.95 | 4702 (20.7) | 1280 (20.4) | 0.64 |
| Paraplegia and hemiplegia | 47 (0.2) | 27 (0.4) | 0.002 | 114 (0.5) | 75 (1.2) | <0.001 |
| Renal disease | 244 (1.1) | 119 (1.9) | <0.001 | 484 (2.1) | 196 (3.1) | <0.001 |
| Cancer | 634 (2.8) | 188 (3.0) | 0.38 | 697 (3.1) | 234 (3.7) | 0.008 |
| AIDS/HIV | 7 (0.0) | 3 (0.0) | 0.46 | 7 (0.0) | 3 (0.0) | 0.46 |
| Hyperlipidaemia | 1264 (5.6) | 287 (4.6) | 0.002 | 2149 (9.5) | 524 (8.4) | 0.008 |
| Gout | 458 (2.0) | 120 (1.9) | 0.613 | 686 (3.0) | 175 (2.8) | 0.349 |
| Depression | 288 (1.3) | 93 (1.5) | 0.183 | 490 (2.2) | 166 (2.7) | 0.02 |
| Valve disorder | 5 (0) | 3 (0) | 0.382 | 290 (1.3) | 9 (0.1) | 0.218 |
| Medications | ||||||
| Angiotensin-converting enzyme inhibitor | 12 788 (56.3) | 3126 (49.9) | <0.001 | 6456 (28.4) | 882 (14.1) | <0.001 |
| Angiotensin receptor blocker | 6782 (29.9) | 2076 (33.1) | <0.001 | 10 505 (46.3) | 2287 (36.5) | <0.001 |
| Calcium channel blocker | 5526 (24.3) | 1812 (28.9) | <0.001 | 4536 (20.0) | 1787 (28.5) | <0.001 |
| Statin | 21 006 (92.5) | 5720 (91.3) | 0.002 | 21 623 (95.2) | 4298 (68.6) | <0.001 |
| Aspirin | 22 237 (97.9) | 6102 (97.4) | 0.02 | 21 542 (94.9) | 4364 (69.7) | <0.001 |
| Clopidogrel | 16 192 (71.3) | 4397 (70.2) | 0.09 | 15 730 (69.3) | 3175 (50.7) | <0.001 |
| Ticagrelor or prasugrel | 5650 (24.9) | 1597 (25.5) | 0.32 | 3282 (14.5) | 555 (8.9) | <0.001 |
| DAPT | 21 391 (94.2) | 5839 (93.2) | 0.004 | |||
| Duration of DAPT, median (IQR) | 434 (351–778) | 344 (93–550) | <0.001 | |||
| Anticoagulant | 420 (1.8) | 144 (2.3) | 0.02 | 323 (1.4) | 98 (1.6) | 0.41 |
| Duration of β-blocker therapy, median (IQR) | 1146 (722–1723) | 72 (12–194) | <0.001 | |||
| At MI | At 1 year after MI | |||||
|---|---|---|---|---|---|---|
| β-blocker ≥1 year (N = 22 707) | β-blocker <1 year (N = 6263) | P-value | β-blocker ≥1 year (N = 22 707) | β-blocker <1 year (N = 6263) | P-value | |
| Age (years), mean (SD) | 60.2 (12.1) | 62.4 (13.1) | <0.001 | |||
| Sex | 0.03 | |||||
| Male | 18 074 (79.6) | 4904 (78.3) | ||||
| Female | 4633 (20.4) | 1359 (21.7) | ||||
| Tertiary hospital | 11 628 (51.2) | 3010 (48.1) | <0.001 | |||
| Hypertension | 6866 (30.2) | 1595 (25.5) | <0.001 | 8091 (35.6) | 1996 (31.9) | <0.001 |
| Atrial fibrillation or flutter | 67 (0.3) | 19 (0.3) | 0.92 | 138 (0.6) | 57 (0.9) | 0.01 |
| Charlson index, mean (SD) | 0.7 (1.0) | 0.7 (1.1) | <0.001 | 1.9 (1.2) | 2.0 (1.2) | <0.001 |
| Charlson comorbiditya | ||||||
