Abstract
Objective:
Our purpose was to examine the association between beta-blocker dose and mortality following acute myocardial infarction.
Methods:
This nationwide cohort study enrolled all patients admitted for first-time acute myocardial infarction in Denmark during 1 July 2004–31 December 2014 using the Danish National Patient Registry. Patients alive 15 days after admission were followed until death, emigration, or 31 December 2014. Patients were categorized according to daily beta-blocker consumption (0%, >0–12.5%, >12.5–25%, >25–50%, >50–100%, or >100% of the currently recommended target dose) based on prescriptions registered in the Danish National Database of Reimbursed Prescriptions. Doses were continuously updated during follow-up. Mortality rate ratios (MRRs) were computed and adjusted for confounders using Cox proportional hazard regression.
Results:
Among 65,125 patients followed, any beta-blocker dose was associated with significant mortality reduction compared with no treatment (adjusted MRR ≤ 0.92 [95% confidence interval (CI): 0.86–0.98]). The largest reduction was observed within the first year for beta-blocker doses >25–50% (adjusted MRR = 0.55 [95% CI: 0.50–0.60]). After one year, doses >50–100% were associated with the largest mortality reduction (adjusted MRR=0.58 [95% CI: 0.50–0.67]), but it did not differ significantly from that associated with doses >25–50% (adjusted MRR=0.68 [95% CI: 0.61–0.77]).
Conclusions:
Any beta-blocker dose was associated with significant mortality reduction following acute myocardial infarction compared with no treatment. Doses >25–50% of the currently recommended target dose were associated with maximal mortality reduction within the first year after acute myocardial infarction, suggesting that higher doses are unnecessary.
Keywords: Acute myocardial infarction, Beta-blocker, Dose, Mortality, Cohort study
Introduction
Randomized clinical trials (RCTs) have documented that beta-blocker therapy after acute myocardial infarction (AMI) improves survival, and contemporary guidelines recommend treatment.1,2 Based on RCTs conducted predominantly before acute revascularization was implemented, recommended target doses (RTD) of frequently used beta-blockers are: metoprolol 200 mg/day,3 carvedilol 50 mg/day,4 bisoprolol 10 mg/day,5 atenolol 100 mg/day,6 and propranolol 180 mg/day.7 Doses prescribed after acute myocardial infarction are often considerably lower,8,9,10,11,12,13,14 but the effects of different doses have not been assessed in RCTs.
Surprisingly, observational studies have shown significant mortality reduction with lower doses than used in RCTs.10,11,12,13,15,16 A study based on the SWEDEHEART registry demonstrated that beta-blocker doses ≥50% of RTD were not associated with superior cardiovascular outcomes up to 5 years after acute myocardial infarction as compared with doses <50% of RTD.15 In the OBTAIN study, the lowest mortality was observed among patients treated with 25% of RTD.10,16
Most observational studies examining the relationship between beta-blocker dose and mortality following acute myocardial infarction have been based on discharge doses, assuming that these represent actual doses taken and that doses remain stable after discharge. However, studies have shown low adherence to beta-blocker therapy among post-acute myocardial infarction patients,9,17,18,19 indicating that the extrapolation of discharge doses may cause misclassification in dose-mortality studies. In order to address this issue, we examined the association between beta-blocker dose and mortality following acute myocardial infarction, using actual rather than prescribed beta-blocker doses in an observational study.
Methods
Study design and setting
The study was a nationwide cohort study conducted in Denmark, with a source population of 5.6 million inhabitants and a tax-funded health care system, partially reimbursing most prescription medication expenses.20 A unique civil registration number of each inhabitant is a prerequisite for receiving health care and enables accurate linkage of national medical registries.20 Study data were derived from the Danish National Patient Registry (DNPR),21 Danish Civil Registration System (DCRS),20 and Danish National Database of Reimbursed Prescription (DNDRP).22 The online supplement includes additional registry information and codes used in the study. No ethics approval was needed.
Patients
All patients ≥18 years admitted for first-time acute myocardial infarction in Denmark between 1 July 2004 and 31 December 2014 were identified in the DNPR by primary in-patient diagnoses only (eTable 1). Outpatient and emergency room diagnoses were not included because of reported lower predictive values.23 Immigrants living in Denmark for <5 years were excluded because of the risk for incomplete data about comorbidity. Using the DCRS, patients alive on day 15 after admission were followed until first occurrence of death, emigration, or 31 December 2014.
eTable 1.
International Classification of Diseases codes (ICD-10) and Nordic Medico-Statistical Committee Classification of Surgical Procedures (NCSP) codes for diseases and procedures identified in the Danish National Patient Registry.
| Disease | ICD-10 |
|---|---|
| First-time myocardial infarction* | I21 |
| STEMI | I21.0;I21.1;I21.2;I21.3 |
| Non-STEMI | I21.4 |
| Unspecified | I21.9 |
| Insulin dependent diabetes mellitus | E10 |
| Non-insulin dependent diabetes mellitus | E11 |
| Hypercholesterolemia | E78.0; E78.2 |
| Hypertension | I10-I15; I67.4 |
| Atrial fibrillation or flutter | I48 |
| Atrioventricular or sinoatrial block | I44.0; I44.1; I44.2; I44.3; I45.5 |
| Sick sinus syndrome | I49.5 |
| Bradycardia | R00.1 |
| Hypotension | I95 |
| Cardiogenic shock | R57.0 |
| Valvular disease | I05; I06; I07; I08; I34; I35; I36; I37; I38; I39 |
| Cardiomyopathy | I42; I43 |
| Stroke | I60; I61; I62; I63; 64; I69.0; I69.1; I69.3; I69.4; G46 |
| Procedure | NCSP code |
| Percutaneous coronary intervention | KFNG; KFNF |
| Coronary artery bypass grafting | KFNA-E; KFNH20 |
No prior registration of myocardial infarction by ICD-8 diagnosis code 410 or ICD-10 diagnosis code I21.
STEMI = ST-elevation myocardial infarction
Comorbidity
Comorbidity was determined, using primary and secondary ICD-10 diagnosis codes registered in the DNPR prior to and during follow-up (eTable 1). Charlson Comorbidity Index scores (eTable 2), excluding diagnoses for acute myocardial infarction, were computed at start of follow-up and updated continuously, as comorbidities were considered present from the day of first registration.24,25 Emergency room diagnoses were not included. Four comorbidity categories were defined: low (0 points), moderate (1 point), severe (2 points), and very severe (≥3 points).
eTable 2.
