ST-segment elevation myocardial infarction (STEMI) is common in older adults and has high age-related mortality. We describe contemporary STEMI care using the National Cardiovascular Data Registry Acute Coronary Treatment and Intervention Network Registry–Get With The Guidelines (ACTION-GWTG) database. Patients with STEMI (n = 30,188) from 285 ACTION-GWTG sites from January 1, 2007 to June 30, 2008 were grouped by age (<75, 75 to 84, and ≥85 years) to compare baseline characteristics, reperfusion, and in-hospital outcomes. In this population, 79.7% (24,070) were <75 years old, 14.2% (4,273) were 75 to 84 years old, and 6.1% (1,845) were ≥85 years old (the oldest old). Compared to younger patients, the oldest-old patients (median age 88 years, interquartile range 86 to 91) were more often women, had more hypertension, and end-organ co-morbidity (heart failure and stroke, p <0.0001 for all). More than 42% of the oldest old were also cited as having contraindications to reperfusion, but with absolute or relative contraindications noted in only 10%, and patient preference was the most common reason indicated (45%). Even in reperfusion-eligible patients, the oldest old were less likely to receive it. Although patients who received reperfusion had better outcomes than those who did not, this was significant only for younger patients (<75 years old, odds ratio 0.58, confidence interval 0.40 to 0.84). In conclusion, >42% of the oldest old have reported contraindications to reperfusion, with neither mortality benefit nor harm in those who receive it. Disparities in process of care and co-morbidity may explain these observational findings. Whether efforts to optimize patient selection and initiate reperfusion therapy can improve outcomes in the oldest old with STEMI is unknown.
Rapid reperfusion remains the cornerstone of evidence-based ST-segment elevation myocardial infarction (STEMI) care. Although absolute benefits associated with reperfusion should ideally increase in high-risk patients, other age-associated factors may limit their realization. In older patients, ambiguous symptoms, delays in presentation or access, co-morbidity, frailty, and polypharmacy are among the factors that can complicate STEMI care. Therefore, by describing treatment and outcomes of patients with STEMI relative to age in community practice, successes, and residual gaps can be recognized. Using the Acute Coronary Treatment and Intervention Network Registry–Get With The Guidelines (ACTION-GWTG) database, we describe reperfusion therapy for 3 age groups (<75, 75 to 84, and ≥85 years [the oldest old]) presenting with STEMI to clarify treatment patterns and age-associated outcomes.
Methods
The ACTION-GWTG is part of the National Cardiovascular Data Registry and was created to evaluate and improve the quality of MI care. Participating ACTION-GWTG hospitals enroll consecutive patients with a primary diagnosis of STEMI or non-STEMI. At most hospitals, patients are retrospectively identified and are deemed ineligible if admitted for conditions unrelated to MI. In all instances, patient care decisions are the responsibility of the treating physicians. Criteria for patients with STEMI include (1) ischemic symptoms at rest, lasting ≥10 minutes, and occurring up to 72 hours before admission, and (2) electrocardiographic changes associated with STEMI (new left bundle branch block or persistent ST-segment elevation ≥1 mm in ≥2 contiguous electrocardiographic leads).
The ACTION-GWTG is a quality improvement initiative, so individual informed consent is not required. Nevertheless, internal review board approval is obtained by participating hospitals in accordance with regulatory policies. Of the 286 participating hospitals enrolling patients with MI for ACTION-GWTG from January 2007 to June 2008, 30,193 patients met STEMI criteria. Five patients were excluded because their age was missing, leaving a final population of 30,188.
Definitions of baseline variables and outcomes are available on the National Cardiovascular Data Registry Web site. Major bleeding is defined as intracranial hemorrhage, retroperitoneal hemorrhage, hematocrit decrease ≥12% (baseline to lowest), any red blood cell transfusion when baseline hematocrit is ≥28%, or any red blood cell transfusion when baseline hematocrit is <28% and a witnessed bleed is also documented. Contraindications for reperfusion are checked as present or absent. Specific reasons for omitting reperfusion are then selected from a list of choices. The list includes absolute contraindications and relative contraindications based on American College of Cardiology/American Heart Association guidelines and other reasons for no reperfusion. Only 1 contraindication for each patient is recorded; if >1 contraindication to reperfusion is present for a patient, then the most substantial contraindication is selected by the chart abstractor.