| Peripheral vascular disease | 298 (1.3) | 112 (1.8) | 0.005 | 458 (2.0) | 163 (2.6) | 0.005 |
| Cerebrovascular disease | 1177 (5.2) | 451 (7.2) | <0.001 | 1394 (6.1) | 577 (9.2) | <0.001 |
| Dementia | 226 (1.0) | 149 (2.4) | <0.001 | 403 (1.8) | 266 (4.2) | <0.001 |
| Chronic pulmonary disease | 2147 (9.5) | 600 (9.6) | 0.77 | 2934 (12.9) | 812 (13) | 0.63 |
| Peptic ulcer disease | 1741 (7.7) | 517 (8.3) | 0.13 | 1372 (6.0) | 392 (6.3) | 0.53 |
| Liver disease | 780 (3.4) | 223 (3.6) | 0.63 | 804 (3.5) | 230 (3.7) | 0.62 |
| Diabetes | 4246 (18.7) | 1169 (18.7) | 0.95 | 4702 (20.7) | 1280 (20.4) | 0.64 |
| Paraplegia and hemiplegia | 47 (0.2) | 27 (0.4) | 0.002 | 114 (0.5) | 75 (1.2) | <0.001 |
| Renal disease | 244 (1.1) | 119 (1.9) | <0.001 | 484 (2.1) | 196 (3.1) | <0.001 |
| Cancer | 634 (2.8) | 188 (3.0) | 0.38 | 697 (3.1) | 234 (3.7) | 0.008 |
| AIDS/HIV | 7 (0.0) | 3 (0.0) | 0.46 | 7 (0.0) | 3 (0.0) | 0.46 |
| Hyperlipidaemia | 1264 (5.6) | 287 (4.6) | 0.002 | 2149 (9.5) | 524 (8.4) | 0.008 |
| Gout | 458 (2.0) | 120 (1.9) | 0.613 | 686 (3.0) | 175 (2.8) | 0.349 |
| Depression | 288 (1.3) | 93 (1.5) | 0.183 | 490 (2.2) | 166 (2.7) | 0.02 |
| Valve disorder | 5 (0) | 3 (0) | 0.382 | 290 (1.3) | 9 (0.1) | 0.218 |
| Medications | ||||||
| Angiotensin-converting enzyme inhibitor | 12 788 (56.3) | 3126 (49.9) | <0.001 | 6456 (28.4) | 882 (14.1) | <0.001 |
| Angiotensin receptor blocker | 6782 (29.9) | 2076 (33.1) | <0.001 | 10 505 (46.3) | 2287 (36.5) | <0.001 |
| Calcium channel blocker | 5526 (24.3) | 1812 (28.9) | <0.001 | 4536 (20.0) | 1787 (28.5) | <0.001 |
| Statin | 21 006 (92.5) | 5720 (91.3) | 0.002 | 21 623 (95.2) | 4298 (68.6) | <0.001 |
| Aspirin | 22 237 (97.9) | 6102 (97.4) | 0.02 | 21 542 (94.9) | 4364 (69.7) | <0.001 |
| Clopidogrel | 16 192 (71.3) | 4397 (70.2) | 0.09 | 15 730 (69.3) | 3175 (50.7) | <0.001 |
| Ticagrelor or prasugrel | 5650 (24.9) | 1597 (25.5) | 0.32 | 3282 (14.5) | 555 (8.9) | <0.001 |
| DAPT | 21 391 (94.2) | 5839 (93.2) | 0.004 | |||
| Duration of DAPT, median (IQR) | 434 (351–778) | 344 (93–550) | <0.001 | |||
| Anticoagulant | 420 (1.8) | 144 (2.3) | 0.02 | 323 (1.4) | 98 (1.6) | 0.41 |
| Duration of β-blocker therapy, median (IQR) | 1146 (722–1723) | 72 (12–194) | <0.001 | |||
AIDS, acquired immune deficiency syndrome; DAPT, dual antiplatelet therapy; HIV, human immunodeficiency virus; IQR, interquartile range; MI, myocardial infarction, SD, standard deviation.