The Charlson Comorbidity Index with corresponding International Classification of Diseases codes (ICD-10).
| 1 point | ICD-10 |
|---|---|
| Myocardial infarction | I21-I23 |
| Congestive heart failure | I50; I11.0; I13.0; I13.2 |
| Peripheral vascular disease | I70; I71; I72; I73; I74; I77 |
| Cerebrovascular disease | I60–I69; G45; G46 |
| Dementia | F00-F03; F05.1; G30 |
| Chronic pulmonary disease | J40–J47; J60–J67; J68.4; J70.1; J70.3; J84.1; J92.0; J96.1; J98.2; J98.3 |
| Connective tissue disease | M05; M06; M08; M09; M30; M31; M32; M33; M34; M35; M36; D86 |
| Ulcer disease | K22.1; K25–K28 |
| Mild liver disease | B18; K70.0–K70.3; K70.9; K71; K73; K74; K76.0 |
| Simple diabetes (without end-organ damage) | E10.0, E10.1; E10.9; E11.0; E11.1; E11.9 |
| 2 points | |
| Hemiplegia | G81; G82 |
| Moderate to severe renal disease | I12-I13; N00-N05; N07; N11; N14; N17-N19; Q61 |
| Complex diabetes (with end-organ damage) | E10.2-E10.8; E11.2-E11.8 |
| Non-metastatic solid tumor | C00–C75 |
| Leukemia | C91–C95 |
| Lymphoma | C81–C85; C88; C90; C96 |
| 3 points | |
| Moderate to severe liver disease | B15.0; B16.0; B16.2; B19.0; K70.4; K72; K76.6; I85 |
| 6 points | |
| Metastatic cancer | C76–C80 |
| AIDS | B21–B24 |
Beta-blocker treatment
Each patient’s daily beta-blocker consumption was determined using the DNDRP, i.e., all beta-blocker prescriptions redeemed between 1 January 2004 and 31 December 2014. Patients were categorized in beta-blocker dose groups: 0%, >0–12.5%, >12.5–25%, >25–50%, >50–100%, or ≥100% of RTD at start of follow-up and every time a prescription was redeemed during follow-up (details provided in online supplement), allowing for each patient to contribute to different dose groups at different times during follow-up.
Patients who redeemed a beta-blocker prescription before admission, were classified as “prior users”. Remaining patients were classified as “new users”. Prior users with beta-blocker prescription redemption ≤90 days before admission, but none between admission and start of follow-up, were categorized as taking an “uncertain dose” at start of follow-up until a new prescription was redeemed or for a maximum of 90 days after prescription redemption, whichever came first, as we assumed they had available beta-blocker from prior prescription redemption. If they did not redeem a new prescription ≤90 days, they were categorized as not taking beta-blocker (dose=0%) until prescription redemption or end of follow-up, whichever came first.
Remaining prior users and new users who had not redeemed a prescription after admission were categorized as taking an “uncertain dose” until three days after hospital discharge, as we assumed that many of these patients were discharged with a small beta-blocker supply. After three days, they were categorized as not taking beta-blocker (dose=0%) unless a prescription was redeemed.
Co-medication
Information about cardiac co-medication was obtained for all patients from the DNDRP (eTable 3). Use of co-medication was defined as ≥1 prescription redemption ≤6 months for the medication in question at any time during the study period.
eTable 3.
Beta-blockers and co-medication with doses and Anatomical Therapeutic Chemical (ATC) codes.
| Beta-Blocker | Target Dose | % of Target Dose (mg/day) | ATC code | ||||
|---|---|---|---|---|---|---|---|
| (mg/day) | >0–1.25 | >12.5–25 | >25–50 | >50–100 | >100–200 | ||
| Any | - | - | - | - | - | C07 | |
| Atenolol | 100 | >0–12.5 | >12.5–25 | >25–50 | >50–100 | >100–200 | C07AB03; C07CB03 |
| Bisoprolol | 10 | >0–1.25 | >1.25–2.5 | >2.5–5 | >5–10 | >10–20 | C07AB07 |
| Carvedilol | 50 | >0–6.25 | >6.25–12.5 | >12.5–25 | >25–50 | >50–100 | C07AG02 |
| Metoprolol | 200 | >0–25 | >25–50 | >50–100 | >100–200 | >200–400 | C07AB02; C07BB02 |
| Propranolol | 180 | >0–22.5 | >22.5–45 | >45–90 | >90–180 | >180–360 | C07AA05 |
| Other (pindolol, sotalol, nebivolol, labetalol) | - | - | - | - | - | C07AA03; C07AA07; C07AB12; C07AG01 | |
| Co-medication | |||||||
| Statins | - | - | - | - | - | C10AA | |
| Platelet inhibitors | - | - | - | - | - | B01AC04 | |
| B01AC06 | |||||||
| B01AC07 | |||||||
| B01AC22 | |||||||
| B01AC24 | |||||||
| B01AC30 | |||||||
| Angiotensin converting enzyme-or renin angiotensin inhibitors | - | - | - | - | - | C09 | |
Statistical Analysis
Patients were categorized according to gender, age, acute myocardial infarction type, cardiac procedures, comorbidity category, conditions of the Charlson Comorbidity Index, other comorbidities, and co-medication. Mortality rates were computed as the number of deaths per 1,000 person-years. Assuming that the underlying assumptions were met, we used Cox proportional hazards regression to compute hazard ratios as measures of mortality rate ratios (MRRs) within beta-blocker dose groups, setting 0% of RTD as reference. In order to reduce confounding, MRRs were adjusted for gender, age, acute myocardial infarction type, prior beta-blocker use, cardiac interventions, comorbidity category, and co-medication. Analyses were performed using SAS version 16.1 (SAS Institute Inc., Cary, NC, USA).
Results
Patient characteristics
A total of 71,359 patients were admitted for first-time acute myocardial infarction during the study period and 65,125 began follow-up (Figure 1). Median follow-up time was 3.6 (interquartile range [IQR]: 1.4–6.5) years.
Figure 1.
Study inclusion and follow-up.
Table 1 presents patient characteristics. New beta-blocker users constituted 51,320 (71.9%) of all patients admitted and 47,179 (72.4%) of patients followed (Table 1).