Absolute or relative contraindications to reperfusion include active or recent bleeding, intracranial neoplasm, arteriovenous malformation or aneurysm, aortic dissection, intravenous contrast allergy, traumatic cardiopulmonary resuscitation, known bleeding diathesis, severe uncontrolled hypertension, recent surgery or trauma, cerebrovascular history, current use of oral anticoagulants, active peptic ulcer disease, and significant co-morbidity. Electrocardiographic reasons for no reperfusion include absence of ST-segment elevation, left bundle branch block, or spontaneous resolution of ST-segment elevation. Symptom and timing reasons for no reperfusion include resolution of chest pain, ambiguous diagnosis of acute coronary syndrome, or onset of acute coronary syndrome >12 hours before assessment. Coronary anatomic reasons for no reperfusion include spontaneous reperfusion (documented by heart catheterization) or finding that coronary anatomy was not suitable to a primary percutaneous coronary intervention (PCI). Patient reasons for no reperfusion include quality-of-life choices, patient or family refusal, or do-not-resuscitate code status.
Characteristics of the clinical presentation, medical histories, medications, reperfusion type and timing, and in-hospital outcomes are presented for patient groups stratified by <75, 75 to 84, and ≥85 years of age. Continuous variables are described as median (25th, 75th percentiles) and categorical variables are described as percentages. In-hospital mortality and complications are shown overall and by specific age groups. A graphic display is used to show the relation between in-hospital mortality and continuous age stratified by those patients who received reperfusion and those who did not. To examine the association among age, reperfusion, and in-hospital mortality, logistic generalized estimating equations method with exchangeable working correlation matrix was used to account for within-hospital clustering because patients at the same hospital are more likely to have similar responses compared to patients at other hospitals (e.g., within-center correlation for responses). This method produces estimates similar to those from ordinary logistic regression, yet variances are adjusted for the correlation of outcomes within a hospital. Adjustments included gender, race, body mass index, heart rate, systolic blood pressure, and signs of heart failure at admission and medical history of hypertension, diabetes mellitus, current/recent smoker, hypercholesterolemia, currently on dialysis, previous peripheral arterial disease, previous MI, previous stroke, previous heart failure, previous PCI, and previous coronary artery bypass graft surgery.
A p value <0.05 was considered statistically significant for all tests, and all tests of statistical significance were 2-tailed. All analyses were performed using SAS 9.2 (SAS Institute, Cary, North Carolina).
Results
Of the 30,188 patients with STEMI selected, 79.7% (24,070) were <75 years old, 14.2% (4,273) were 75 to 84 years old, and 6.1% (1,845) were ≥85 years old (e.g., the oldest old). The oldest-old patients were mostly women, were not recent smokers, and had less diabetes mellitus and hyperlipidemia. This group had higher prevalence of end-organ co-morbidity (previous heart failure and stroke; Table 1 ). They also had lower creatinine clearance and higher B-type natriuretic peptide levels than the younger age strata. With increasing age, patients with STEMI also had more signs of heart failure at presentation and lower systolic blood pressure at admission. An age-associated increase in time from symptom onset to presentation was also noted.