Charlson comorbidity were defined as presence of disease within 1 year prior to index MI.
At 1 year after the index MI, patients receiving β-blocker for ≥1 year had lower prevalence of atrial fibrillation or flutter (0.6% vs. 0.9%; P = 0.01) and lower Charlson comorbidity index (1.9 vs. 2.0; P < 0.001) compared with those receiving β-blocker for <1 year. All other medications except anticoagulant were more frequently prescribed in patients receiving β-blocker for ≥1 year.
Outcomes
During a median follow-up of 3.5 years (interquartile range 2.2–5.0 years), a total of 1684 deaths were observed. Patients receiving β-blocker for ≥1 year had a significantly lower risk of all-cause death than those receiving β-blocker for <1 year [13.1 vs. 25.7 per 1000 person-years; hazard ratio (HR) 0.50; 95% confidence interval (CI) 0.46–0.55; P < 0.001; Table 2 and Figure 2]. The lower risk of all-cause death in patients receiving β-blocker for ≥1 year remained significant in multivariable analysis (adjusted HR 0.81; 95% CI 0.72–0.91; P < 0.001). Of 2560 events of composite of death, recurrent MI, or hospitalization for new HF, 1748 cases occurred in patients receiving β-blocker for ≥1 year, and 812 cases occurred in patients receiving β-blocker for <1 year (20.9 vs. 37.0 per 1000 person-years; adjusted HR 0.82; 95% CI 0.75–0.89; P < 0.001). Patients receiving β-blocker for ≥1 year had a strong trend towards a lower risk of recurrent MI (5.2 vs. 8.2 per 1000 person-years; adjusted HR 0.82; 95% CI 0.67–1.01; P = 0.06). The association between the β-blocker use for ≥1 year and risk of all-cause death was consistent in various subgroups (Figure 3). The results were similar by type of β-blocker (Supplementary material online, Table S3).
Cumulative incidences of clinical outcomes since 1 year after myocardial infarction. (A) All-cause death, (B) recurrent MI, (C) hospitalization for new heart failure, and (D) a composite of all-cause death, recurrent MI, or hospitalization for new heart failure. MI, myocardial infarction.
Subgroup analysis for a risk of all-cause death. ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; CABG, coronary artery bypass grafting; CI, confidence interval; HR, hazard ratio.
Clinical outcomes since 1 year after index MI
| Outcomes | β-blocker ≥1 year (N = 22 707) | β-blocker <1 year (N = 6263) | ||||||
|---|---|---|---|---|---|---|---|---|
| Number of cases | Incidence rate (/1000 person-years) | Number of cases | Incidence rate (/1000 person-years) | Unadjusted | Adjustedb | |||
| HR (95% CI) | P-value | HR (95% CI) | P-value | |||||
| All-cause death | 1109 | 13.1 | 575 | 25.7 | 0.50 (0.46–0.55) | <0.001 | 0.81 (0.72–0.91) | <0.001 |
| Recurrent MI | 440 | 5.2 | 180 | 8.2 | 0.66 (0.55–0.79) | <0.001 | 0.82 (0.67–1.01) | 0.06 |
| Hospitalization for new heart failure | 333 | 3.9 | 121 | 5.4 | 0.74 (0.59–0.92) | 0.007 | 0.92 (0.73–1.17) | 0.51 |
| A composite outcomea | 1748 | 20.9 | 812 | 37.0 | 0.56 (0.51–0.60) | <0.001 | 0.82 (0.75–0.89) | <0.001 |
| Outcomes | β-blocker ≥1 year (N = 22 707) | β-blocker <1 year (N = 6263) | ||||||
|---|---|---|---|---|---|---|---|---|
| Number of cases | Incidence rate (/1000 person-years) | Number of cases | Incidence rate (/1000 person-years) | Unadjusted | Adjustedb | |||
| HR (95% CI) | P-value | HR (95% CI) | P-value | |||||
| All-cause death | 1109 | 13.1 | 575 | 25.7 | 0.50 (0.46–0.55) | <0.001 | 0.81 (0.72–0.91) | <0.001 |
| Recurrent MI | 440 | 5.2 | 180 | 8.2 | 0.66 (0.55–0.79) | <0.001 | 0.82 (0.67–1.01) | 0.06 |
| Hospitalization for new heart failure | 333 | 3.9 | 121 | 5.4 | 0.74 (0.59–0.92) | 0.007 | 0.92 (0.73–1.17) | 0.51 |
| A composite outcomea | 1748 | 20.9 | 812 | 37.0 | 0.56 (0.51–0.60) | <0.001 | 0.82 (0.75–0.89) | <0.001 |
Landmark analyses were performed among patients who were event-free during 1 year after index MI.