Table 1.
Patient characteristics at start of follow-up.
| Beta-blocker dose group, n (%)a | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Patient characteristic | Uncertai nb | 0% | >0–12.5% | >12.5–25% | >25–50% | >50–100% | Overall | ||
| Total | 16,066 (24.7) |
11,131 (17.1) |
9,334 (14.3) |
20,384 (31.3) |
6,873 (10.6) |
708 (1.1) |
65,125 (100.0) |
||
| Gender | |||||||||
| Male | 9,635 (23.2) |
6,494 (15.6) |
5,914 (14.2) |
13,928 (33.6) |
4,778 (11.5) |
469 (1.1) |
41,505 (100.0) |
||
| Female | 6,431 (27.2) |
4,637 (19.6) |
3,420 (14.5) |
6,456 (27.3) |
2,095 (8.9) |
239 (1.0) |
23,620 (100.0) |
||
| Age in years | |||||||||
| <60 | 2,416 (13.5) |
2,960 (16.5) |
2,726 (15.2) |
7,146 (39.9) |
2,380 (13.3) |
219 (1.2) |
17,912 (100.0) |
||
| 60–64 | 1,638 (20.2) |
1,199 (14.8) |
1,252 (15.4) |
2,920 (36.0) |
979 (12.1) |
95 (1.2) |
8,117 (100.0) |
||
| 65–69 | 2,141 (25.4) |
1,247 (14.8) |
1,229 (14.6) |
2,738 (32.5) |
930 (11.0) |
97 (1.2) |
8,433 (100.0) |
||
| 70–74 | 2,472 (29.8) |
1,349 (16.2) |
1,110 (13.4) |
2,388 (28.8) |
831 (10.0) |
89 (1.1) |
8,304 (100.0) |
||
| 75–79 | 2,638 (33.6) |
1,276 (16.3) |
1,010 (12.9) |
2,100 (26.7) |
670 (8.5) |
76 (1.0) |
7,852 (100.0) |
||
| ≥80 | 4,761 (32.8) |
3,100 (21.4) |
2,007 (13.8) |
3,092 (21.3) |
1,083 (7.5) |
132 (0.9) |
14,507 (100.0) |
||
| Type of MI | |||||||||
| STEMI | 4,075 (16.9) |
3,154 (13.1) |
4,684 (19.5) |
9,037 (37.6) |
2,723 (11.3) |
244 (1.0) |
24,063 (100.0) |
||
| Non-STEMI | 8,646 (29.1) |
5,370 (18.1) |
3,455 (11.6) |
8,564 (28.8) |
3,026 (10.2) |
338 (1.1) |
29,722 (100.0) |
||
| Unspecifiedc | 3,345 (29.5) |
2,607 (23.0) |
1,195 (10.5) |
2,783 (24.5) |
1,124 (9.9) |
126 (1.1) |
11,340 (100.0) |
||
| Beta-blocker use | |||||||||
| Prior user | 8,330 (46.4) |
2,456 (13.7) |
1,342 (7.5) |
3,180 (17.7) |
1,820 (10.1) |
369 (2.1) |
17,946 (100.0) |
||
| New user | 7,736 (16.4) |
8,675 (18.4) |
7,992 (16.9) |
17,204 (36.5) |
5,053 (10.7) |
339 (0.7) |
47,179(100.0) | ||
| Cardiac interventions | |||||||||
| Percutaneous coronary intervention | 1,994 (21.1) |
1,275 (13.5) |
1,509 (15.9) |
3,220 (34.0) |
1,257 (13.3) |
136(1.4) | 9.465 (100.0) |
||
| Coronary artery bypass grafting | 985 (67.3) |
160 (10.9) |
85 (5.8) |
159 (10.9) |
49 (3.3) |
11 (0.8) |
1,463 (100.0) |
||
| Comorbidity category | |||||||||
| Low | 5,188 | 4,950 | 4,666 | 12,228 | 3,720 | 325 | 31,234 | ||
| (16.6) | (15.8) | (14.9) | (39.1) | (11.9) | (1.0) | (100.0) | |||
| Moderate | 3,954 (26.5) |
2,613 (17.5) |
2,392 (16.0) |
4,199 (28.1) |
1,494 (10.0) |
142 (0.9) |
14,948 (100.0) |
||
| Severe | 2,582 (32.1) |
1,462 (18.2) |
1,081 (13.4) |
1,959 (24.3) |
750 (9.3) | 100 (1.2) |
8,052 (100.0) |
||
| Very severe | 4,342 (39.9) |
2,106 (19.3) |
1,195 (11.0) |
1,998 (18.3) |
909 (8.3) | 141 (1.3) |
10,891 (100.0) |
||
| Charlson diseases | |||||||||
| Heart failure | 4,497 (36.1) |
1,732 (13.9) |
2,127 (17.1) |
2,658 (21.3) |
1,093 (8.8) |
149 (1.2) |
12,461 (100.0) |
||
| Peripheral vascular disease | 2,236 (40.2) |
995 (17.9) |
588 (10.6) |
1,096 (19.7) |
483 (8.7) | 80 (1.4) | 5,558 (100.0) |
||
| Cerebrovascular disease | 2,647 (37.1) |
1,403 (19.7) |
794 (11.1) |
1,513 (21.2) |
533 (7.5) | 86 (1.2) | 7,126 (100.0) |
||
| Chronic pulmonary disease | 2,493 (32.7) |
1,933 (25.4) |
906 (11.9) |
1,440 (18.9) |
673 (8.8) | 69 (0.9) | 7,620 (100.0) |
||
| Connective tissue disease | 887 (31.2) |
576 (20.3) |
370 (13.0) |
700 (24.6) |
248 (8.7) | 31 (1.1) | 2,840 (100.0) |
||
| Ulcer disease | 375 (40.0) |
185 (19.5) |
105 (11.1) |
170 (17.9) |
87 (9.2) | 11 (1.2) | 949 (100.0) | ||
| Simple diabetes | 2,897 (35.7) |
1,452 (17.9) |
842 (10.4) |
1,820 (22.4) |
838 (10.3) | 120 (1.5) |
8,118 (100.0) |
||
| Complex diabetes | 1,990 (39.7) |
889 (17.7) |
497 (9.9) | 977 (19.5) |
477 (9.5) | 79 (1.6) | 5,012 (100.0) |
||
| Moderate/severe renal disease | 1,365 (50.9) |
480 (17.9) |
200 (7.5) | 370 (13.8) |
198 (7.4) | 33 (1.2) | 2,680 (100.0) |
||