| Baseline Characteristics | Total Population (n = 30,188) | Age (years) | p Value for Trend | ||
|---|---|---|---|---|---|
| <75 (n = 24,070) | 75–84 (n = 4,273) | ≥85 (n = 1,845) | |||
| Age (years) ⁎ | 60 (52–72) | 57 (50–64) | 79 (77–82) | 88 (86–91) | <0.0001 |
| Women | 30% | 25% | 48% | 61% | <0.0001 |
| Caucasian | 85% | 84% | 89% | 91% | <0.0001 |
| Body mass index (kg/m 2 ) ⁎ | 28 (25–32) | 29 (26–33) | 26 (24–30) | 24 (22–27) | <0.0001 |
| Current/recent smoker | 42% | 50% | 13% | 5% | <0.0001 |
| Previous myocardial infarction | 19% | 18% | 21% | 22% | <0.0001 |
| Previous revascularization | 23% | 23% | 26% | 20% | 0.0047 |
| Previous hypertension † | 60% | 57% | 73% | 76% | <0.0001 |
| Previous hyperlipidemia ‡ | 47% | 47% | 51% | 40% | 0.4820 |
| Previous heart failure | 5% | 3% | 10% | 17% | <0.0001 |
| Previous peripheral artery disease | 6% | 5% | 11% | 10% | <0.0001 |
| Previous stroke | 5% | 4% | 10% | 13% | <0.0001 |
| Diabetes mellitus | 22% | 22% | 27% | 20% | 0.0031 |
| Creatinine clearance (ml/min) ⁎ § | 84 (60–111) | 93 (73–118) | 51 (39–65) | 34 (25–45) | <0.0001 |
| Brain natriuretic peptide (pg/ml) ⁎ | 181 (46–555) | 115 (31–365) | 401 (151–923) | 602 (243–1,390) | <0.0001 |
| Initial systolic blood pressure (mm Hg) ⁎ | 138 (117–157) | 138 (118–158) | 136 (114–157) | 133 (110–155) | <0.0001 |
| Initial heart rate (beats/min) ⁎ | 78 (65–93) | 78 (65–92) | 78 (64–94) | 81 (66–98) | 0.843 |
| Initial hematocrit (%) ⁎ | 43 (39–46) | 43 (40–46) | 40 (36–43) | 38 (35–42) | <0.0001 |
| Left bundle branch block on presentation | 4% | 3% | 9% | 13% | <0.0001 |
| Door-to-electrocardiogram time (min) ⁎ | 7 (3–14) | 6 (3–13) | 8 (4–16) | 10 (4–21) | <0.0001 |
| Symptom onset to presentation time (hours) ⁎ ∥ | 1.70 (0.98–3.60) | 1.63 (0.97–3.42) | 1.93 (1.02–4.05) | 2.20 (1.05–4.97) | <0.0001 |
⁎ Median (25th–75th percentiles).
† Systolic blood pressure >140 mm Hg or diastolic blood pressure >90 mm Hg on repeated measurements, or hypertension treated long term with antihypertensive medications.
‡ Known total cholesterol level >200 mg/dl (5.2 mmol/L) or long-term treatment with lipid-lowering agent.
§ Creatinine clearance calculated using the Cockcroft-Gault formula in patients not on dialysis: ([140 – age] × actual body weight) − (baseline serum creatinine × 72) multiplied by a correction factor of 0.85 in women and 1.0 in men.
In Table 2 , acute and discharge anticoagulant, antiplatelet, and adjunctive medications are presented across age groups for those without contraindications. Older age was associated with decreases in the use of certain therapies, yet usage rates remained generally high, including heparin, aspirin, and β blockers in hospital, and aspirin, β blockers, clopidogrel, and statins at discharge.
| Medications | Total Population (n = 30,188) | Age (years) | p Value for Trend | ||
|---|---|---|---|---|---|
| <75 (n = 24,070) | 75–84 (n = 4,273) | ≥85 (n = 1,845) | |||
| Acute (≤24 hours) (%) ⁎ | |||||
| Heparin † | 89 | 89 | 89 | 87 | 0.012 |
| Glycoprotein IIb/IIa | 73 | 76 | 65 | 52 | <0.0001 |
| Aspirin | 98 | 98 | 98 | 96 | 0.002 |
| Clopidogrel | 86 | 88 | 80 | 72 | <0.0001 |
| β Blocker | 96 | 96 | 94 | 93 | <0.0001 |
| Angiotensin-converting enzyme inhibitor or angiotensin receptor blocker | 56 | 57 | 52 | 50 | <0.0001 |
| Statin | 67 | 69 | 62 | 53 | <0.0001 |
| Discharge (%) ⁎ | |||||
| Aspirin | 99 | 99 | 98 | 97 | 0.004 |
| Clopidogrel | 91 | 92 | 87 | 80 | <0.0001 |
| Warfarin | 7 | 7 | 12 | 8 | <0.0001 |
| β Blocker | 97 | 97 | 97 | 96 | NS |
| Angiotensin-converting enzyme inhibitor or angiotensin receptor blocker | 79 | 79 | 79 | 75 | NS |
| Aldosterone blocker | 5 | 4 | 6 | 6 | <0.0001 |
| Statin | 92 | 93 | 89 | 83 | <0.0001 |
⁎ In those without contraindications; for discharge medications, we also excluded patients transferred out and patients who died during hospitalization.
† Recorded as any time during hospitalization rather than within 24 hours and heparin is low-molecular-weight or unfractionated heparin.