CI, confidence interval; HR, hazard ratio; MI, myocardial infarction.
A composite of all-cause death, recurrent MI, or hospitalization for new heart failure.
Adjusted for comorbidities and medications at 1 year after index MI.
Clinical outcomes since 1 year after index MI
| Outcomes | β-blocker ≥1 year (N = 22 707) | β-blocker <1 year (N = 6263) | ||||||
|---|---|---|---|---|---|---|---|---|
| Number of cases | Incidence rate (/1000 person-years) | Number of cases | Incidence rate (/1000 person-years) | Unadjusted | Adjustedb | |||
| HR (95% CI) | P-value | HR (95% CI) | P-value | |||||
| All-cause death | 1109 | 13.1 | 575 | 25.7 | 0.50 (0.46–0.55) | <0.001 | 0.81 (0.72–0.91) | <0.001 |
| Recurrent MI | 440 | 5.2 | 180 | 8.2 | 0.66 (0.55–0.79) | <0.001 | 0.82 (0.67–1.01) | 0.06 |
| Hospitalization for new heart failure | 333 | 3.9 | 121 | 5.4 | 0.74 (0.59–0.92) | 0.007 | 0.92 (0.73–1.17) | 0.51 |
| A composite outcomea | 1748 | 20.9 | 812 | 37.0 | 0.56 (0.51–0.60) | <0.001 | 0.82 (0.75–0.89) | <0.001 |
| Outcomes | β-blocker ≥1 year (N = 22 707) | β-blocker <1 year (N = 6263) | ||||||
|---|---|---|---|---|---|---|---|---|
| Number of cases | Incidence rate (/1000 person-years) | Number of cases | Incidence rate (/1000 person-years) | Unadjusted | Adjustedb | |||
| HR (95% CI) | P-value | HR (95% CI) | P-value | |||||
| All-cause death | 1109 | 13.1 | 575 | 25.7 | 0.50 (0.46–0.55) | <0.001 | 0.81 (0.72–0.91) | <0.001 |
| Recurrent MI | 440 | 5.2 | 180 | 8.2 | 0.66 (0.55–0.79) | <0.001 | 0.82 (0.67–1.01) | 0.06 |
| Hospitalization for new heart failure | 333 | 3.9 | 121 | 5.4 | 0.74 (0.59–0.92) | 0.007 | 0.92 (0.73–1.17) | 0.51 |
| A composite outcomea | 1748 | 20.9 | 812 | 37.0 | 0.56 (0.51–0.60) | <0.001 | 0.82 (0.75–0.89) | <0.001 |
Landmark analyses were performed among patients who were event-free during 1 year after index MI.
CI, confidence interval; HR, hazard ratio; MI, myocardial infarction.
A composite of all-cause death, recurrent MI, or hospitalization for new heart failure.
Adjusted for comorbidities and medications at 1 year after index MI.