| Non-metastatic solid Tumor | 1,565 (30.1) |
1,044 (20.1) |
743 (14.3) |
1,284 (24.7) |
441 (8.5) | 52 (1.0) | 5,203 (100.0) |
||
| Other diseases | |||||||||
| Insulin dependendent diabetes mellitus | 1,376 (40.4) |
630 (18.5) |
325 (9.5) | 668 (19.6) |
306 (9.0) | 45 (1.3) | 3,407 (100.0) |
||
| Non-insulin dependent diabetes mellitus | 2,975 (35.9) |
1,441 (17.4) |
868 (10.5) |
1,843 (22.2) |
871 (10.5) | 128 (1.5) |
8,286 (100.0) |
||
| Hypercholesterolemia | 4,701 (23.7) |
2,975 (15.0) |
2,805 (14.2) |
6,316 (31.9) |
2,580 (13.0) |
262 (1.3) |
19,806 (100.0) |
||
| Hypertension | 848 (42.4) |
373 (18.7) |
173 (8.7) | 333 (16.7) |
201 (10.1) | 36 (1.8) | 2,000 (100.0) |
||
| Atrial fibrillation/flutter | 3,190 (42.6) |
1,268 (16.9) |
691 (9.2) | 1,349 (18.0) |
728 (9.7) | 110 (1.5) |
7,486 (100.0) |
||
| Valvular heart disease | 1,899 (40.4) |
927 (19.7) |
467 (9.9) | 904 (19.2) |
366 (7.8) | 56 (1.2) | 4,697 (100.0) |
||
| Cardiomyopathy | 477 (40.4) |
192 (16.2) |
148 (12.5) |
219 (18.5) |
105 (8.9) | 24 (2.0) | 1,182 (100.0) |
||
| Stroke | 2,021 (38.9) |
1,024 (19.7) |
560 (10.8) |
1,049 (20.2) |
353 (6.8) | 62 (1.2) | 5,189 (100.0) |
||
| Co-medication | |||||||||
| Statins | 7,096 (15.1) |
7,029 (15.0) |
8,038 (17.1) |
17,920 (38.1) |
5,871 (12.5) |
579 (1.2) |
46,973 (100.0) |
||
| Platelet inhibitors | 8,186 (16.2) |
8,253 (16.3) |
8,528 (16.9) |
18,333 (36.3) |
6,007 (11.9) |
616 (1.2) |
50,488 (100.0) |
||
| Angiotensin converting enzyme- or renin angiotensin inhibitors | 7,403 (23.2) |
4,866 (15.2) |
4,892 (15.3) |
10,038 (31.4) |
3,871 (12.1) |
472 (1.5) |
31,925 (100.0) |
||
Patients categorized as taking other beta-blockers (n =60), unknown dose (n = 494), or >100% of target dose (n = 75) not included. Charlson diseases with a prevalence < 1% not included (dementia, hemiplegia, mild liver disease, moderate to severe liver disease, leukemia, lymphoma, metastatic cancer, and AIDS). STEMI = ST-elevation myocardial infarction.
Percentage of the recommended target dose.
Prior users with a beta-blocker prescription redemption ≤90 days before admission, but none between admission and start of follow-up, were categorized as taking an “uncertain dose” at start of follow-up until a new prescription was redeemed or for a maximum of 90 days after prescription redemption. Remaining prior users and new users who had not redeemed a prescription after admission were categorized as taking an “uncertain dose” until three days after hospital discharge.
Not specified as STEMI or non-STEMI
For patients who redeemed a beta-blocker prescription after admission, the median number of days from hospital discharge to first prescription redemption was 1 (IQR: 0–4 days). One month after discharge, 13,652 (26.6%) of potential new beta-blocker users had still not redeemed a prescription.
At start of follow-up, 31.3% of patients took beta-blocker doses >12.5–25% of RTD, while 1.1% took daily doses >50% of RTD, and 17.1% were not taking a beta-blocker (Table 1). Metoprolol was the most commonly used beta-blocker (86.8%), followed by carvedilol (7.1%), bisoprolol (5.3%), atenolol (0.4%), and propranolol (0.2%). Only 0.2% were other beta-blockers.
Overall mortality according to beta-blocker treatment
A total of 6,014 (8.4%) patients died ≤14 days after admission (eTable 4 presents patient characteristics), and 18,913 (29.0%) patients died during follow-up (Figure 1), corresponding to an overall mortality rate during follow-up of 70.9 (95% CI: 69.9–71.9) per 1,000 person-years. Patients not taking beta-blockers had a significantly higher mortality rate during 10-year-follow-up than patients taking any beta-blocker dose, and MRRs remained significantly lower for patients taking any beta-blocker dose after adjustment for potential confounders (Table 2).
eTable 4.
Characteristics for patients not followed, i.e., patients who died ≤14 days after admission or were admitted ≤14 days of end of the study period.