Use of diagnostic catheterization and PCI decreased more prominently with age ( Table 3 ). More than 90% of patients <75 years of age underwent coronary angiography compared to 79% of the oldest old (p <0.0001). Likewise, 90% of patients <75 years of age had an assessment of left ventricular ejection fraction compared to 81% of the oldest old (p <0.0001).
| Procedures | Total Population (n = 30,188) | Age (years) | p Value for Trend | ||
|---|---|---|---|---|---|
| <75 (n = 24,070) | 75–84 (n = 4,273) | ≥85 (n = 1,845) | |||
| Cardiac catheterization ⁎ | 92% | 93% | 90% | 79% | <0.0001 |
| Percutaneous coronary intervention † | 85% | 88% | 78% | 69% | <0.0001 |
| Coronary artery bypass grafting † | 8% | 8% | 9% | 3% | 0.298 |
| Assessment of left ventricular ejection fraction | 89% | 90% | 87% | 81% | <0.0001 |
⁎ In those without catheter contraindications.
Table 4 presents those patients for whom contraindications to reperfusion were reported. Overall, 17% were reported to have contraindications, but a significant age-associated progression in contraindications was also evident (13% of those <75 years old, 25% of those 75 to 84 years old, and 42% of those ≥85 years old). However, increased absolute and relative contraindications explain only 9% of the age-related discrepancy in contraindications; age-specific differences in patient preferences were a greater determinant of treatment differences. Of patients <75 years old, reperfusion was most likely to be contraindicated due to symptoms, timing, electrocardiographic criteria, and coronary anatomy, whereas in patients ≥85 years old, reperfusion was more likely to be contraindicated based on patient preferences (4% in patients <75 years old, 15% in those 75 to 84 years old, and 45% in those ≥85 years old). Furthermore, as presented in Table 5 , even in eligible older patients with STEMI without contraindications, reperfusion was less likely to be performed (95% of those <75 years old, 89% of those 75 to 84 years old, and 84% of those ≥85 years old, p <0.0001).
| Contraindications to reperfusion | Total Population (n = 30,188) | Age (years) | p Value for Trend | ||
|---|---|---|---|---|---|
| <75 (n = 24,070) | 75–84 (n = 4,273) | ≥85 (n = 1,845) | |||
| Total contraindications | 17% (5,078) | 13% (3,239) | 25% (1,067) | 42% (772) | <0.0001 |
| Reason for contraindication | |||||
| Absolute or relative † | 9% (465) | 8% (269) | 11% (120) | 10% (76) | |
| Symptom, timing, or electrocardiographic factors | 31% (1,582) | 34% (1,089) | 30% (319) | 23% (174) | |
| Coronary anatomy factors | 30% (1,505) | 35% (1,123) | 27% (293) | 12% (89) | |
| Patient preferences | 13% (639) | 4% (129) | 15% (162) | 45% (348) | |
| Other | 17% (865) | 19% (620) | 16% (170) | 10% (75) | |
⁎ Missing contraindications excluded from the denominator (n = 22).
† Absolute or relative contraindications as described in guidelines.
| Reperfusion | Total Population (n = 30,188) | Age (years) | p Value for Trend | ||
|---|---|---|---|---|---|
| <75 (n = 24,070) | 75–84 (n = 4,273) | ≥85 (n = 1,845) | |||
| Reperfusion eligible | 83% (25,110) | 87% (20,831) | 75% (3,206) | 58% (1,073) | <0.0001 |
| Reperfusion given ⁎ | 93% (23,449) | 95% (19,686) | 89% (2,863) | 84% (900) | <0.0001 |
| Reperfusion type † | 0.57 | ||||
| Percutaneous coronary intervention | 85% (19,486) | 85% (16,264) | 86% (2,415) | 91% (807) | |
| Lytic | 8% (1,827) | 8% (1,531) | 9% (243) | 6% (53) | |
| Percutaneous coronary intervention and lytic | 7% (1,615) | 8% (1,447) | 5% (141) | 3% (27) | |
| Reperfusion timing (min) | |||||
| Door to balloon (min) ‡ § | 73 (56–93) | 72 (55–91) | 77 (59–98) | 80 (63–103) | <0.0001 |
| Door to lytic (min) ‡ § | 28 (17–48) | 26 (17–45) | 38 (19–66) | 42 (26–60) | 0.007 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