Long-term β-blocker use beyond 2 years after index myocardial infarction
Figure 4 presents patients with persistent vs. discontinued β-blocker therapy at each time point during follow-up. Among 22 769 patients who were event-free from death, recurrent MI, or hospitalization for new HF during 2 years after index MI, those receiving persistent β-blocker therapy for ≥2 years had significantly lower risks of all-cause death (adjusted HR 0.86, 0.75 to 0.99; P = 0.04), recurrent MI (adjusted HR 0.67; 95% CI 0.51–0.87; P = 0.003), and composite of all-cause death, recurrent MI, or hospitalization for new HF (adjusted HR 0.82; 95% CI 0.73–0.92; P < 0.001) than those receiving β-blocker for <2 years (Table 3).
Persistent β-blocker therapy and clinical outcomes after myocardial infarction. A composite outcome includes all-cause death, recurrent myocardial infarction, or hospitalization for new heart failure.
Clinical outcomes according to duration of β-blocker therapy after myocardial infarction. CI, confidence interval; HF, heart failure; HR, hazard ratio; MI, myocardial infarction.
Clinical outcomes since 2 years and 3 years after index MI
| Landmark time from index MI | |||||||
|---|---|---|---|---|---|---|---|
| At 2 years | At 3 years | ||||||
| Duration of β-blocker therapy | Duration of β-blocker therapy | ||||||
| ≥2 years (persistent use) (N = 16 156) | <2 year (N = 6613) | P-value | ≥3 years (persistent use) (N = 11 290) | <3 year (N = 5719) | P-value | ||
| All-cause death | Incidence rate | 9.2 | 16.9 | 6.6 | 12.3 | ||
| Unadjusted HR | 0.53 (0.47–0.60) | 1 (ref) | <0.001 | 0.53 (0.45–0.62) | 1 (ref) | <0.001 | |
| Adjusted HRa | 0.86 (0.75–0.99) | 1 (ref) | 0.04 | 0.87 (0.73–1.03) | 1 (ref) | 0.12 | |
| Recurrent MI | Incidence rate | 2.8 | 5.5 | 2.1 | 3.9 | ||
| Unadjusted HR | 0.51 (0.41–0.64) | 1 (ref) | <0.001 | 0.53 (0.40–0.69) | 1 (ref) | <0.001 | |
| Adjusted HR | 0.66 (0.51–0.87) | 1 (ref) | 0.003 | 0.74 (0.53–1.01) | 1 (ref) | 0.06 | |
| Hospitalization for new heart failure | Incidence rate | 2.4 | 3.3 | ||||
| Unadjusted HR | 0.72 (0.56–0.91) | 1 (ref) | 0.007 | 0.76 (0.55–1.04) | 1 (ref) | 0.08 | |
| Adjusted HR | 0.95 (0.73–1.24) | 1 (ref) | 0.72 | 1.04 (0.75–1.46) | 1 (ref) | 0.81 | |
| A composite outcomeb | Incidence rate | 13.6 | 24.2 | ||||
| Unadjusted HR | 0.55 (0.49–0.61) | 1 (ref) | <0.001 | 0.55 (0.48–0.63) | 1 (ref) | <0.001 | |
| Adjusted HR | 0.82 (0.73–0.92) | 1 (ref) | <0.001 | 0.85 (0.74–0.98) | 1 (ref) | 0.03 | |
| Landmark time from index MI | |||||||
|---|---|---|---|---|---|---|---|
| At 2 years | At 3 years | ||||||
| Duration of β-blocker therapy | Duration of β-blocker therapy | ||||||