| n (%) | |
|---|---|
| Total | 6,234 (100) |
| Gender | |
| Male | 3,239 (52.0) |
| Female | 2,995 (48.0) |
| Age in years | |
| <60 | 448 (7.2) |
| 60–64 | 363 (5.8) |
| 65–69 | 457 (7.3) |
| 70–74 | 644 (10.3) |
| 75–79 | 945 (15.2) |
| ≥80 | 3,377 (54.2) |
| Type of myocardial infarction | |
| STEMI | 2,146 (34.4) |
| Non-STEMI | 1,308 (21.0) |
| Unspecified | 2,780 (44.6) |
| Beta-blocker use | |
| Previous | 2,093 (33.6) |
| No previous | 4,141 (66.4) |
| Cardiac interventions | |
| Percutaneous coronary intervention | 250 (4.0) |
| Coronary artery bypass grafting | 104 (1.7) |
| Comorbidity category | |
| Low | 1,876 (30.1) |
| Moderate | 1,549 (24.8) |
| Severe | 1,046 (16.8) |
| Very severe | 1,763 (28.3) |
| Charlson diseases | |
| Heart failure | 1,556 (25.0) |
| Peripheral vascular disease | 867 (13.9) |
| Cerebrovascular disease | 1,242 (19.9) |
| Dementia | 183 (2.9) |
| Chronic pulmonary disease | 1,006 (16.1) |
| Connective tissue disease | 346 (5.6) |
| Ulcer disease | 202 (3.2) |
| Mild liver disease | 82 (1.3) |
| Simple diabetes | 932 (15.0) |
| Complex diabetes | 674 (10.8) |
| Moderate/severe renal disease | 522 (8.4) |
| Hemiplegia | 37 (0.6) |
| Non-metastatic solid tumor | 797 (12.8) |
| Leukemia | 44 (0.7) |
| Lymphoma | 80 (1.3) |
| Moderate/severe liver disease | 38 (0.6) |
| Metastatic cancer | 105 (1.7) |
| AIDS | 4 (0.06) |
| Other diseases | |
| Insulin-dependent diabetes mellitus | 472 (7.6) |
| Non-insulin-dependent diabetes mellitus | 971 (15.6) |
| Hypercholesterolemia | 640 (10.3) |
| Hypertension | 241 (3.9) |
| Atrial fibrillation/flutter | 1,117 (17.9) |
| Valvular heart disease | 700 (11.2) |
| Cardiomyopathy | 88 (1.4) |
| n (%) | |
| Stroke | 964 (15.5) |
| Co-medication | |
| Statins | 1,545 (24.8) |
| Platelet inhibitors | 2,500 (40.1) |
| Angiotensin converting enzyme- or renin angiotensin inhibitors | 2,203 (35.3) |
AIDS = acquired immune deficiency syndrome; STEMI = ST-elevation myocardial infarction.
Table 2.
Mortality rates and rate ratios following acute myocardial infarction according to daily beta-blocker dose.
| Beta-blocker dose group, % of recommended target dose. | ||||||
|---|---|---|---|---|---|---|
| 0 | >0–12.5 | >12.5–25 | >25–50 | >50–100 | >100 | |
| Year 1 | ||||||
| Deaths, n | 3,018 | 445 | 902 | 672 | 386 | 97 |
| MR† (95% CI) | 163.9 (158.1–169.8) | 155.7 (141.8–170.8) | 83.9 (78.6–89.5) | 54.8 (50.8–59.1) | 59.6 (53.9–65.9) | 60.2 (49.3–73.4) |
| MRR (95% CI) | Reference | 0.68 (0.62–0.76) | 0.42 (0.38–0.45) | 0.32 (0.30–0.35) | 0.38 (0.35–0.43) | 0.39 (0.32–0.48) |
| MRRa (95% CI) | Reference | 0.91 (0.82–1.00) | 0.70 (0.65–0.76) | 0.55 (0.50–0.60) | 0.62 (0.55–0.69) | 0.54 (0.44–0.67) |
| Year 2 | ||||||
| Deaths, n | 1,515 | 148 | 325 | 385 | 225 | 77 |
| MR† (95% CI) | 84.1 (80.0–88.4) | 67.3 (57.3–79.1) | 41.8 (37.5–46.6) | 40.2 (36.4–44.5) | 35.7 (31.3–40.7) | 40.9 (32.8–51.2) |
| MRR (95% CI) | Reference | 0.80 (0.68–0.95) | 0.50 (0.44–0.56) | 0.48 (0.43–0.53) | 0.42 (0.37–0.49) | 0.49 (0.39–0.61) |
| MRRa (95% CI) | Reference | 0.91 (0.77–1.08) | 0.72 (0.64–0.82) | 0.68 (0.61–0.77) | 0.58 (0.50–0.67) | 0.60 (0.47–0.75) |
| Year 3 | ||||||
| Deaths, n | 1,158 | 119 | 238 | 267 | 182 | 64 |
| MR† (95% CI) | 74.3 (70.1–78.7) | 62.9 (52.5–75.2) | 38.8 (34.2–44.1) | 34.3 (30.5–38.7) | 34.5 (29.9–39.9) | 39.5 (30.9–50.4) |
| MRR (95% CI) | Reference | 0.85 (0.70–1.02) | 0.52 (0.45–0.60) | 0.46 (0.40–0.53) | 0.46 (0.40–0.54) | 0.53 (0.41–0.68) |
| MRRa (95% CI) | Reference | 0.95 (0.79–1.15) | 0.75 (0.65–0.86) | 0.65 (0.57–0.74) | 0.61 (0.52–0.71) | 0.60 (0.46–0.77) |
| Year 4–5 | ||||||
| Deaths, n | 1,620 | 166 | 315 | 368 | 277 | 109 |
| MR† (95% CI) | 67.3 (64.1–70.6) | 58.8 (50.5–68.4) | 36.6 (32.8–40.9) | 32.5 (29.3–35.9) | 34.9 (31.0–39.3) | 43.9 (36.4–52.9) |
| MRR (95% CI) | Reference | 0.87 (0.74–1.02) | 0.54 (0.48–0.61) | 0.48 (0.43–0.54) | 0.52 (0.46–0.59) | 0.65 (0.54–0.79) |
| MRRa (95% CI) | Reference | 0.99 (0.84–1.16) | 0.76 (0.67–0.86) | 0.67 (0.59–0.75) | 0.66 (0.58–0.75) | 0.73 (0.60–0.88) |
| Year 1–10 | ||||||
| Deaths, n | 8,899 | 1,016 | 2,057 | 2,107 | 1,342 | 451 |
| MR† (95% CI) | 88.3 (86.5–90.2) | 80.9 (76.1–86.0) | 50.1 (48.0–52.3) | 40.7 (39.0–42.4) | 39.3 (37.3–41.5) | 43.6 (39.8–47.8) |
| MRR (95% CI) | Reference | 0.77 (0.72–0.83) | 0.48 (0.46–0.51) | 0.43 (0.41–0.46) | 0.45 (0.42–0.48) | 0.51 (0.47–0.57) |
| MRRa (95% CI) | Reference | 0.92 (0.86–0.98) | 0.73 (0.69–0.76) | 0.65 (0.62–0.68) | 0.63 (0.59–0.66) | 0.61 (0.55–0.67) |
Per 1,000 person-years.
Adjusted for gender, age, type of myocardial infarction, beta-blocker use prior to myocardial infarction, cardiac interventions, comorbidity category, and co-medication.