| ≥2 years (persistent use) (N = 16 156) | <2 year (N = 6613) | P-value | ≥3 years (persistent use) (N = 11 290) | <3 year (N = 5719) | P-value | ||
| All-cause death | Incidence rate | 9.2 | 16.9 | 6.6 | 12.3 | ||
| Unadjusted HR | 0.53 (0.47–0.60) | 1 (ref) | <0.001 | 0.53 (0.45–0.62) | 1 (ref) | <0.001 | |
| Adjusted HRa | 0.86 (0.75–0.99) | 1 (ref) | 0.04 | 0.87 (0.73–1.03) | 1 (ref) | 0.12 | |
| Recurrent MI | Incidence rate | 2.8 | 5.5 | 2.1 | 3.9 | ||
| Unadjusted HR | 0.51 (0.41–0.64) | 1 (ref) | <0.001 | 0.53 (0.40–0.69) | 1 (ref) | <0.001 | |
| Adjusted HR | 0.66 (0.51–0.87) | 1 (ref) | 0.003 | 0.74 (0.53–1.01) | 1 (ref) | 0.06 | |
| Hospitalization for new heart failure | Incidence rate | 2.4 | 3.3 | ||||
| Unadjusted HR | 0.72 (0.56–0.91) | 1 (ref) | 0.007 | 0.76 (0.55–1.04) | 1 (ref) | 0.08 | |
| Adjusted HR | 0.95 (0.73–1.24) | 1 (ref) | 0.72 | 1.04 (0.75–1.46) | 1 (ref) | 0.81 | |
| A composite outcomeb | Incidence rate | 13.6 | 24.2 | ||||
| Unadjusted HR | 0.55 (0.49–0.61) | 1 (ref) | <0.001 | 0.55 (0.48–0.63) | 1 (ref) | <0.001 | |
| Adjusted HR | 0.82 (0.73–0.92) | 1 (ref) | <0.001 | 0.85 (0.74–0.98) | 1 (ref) | 0.03 | |
Landmark analyses were performed among patients who were event-free for 2 years or 3 years after index MI, respectively. Incidence rates were presented as events per 1000 person-years.
HR, hazard ratio; MI, myocardial infarction; ref, reference.
Adjusted for comorbidities and medications at 2 years or 3 years after index MI, respectively.
A composite of all-cause death, recurrent MI, or hospitalization for new heart failure.
Clinical outcomes since 2 years and 3 years after index MI
| Landmark time from index MI | |||||||
|---|---|---|---|---|---|---|---|
| At 2 years | At 3 years | ||||||
| Duration of β-blocker therapy | Duration of β-blocker therapy | ||||||
| ≥2 years (persistent use) (N = 16 156) | <2 year (N = 6613) | P-value | ≥3 years (persistent use) (N = 11 290) | <3 year (N = 5719) | P-value | ||
| All-cause death | Incidence rate | 9.2 | 16.9 | 6.6 | 12.3 | ||
| Unadjusted HR | 0.53 (0.47–0.60) | 1 (ref) | <0.001 | 0.53 (0.45–0.62) | 1 (ref) | <0.001 | |
| Adjusted HRa | 0.86 (0.75–0.99) | 1 (ref) | 0.04 | 0.87 (0.73–1.03) | 1 (ref) | 0.12 | |
| Recurrent MI | Incidence rate | 2.8 | 5.5 | 2.1 | 3.9 | ||
| Unadjusted HR | 0.51 (0.41–0.64) | 1 (ref) | <0.001 | 0.53 (0.40–0.69) | 1 (ref) | <0.001 | |
| Adjusted HR | 0.66 (0.51–0.87) | 1 (ref) | 0.003 | 0.74 (0.53–1.01) | 1 (ref) | 0.06 | |
| Hospitalization for new heart failure | Incidence rate | 2.4 | 3.3 | ||||
| Unadjusted HR | 0.72 (0.56–0.91) | 1 (ref) | 0.007 | 0.76 (0.55–1.04) | 1 (ref) | 0.08 | |