CI = confidence interval; MR = mortality rate; MRR = mortality rate ratio.
The lowest overall MRR was observed for patients taking >25–50% of RTD. After adjustment for potential confounders the ratio did not differ significantly among daily doses >25–50% of RTD (Table 2).
Mortality according to follow-up time and beta-blocker dose
The mortality rate decreased during follow-up from 126.6 (95% CI: 123.7–129.6) during year 1 to 61.9 (95% CI: 59.8–64.2) during year 2, 57.3 (95% CI: 55.1–59.7) during year 3, and 54.5 (95% CI: 52.6–56.3) per 1,000 person-years during years 4–5 (Table 2). Mortality rate ratios were particularly low for patients taking any beta-blocker dose within the first year of follow-up, during which ratios also decreased significantly with increasing doses up to 50% of RTD.
The lowest MRR was observed within the first year of follow-up for patients taking >25–50% of RTD and remained almost half of that for patients not taking beta-blockers after adjustment for potential confounders (Table 2 and Figure 2).
Figure 2.
Adjusted mortality rate ratios according to daily beta-blocker dose during year 1 and years 1–10 after acute myocardial infarction. Whisker bars illustrate upper and lower 95% confidence intervals.
*Adjusted for gender, age, type of myocardial infarction, beta-blocker use prior to myocardial infarction, cardiac interventions, comorbidity category, and co-medication. ref = reference.
Doses >50% of RTD were not associated with further significant reductions in MRRs, neither within the first year of follow-up, nor later, and adjustment for potential confounders did not change this finding (Table 2 and Figure 2).
Discussion
This nationwide study of beta-blocker treatment following acute myocardial infarction in the modern era showed that any beta-blocker dose was associated with significant mortality reduction compared with no beta-blocker. Maximal mortality reduction was observed with doses below RTD, particularly within the first year after acute myocardial infarction where doses >25–50% of RTD were associated with an almost 50% lower mortality compared with no treatment. Doses >50% of RTD were not associated with further significant reductions in MRRs, neither within the first year of follow-up, nor later.
Most RCT data supporting the use of beta-blockers after acute myocardial infarction was accumulated in the late 1970s and early 1980s. As such, there are many contemporary questions about the use of beta-blockers in the era of modern medical management for acute myocardial infarction which includes therapies that did not exist at that time, such as acute revascularization therapies and statins. The major question is whether beta-blockers confer any benefit in the modern era, and results of ongoing RCTs, e.g. DANBLOCK26 and REBOOT,27 are highly anticipated. Our study confirms previous reports of significant mortality reduction associated with beta-blockers following acute myocardial infarction.10,12,134,15,16,28,29
Another important question is the optimal dose required to achieve beta-blocker benefit. The initial RCTs generally titrated patients to high target doses. Several recent observational studies suggest that this maximal target dose is not required to achieve maximal benefit from beta-blocker therapy.10,11,16 The present study is among the largest, to date, to confirm this observation, and the only one to confirm of actual, rather than prescribed, therapy based on prescription refills. In accordance with the SWEDEHEART study,15 our study demonstrated that higher beta-blocker doses are not associated with improved outcomes. Furthermore, our study included a refined dose categorization, allowing for identification of significant differences in mortality between individual dose groups. There is therefore now a large body of observational data showing that maximal mortality reduction is achieved with beta-blocker doses below RTD used in the RCTs establishing their efficacy. In particular, a beta-blocker dose >25–50% of RTD was associated with an almost 50% lower mortality than no beta-blocker treatment one year after acute myocardial infarction. These findings are reassuring because of consistent reports of low-dose beta-blocker treatment at discharge and infrequent up-titration.8–19 However, our findings also underscore the need for reinforced efforts to increase the proportion of patients taking any beta-blocker following acute myocardial infarction, as one in four potential new beta-blocker users was alive one month after discharge, but had not redeemed a prescription and was thus unlikely to receive any beta-blocker treatment.
Beta-blocker dose and comorbidity
While beta-blockers appear beneficial in the modern era, it is unknown whether all patients with acute myocardial infarction benefit. For example, patients with preserved left ventricular function may benefit less and RCTs examining the effect of beta-blocker therapy in these patients are ongoing.26,27 In our study, we found that low-dose beta-blocker treatment following acute myocardial infarction remained associated with significantly lower mortality after adjustment for differences in both prevalence of heart failure and other suggested confounders, e.g. percutaneous coronary intervention (PCI),30 type of acute myocardial infarction, and co-medication.17 In contrast, Grall et al. previously found no survival benefit of beta-blockers up to one year after acute myocardial infarction without heart failure.12 This difference in study results may reflect, at least in part, differences in determination of beta-blocker use. While our study relied on actually redeemed prescriptions during the study period, the study by Grall et al. relied on hospital discharge prescriptions and doses reported at 6-months and 1-year follow-up, respectively.12 Patient reports are susceptible to recall bias and studies of compliance among post-MI patients have shown low adherence to prescribed medicine.8,17,18,19 At one-year follow-up, <70% of acute myocardial infarction patients adhere to beta-blocker therapy,17 suggesting that differential misclassification may attenuate a beneficial effect of beta-blocker treatment in studies not based on actually redeemed prescriptions.
The potential bias associated with extrapolation of discharge beta-blocker doses during long-term follow-up was also demonstrated in a study of 192,746 acute myocardial infarction survivors, as 27% filled no prescriptions for beta-blockers within 30 days after discharge, and 24% of those who initially filled a prescription were non-adherent to beta-blockers 180 days after hospital discharge.18 Low adherence to beta-blocker therapy among post-acute myocardial infarction patients necessitates data on actual beta-blocker use after discharge in studies examining dose-mortality associations.
Duration of therapy
The recommended duration of therapy with beta-blockers has not been rigorously evaluated, but guidelines generally recommend long-term treatment. Several one-year landmark analyses suggest no benefit when beta-blocker dose is not accounted for. In one report, benefit was noted for lower doses (>12.5–25% of RTD), while another 3-year landmark analysis found no difference in outcome among those not treated with beta-blockers versus either low or high dose therapy. The present study provides one of the largest analyses of patients treated with beta-blockers showing continued dose-dependent benefit of beta-blockers for at least 5 years after acute myocardial infarction, with no excess benefit at doses exceeding >25–50% of RTD.
Strengths and limitations
Our study was nationwide, including a much larger cohort than most previous studies of beta-blocker dose and mortality post-acute myocardial infarction. The use of national registries in a country with tax-funded universal health care minimized selection bias, and prospective data recording eliminated recall bias. All-cause mortality data from the DCRS is complete,20 and reported positive predictive values of diagnoses in the DNPR are high, i.e., 92% for acute myocardial infarction 23 and 98% for comorbidities in the Charlson Comorbidity Index.23 Data completeness for cardiovascular medication in the DNDRP is very high.22 Cross-linking of these administrative databases provided not only completeness of high quality data, but also virtually complete long-term follow-up over a 10-year period, which is longer than most previous studies of beta-blocker dose and mortality after acute myocardial infarction.
The most important limitation of our study is its observational design, precluding causal inference. Another limitation is that although we adjusted for several potential confounders, residual confounding may exist. Clinical data such as left ventricular ejection fraction could have strengthened the study, but were not available in the registries. Instead, statistical adjustments were made for presence of diagnosed heart failure. With regards to the registration of PCI, a positive predictive value of 98% has been reported for procedure codes in the DNPR,31 but the low number of PCI registered in our study group, may reflect a lower negative predictive value. Interpretation of our results should also take into account that daily beta-blocker doses were not directly available in the DNDRP, but were computed based on certain assumptions. Still, the use of redeemed prescriptions, reflecting actual beta-blocker intake, instead of discharge prescriptions is a major strength, particularly because beta-blockers are prescription only in Denmark.
Conclusions
We showed that any beta-blocker dose was associated with significantly lower mortality than no beta-blocker, and that maximal mortality reduction was associated with beta-blocker doses below RTD based on the initial RCTs documenting their efficacy. Our findings support continued use of beta-blocker therapy following acute myocardial infarction to all patients in the absence of contraindications, and indicate that maximal mortality reduction may be achieved with mid-dose therapy, potentially reducing health care costs and beta-blocker side effects. Randomized controlled trials are needed to resolve the clinical questions remaining about the use of beta-blockers following acute myocardial infarction in the modern era of acute myocardial infarction therapeutics.
ClinicalSignificance.
Any beta-blocker dose was associated with significant mortality reduction after AMI
Doses >25–50% of target dose were associated with maximal mortality reduction
Higher doses may be unnecessary, reducing health care costs and side effects
Funding sources
Dr. Goldberger was supported by grant #5U01HL080416 from the National Heart, Lung, and Blood Institute of the National Institutes of Health and the Miami Heart Research Institute. None of the funders had any role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, approval of the manuscript, or decision to submit the manuscript for publication.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Declaration of Competing Interest
none.
Data sources
The Danish National Patient Registry maintains records of all hospitalizations performed in Denmark since 1977. Upon hospital discharge, the treating physician records a primary diagnosis describing the main reason for diagnostic work-up and treatment, and up to several secondary diagnoses describing comorbid conditions. Diagnoses are coded according to the World Health Organization’s International Classification of Diseases, Eighth Revision (ICD-8) before 1993 and Tenth Revision (ICD-10) thereafter. The Danish National Database of Reimbursed Prescriptions contains electronic records of all prescriptions redeemed in Denmark since 1 January 2004. Information about the person redeeming the prescription, including civil registration number, is registered, and a barcode identifier on each medication package enables automatic registration of the Anatomical Therapeutic Chemical (ATC) code, strength, pack size, pack quantity, and date of redemption.
The Danish Civil Registration System has tracked all changes in vital status for all Danish inhabitants since 1968 with daily electronic updates, and the information is transferred from the registry to electronic medical patient files on a daily basis.
Beta-blocker dose
From the day a first beta-blocker prescription was redeemed after admission, each patients’ daily beta-blocker dose was assumed to be one pill per day for metoprolol, atenolol, and bisoprolol, two pills per day for carvedilol, and three pills per day for propranolol until a second prescription was redeemed. If no redemption occurred within a number of days corresponding to the number of daily doses picked up at the pharmacy, the daily dose was assumed to be 0 mg until prescription redemption.
When a second beta-blocker prescription was redeemed, the daily dose was computed as the prescription content (i.e. strength of pills x pack size x pack quantity) divided by the number of days to the third prescription redemption, etc. The final daily dose during follow-up was computed as the content of the before-last prescription redeemed, divided by the number of days between the before-last and the last prescription redeemed. The daily dose was assumed constant between two prescription redemptions. When a patient changed beta-blocker type, the latest daily dose before the change was categorized as “unknown”, because treatment was assumed likely interrupted due to side effects and a new daily dose was not computed until two consecutive prescriptions for the same beta-blocker had been redeemed. Similarly, daily doses were categorized as “unknown” in case of a registered pack number <1, as this signified dose dispensing made by the pharmacy, e.g. for patients in a nursing home, making daily dose approximation impossible.
Patients, who redeemed prescriptions for beta-blockers other than metoprolol, carvedilol, bisoprolol, atenolol, or propranolol, i.e., pindolol, sotalol, nebivolol, or labetalol, were categorized as taking “other beta-blockers” from the day of redemption and were not categorized according to daily beta-blocker dose because the target doses for these beta-blockers have not been defined by randomized clinical trials.