| Adjusted HR | 0.95 (0.73–1.24) | 1 (ref) | 0.72 | 1.04 (0.75–1.46) | 1 (ref) | 0.81 | |
| A composite outcomeb | Incidence rate | 13.6 | 24.2 | ||||
| Unadjusted HR | 0.55 (0.49–0.61) | 1 (ref) | <0.001 | 0.55 (0.48–0.63) | 1 (ref) | <0.001 | |
| Adjusted HR | 0.82 (0.73–0.92) | 1 (ref) | <0.001 | 0.85 (0.74–0.98) | 1 (ref) | 0.03 | |
| Landmark time from index MI | |||||||
|---|---|---|---|---|---|---|---|
| At 2 years | At 3 years | ||||||
| Duration of β-blocker therapy | Duration of β-blocker therapy | ||||||
| ≥2 years (persistent use) (N = 16 156) | <2 year (N = 6613) | P-value | ≥3 years (persistent use) (N = 11 290) | <3 year (N = 5719) | P-value | ||
| All-cause death | Incidence rate | 9.2 | 16.9 | 6.6 | 12.3 | ||
| Unadjusted HR | 0.53 (0.47–0.60) | 1 (ref) | <0.001 | 0.53 (0.45–0.62) | 1 (ref) | <0.001 | |
| Adjusted HRa | 0.86 (0.75–0.99) | 1 (ref) | 0.04 | 0.87 (0.73–1.03) | 1 (ref) | 0.12 | |
| Recurrent MI | Incidence rate | 2.8 | 5.5 | 2.1 | 3.9 | ||
| Unadjusted HR | 0.51 (0.41–0.64) | 1 (ref) | <0.001 | 0.53 (0.40–0.69) | 1 (ref) | <0.001 | |
| Adjusted HR | 0.66 (0.51–0.87) | 1 (ref) | 0.003 | 0.74 (0.53–1.01) | 1 (ref) | 0.06 | |
| Hospitalization for new heart failure | Incidence rate | 2.4 | 3.3 | ||||
| Unadjusted HR | 0.72 (0.56–0.91) | 1 (ref) | 0.007 | 0.76 (0.55–1.04) | 1 (ref) | 0.08 | |
| Adjusted HR | 0.95 (0.73–1.24) | 1 (ref) | 0.72 | 1.04 (0.75–1.46) | 1 (ref) | 0.81 | |
| A composite outcomeb | Incidence rate | 13.6 | 24.2 | ||||
| Unadjusted HR | 0.55 (0.49–0.61) | 1 (ref) | <0.001 | 0.55 (0.48–0.63) | 1 (ref) | <0.001 | |
| Adjusted HR | 0.82 (0.73–0.92) | 1 (ref) | <0.001 | 0.85 (0.74–0.98) | 1 (ref) | 0.03 | |
Landmark analyses were performed among patients who were event-free for 2 years or 3 years after index MI, respectively. Incidence rates were presented as events per 1000 person-years.
HR, hazard ratio; MI, myocardial infarction; ref, reference.
Adjusted for comorbidities and medications at 2 years or 3 years after index MI, respectively.
A composite of all-cause death, recurrent MI, or hospitalization for new heart failure.
For 17 009 patients who were event-free during 3 years after index MI, the risk of composite of all-cause death, recurrent MI, or hospitalization for new HF was significantly lower in patients receiving persistent β-blocker therapy for ≥3 years (adjusted HR 0.85; 95% CI 0.74–0.98; P = 0.03) than those who were not.
Discussion
In this nationwide cohort study, β-blocker therapy for ≥1 year was associated with a lower risk of all-cause death compared with β-blocker for <1 year in patients without HF after AMI. The beneficial effects of β-blocker were consistent across various subgroups.