References
- 1.Amsterdam EA, Wenger NK, Brindis RG, et al. 2014 AHA/ACC Guideline for the management of patients with non-ST-elevation acute coronary syndromes. JACC 2014;64(24):e139–228. [DOI] [PubMed] [Google Scholar]
- 2.O’Gara PT, Kushner FG, Casey DE, et al. 2013 ACCF/AHA Guideline for the management of ST-elevation myocardial infarction. JACC 2013;61(4)e78–140. [DOI] [PubMed] [Google Scholar]
- 3.Hjalmarson A, Elmfeldt D, Herlitz J, et al. Effect on mortality of metoprolol in acute myocardial infarction: A double-blind randomized trial. Lancet 1981;2:823–827. [DOI] [PubMed] [Google Scholar]
- 4.The CAPRICORN Investigators. Effect of carvedilol on outcome after myocardial infarction in patients with left-ventricular dysfunction: the CAPRICORN randomised trial. Lancet 2001;357:1385–1390. [DOI] [PubMed] [Google Scholar]
- 5.CIBIS-II Investigators and Committees. The Cardiac Insufficiency Bisoprolol Study II (CIBIS II): A randomized trial. Lancet 1999;353:9–13. [PubMed] [Google Scholar]
- 6.Wilcox RG, Roland JM, Banks DC, et al. Randomized trial comparing propranolol with atenolol in immediate treatment of suspected myocardial infarction. Brit Med J 1980;885–888. [DOI] [PMC free article] [PubMed]
- 7.Β-blocker Heart Attack Trial Research Group. A randomized trial of propranolol in patients with acute myocardial infarction. I. Mortality results. JAMA 1982;247:1707–14. [DOI] [PubMed] [Google Scholar]
- 8.Goldberger JJ, Bonow RO, Cuffe M, et al. Β-blocker use following myocardial infarction: Low prevalence of evidence-based dosing. Am Heart J 2010;160:435–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Arnold SV, Spertus JA, Masoudi FA, et al. Beyond medication prescription as performance measures. J Am Coll Cardiol 2013;62:1791–1801. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Goldberger JJ, Bonow RO, Cuffe M, et al. Effect of beta-blocker dose on survival after acute myocardial infarction. J Am Coll Cardiol 2015;66:1431–1441. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Grall S, Biere L, Nezet ML, et al. Relationship between beta-blocker and angiotensin-converting enzyme inhibitor dose and clinical outcome following acute myocardial infarction. Circ J 2015;79:632–640. [DOI] [PubMed] [Google Scholar]
- 12.Li C, Sun Y, Shen X, et al. Relationship between beta-blocker therapy at discharge and clinical outcomes in patients with acute coronary syndrome undergoing percutaneous coronary intervention. J Am Heart Assoc 2016;5:pii: e004190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Pedersen SB, Nielsen JC, Bøtker HE, et al. Beta-blocker therapy early after myocardial infarction: a comparison between medication at hospital discharge and subsequent pharmacy-dispensed medication. Drugs – Real World Outcomes 2016;3:279–288. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Allen JE, Knight S, McCubrey RO, et al. Beta-blocker dosage and outcomes after acute coronary syndrome. Am Heart J 2017;184:26–36. [DOI] [PubMed] [Google Scholar]
- 15.Mars K, Wallert J, Held C, et al. Association between beta-blocker dose and cardiovascular outcomes after myocardial infarction: insights from the SWEDEHEART registry. Eur Heart J Acute Cardiovasc Care 2021;10:372–379. [DOI] [PubMed] [Google Scholar]
- 16.Goldberger JJ, Subačius H, Marroquin O, et al. , OBTAIN (Outcomes of Beta-Blocker Therapy After Myocardial Infarction) Investigators. One-year landmark analysis of the effect of beta-blocker dose on survival after acute myocardial infarction. J Am Heart Assoc 2021;10(14):e019017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Korhonen MJ, Robinson JG, Annis IE, et al. Adherence tradeoff to multiple preventive therapies and all-cause mortality after acute myocardial infarction. J Am Coll Cardiol 2017; 70: 1543–1554. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Butler J, Arbogast PG, BeLue R, et al. Outpatient adherence to beta-blocker therapy after acute myocardial infarction. J Am Coll Cardiol 2002;40:1589–1595. [DOI] [PubMed] [Google Scholar]
- 19.Desta L, Khedri M, Jernberg T, et al. Adherence to beta-blockers and long-term risk of heart failure and mortality after a myocardial infarction. Eur Soc Cardiol Heart Fail 2021;8:344–355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Schmidt M, Pedersen L, Sørensen HT. The Danish Civil Registration System as a tool in epidemiology. Eur J Epidemiol 2014;29:541–549. [DOI] [PubMed] [Google Scholar]
- 21.Schmidt M, Schmidt SA, Sandegaard JL, et al. The Danish National Patient Registry: a review of content, data quality, and research potential. J Clin Epidemiol 2015;7:449–490. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Johannesdottir SA, Horváth-Puhó E, Ehrenstein V, et al. Existing data sources for clinical epidemiology: the Danish National Database of Reimbursed Prescriptions. J Clin Epidemiol 2012;4:303–313. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Joensen AM, Jensen MK, Overvad K, et al. Predictive values of acute coronary syndrome discharge diagnoses differed in the Danish National Patient Registry. J Clin Epidemiol 2009;62:188–194. [DOI] [PubMed] [Google Scholar]
- 24.Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373–383. [DOI] [PubMed] [Google Scholar]
- 25.Thygesen SK, Christiansen CF, Christensen S, et al. The predictive value of ICD-10 diagnostic coding used to assess Charlson comorbidity index conditions in the population-based Danish National Registry of Patients. BMC Med Res Methodol 2011;11:83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kristensen AMD, Bovin A, Zwisler AD, et al. Design and rationale of the Danish trial of beta-blocker treatment after myocardial infarction without reduced ejection fraction: study protocol for a randomized controlled trial. Trials 2020;21(1):415. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Rosello X, Raposeiras-Roubin S, Latini R, et al. Rationale and design of the pragmatic clinical trial tREatment with Beta-blockers after myOcardial infarction withOut reduced ejection fracTion (REBOOT). Eur Heart J Cardiovasc Pharmacother 2022;8(3):291–301. [DOI] [PubMed] [Google Scholar]
- 28.Maqsood MH, Alam M, Atar D, et al. Efficacy of long-term oral beta-blocker therapy in patients who underwent percutaneous coronary intervention for ST-segment elevation myocardial infarction with preserved left ventricular ejection fraction: a systematic review and meta-analysis. J Cardiovasc Pharmacol 2021;77(1):87–93. [DOI] [PubMed] [Google Scholar]
- 29.Park JJ, Kim SH, Kang SH, et al. Effect of beta-blockers beyond 3 years after acute myocardial infarction. J Am Heart Assoc 2018;7(5). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Shavadia JS, Holmes DN, Thomas L, et al. Comparative effectiveness of beta-blocker use beyond 3 years after myocardial infarction and long-term outcomes among elderly patients. Circ Cardiovasc Qual Outcomes 2019;12(7):e005103. [DOI] [PubMed] [Google Scholar]
- 31.Adelborg K, Sundbøll J, Munch T, et al. Positive predictive value of cardiac examination, procedure and surgery codes in the Danish National Patient Registry: a population-based validation study. BMJ Open 2016;6(12):e012817. [DOI] [PMC free article] [PubMed] [Google Scholar]