Very limited data are available on the effects of long-term β-blocker therapy on mortality in the reperfusion era. Among 824 and 556 patients who were followed-up at 1 and 3 years after AMI in a single-centre registry,11 the HR for all-cause mortality in patients with β-blocker therapy at 1 and 3 years after AMI were 0.85 (95% CI 0.53–1.36) and 0.89 (95% CI 0.44–1.78), respectively. However, this cohort included patients with HF who might have a worse prognosis despite long-term β-blocker therapy than those without HF not receiving long-term β-blocker therapy. In the FAST-MI registry (Nationwide French registry of Acute ST-and non-ST-elevation Myocardial Infarction) with data from 1383 patients with AMI and without HF or left ventricular dysfunction alive at 1 year,3 the 5-year mortality in patients who were still taking β-blockers at 1 year was 7.6%, compared with 9.2% in those who were not taking β-blockers at 1 year (adjusted HR 1.19; 95% CI 0.65–2.18). However, there were no data on β-blocker use beyond 1 year after the index MI and status of β-locker therapy from 1 year to 5 years could not be identified. It is known that a substantial number of patients discontinue β-blockers during follow-up,11 and β-blockers could also be restarted due to the development of HF during follow-up. Another study reporting β-blocker therapy beyond 3 years post-MI,12 conducted in patients ≥65 years of age including 21.8% of patients with reduced left ventricular ejection fraction and with a single assessment of β-blocker therapy, also found no association of β-blocker therapy with improved outcomes. On the other hand, our study was large enough to reliably estimate clinically relevant associations in a contemporary practice setting. Moreover, we were able to compare clinical outcomes by type of β-blocker, an issue for which there is very limited data available. Most patients in our study received relatively new types of β-blockers recommended by clinical guidelines for HF,20 and we found no significant differences across types of β-blockers.
There are several plausible explanations for the current results. First, a considerable number of AMI patients who were discharged from hospital without HF experience recurrent MI, or hospitalization or death because of HF during follow-up.21 , 22 Therefore, continuation of β-blocker as a secondary prevention would be beneficial even in patients without HF at the time of hospital discharge. Second, blood pressure might be better controlled in patients receiving long-term β-blocker therapy than those not. The control of blood pressure is an essential part of risk reduction therapy for patients with coronary artery disease,23 and intensive blood pressure control improves cardiovascular outcome in patients at high risk for cardiovascular events.24 Third, β-blocker therapy might slow progression of coronary atherosclerosis, as pooled analysis from intravascular ultrasound trials demonstrated that β-blocker therapy was associated with a regression of atheroma volume in patients with coronary artery disease.25
Limitations
Our study has limitations. First, we could not identify reasons for β-blocker use in this study. Patients receiving long-term β-blocker therapy might have less severe risk profiles to tolerate β-blocker therapy compared with those stopping β-blockers during the follow-up. To overcome this potential bias, we excluded patients who had not started their β-blocker therapy at early stage after AMI, and conducted landmark analysis to avoid immortal time bias. Second, there is a potential selection bias induced by unrecorded confounders. For example, we lacked information on type of AMI and angiographic severity, anthropometric and behavioural factors, and we had only limited information on disease management based on claims. Finally, there was no information in our study about left ventricular ejection fraction. However, the majority of patients experiencing MI in Korea have preserved left ventricular ejection fraction.26 Furthermore, since patients with low ejection fraction are likely to receive β-blocker therapy and have poor prognosis, this type of bias is unlikely to explain our findings.
Future directions
The CAPITAL-RCT (Carvedilol Post-Intervention Long-Term Administration in Large-scale Randomized Controlled Trial) was a multicentre, open-label, randomized controlled trial evaluating the efficacy of long-term oral beta-blocker therapy in preventing cardiovascular events in STEMI patients with preserved left ventricular systolic function who underwent primary PCI,13 but its sample size was relatively small (801 patients) and the event rate was low. This trial did not provide robust evidence for a definite conclusion. A large randomized clinical trial will be needed to investigate the effect and optimal duration of β-blocker therapy after MI in patients without HF or left ventricular systolic dysfunction who receive current guideline-based treatments.
Conclusions
Among patients with AMI but without HF, β-blocker therapy for ≥1 year after MI was associated with reduced mortality. However, a large randomized clinical trial will be needed to establish the optimal duration of routine β-blocker therapy after AMI.
Data availablity
The data from the Korean National Health Insurance Service are only available to authorized personnel and access to the data is restricted.
Supplementary material is available at European Heart Journal online.
Conflict of interest: none declared.
References
Author notes
Jihoon Kim and Danbee Kang contributed equally to this work.
