Guidelines mandate urgent revascularization in patients presenting with ST-elevation myocardial infarction (STEMI) irrespective of age. Whether this strategy is optimal in patients aged ≥85 years remains uncertain. We aimed to assess the clinical characteristics and outcomes of patients aged ≥85 years with STEMI stratified by their management strategy. We analyzed baseline clinical characteristics of 101 consecutive patients aged ≥85 years who presented with STEMI to a tertiary Australian hospital. Patients were stratified based on whether they underwent invasive management with urgent coronary angiography ± percutaneous coronary intervention or conservative management. Our primary outcome was long-term mortality. Independent predictors of conservative management and long-term mortality were assessed by multivariate logistic regression and Cox proportional hazard modeling respectively. Of the 101 patients included, 45 underwent invasive management. Independent predictors of having conservative management were older age, anterior STEMI, and cognitive impairment (all p <0.01). Patients managed invasively had lower in-hospital (13.3% vs 32.1%, p = 0.03), 30-day (13.3% vs 37.5%, p <0.01), 12-month (22.2% vs 57.1%, p <0.01), and long-term (40.0% vs 75.0%, p <0.01) mortality. Invasive management was an independent predictor of lower long-term mortality (hazard ratio 0.29, 95% CI 0.11 to 0.76, p <0.01). In conclusion, patients aged ≥85 years with STEMI who were older, had cognitive impairment or presented with anterior ST-elevation were more likely to be managed conservatively. Those who underwent invasive management had reasonable short- and long-term outcomes.
In this single-center retrospective analysis, we aimed to assess the clinical characteristics and outcomes of patients aged ≥85 years presenting with ST-elevation myocardial infarction (STEMI) stratified by invasive or conservative management.
Methods
The cohort studied included consecutive patients aged ≥85 years who presented with STEMI to Austin Health from November 2011 to July 2015. Austin Health is a large metropolitan tertiary hospital affiliated with the University of Melbourne; it services a population of approximately 1.25 million people.
Scrutiny of coding data identified patients aged ≥85 years with a diagnosis of STEMI (International Classification of Disease [ICD] 10 code I21.3). A manual chart review was performed to confirm the diagnosis and to obtain baseline clinical characteristics including functional state, management strategies, and in-hospital outcomes. STEMI was defined as symptoms suggestive of ischemia and ST-segment elevation ≥0.1 mV in ≥2 contiguous leads or new left bundle branch block on electrocardiogram. Patients who did not meet the criteria were excluded from the study, even if they had an ICD 10 code I21.3.
Patients who underwent angiography with a view to percutaneous coronary intervention (PCI) were included in the invasive group, whereas the remaining patients were included in the conservative group.
In-hospital, 30-day, 12-month, and long-term mortality and other outcomes were ascertained from review of clinical follow-up data at Austin Health. The last follow-up day was August 30, 2015. If a patient’s status was unknown at follow-up, the nominated general practitioner was contacted for additional information.
In-hospital bleeding complications as well as recurrent myocardial infarction, cerebrovascular accident diagnosed by computer tomography (CT scan) and cardiac rhythm disturbances at 12 months were also recorded. Recurrent myocardial infarction was defined as new symptoms of ischemia and/or associated electrocardiographic changes and/or biomarker elevation >20% above previous biomarker level, in the context of stable or falling levels. Cardiac rhythm disturbances included new-onset atrial fibrillation and sustained ventricular arrhythmias. In-hospital bleeding was defined as any bleeding requiring a transfusion and/or a significant hemoglobin decrease (>3 g/dl) and/or vascular surgery intervention.
Continuous variables are expressed as mean ± SD or SEM, and categorical variables as counts and percentages. Continuous variables were compared using the Student t test. Categorical variables were compared using the Fisher’s exact or chi-square tests. Cumulative incidence of mortality was estimated by the Kaplan–Meier method, and the log-rank test was used to evaluate differences between groups.
A multivariate logistic regression analysis was undertaken to determine predictors of receiving conservative management. Patient characteristics, functional status, and selected clinical variables at presentation were included in the model. As individual components of the combined Charlson age-comorbidity index were included in the multivariate analysis, the index itself was not included. Those with a p value <0.10 on univariate logistic regression analyses were entered into a stepwise backward selection multivariate logistic regression model.
A Cox proportional hazard model was used to estimate the adjusted hazard ratio (HR) and 95% CI of invasive strategy for long-term mortality using 24 baseline patient and procedural characteristics.
Statistical analysis was performed using IBM SPSS Statistics version 21.0 (IBM Corp., Armonk, New York).
Ethics approval for this study was obtained from the Austin Health Human Research Ethics Committee.
Results
A total of 115 patients aged ≥85 years were assigned the ICD 10 code I21.3 for STEMI. Of these patients, 2 (1.7%) were excluded as they were transferred from the emergency department to another institution. A further 12 patients (10.6%) were excluded as their electrocardiogram and/or clinical history did not meet criteria for an STEMI. Of the remaining 101 patients included in our analysis, 56 (55%) were managed conservatively and 45 (45%) received invasive management. Follow-up data were available for all patients.
Among patients managed invasively, 3 had mild nonobstructive coronary atherosclerosis, one had takotsubo cardiomyopathy, and another had a thrombotic left main lesion treated with anticoagulants alone. Thirty-seven of the 40 remaining patients underwent successful PCI with bare-metal stents inserted in 26 patients, drug-eluting stents inserted in 9 patients, and plain old balloon angioplasty used in 2 patients. Unsuccessful PCI occurred in 3 patients: one died in the cardiac catheterization laboratory with refractory ventricular fibrillation, whereas 2 patients had culprit lesions that could not be crossed.
Baseline medical characteristics are presented in Table 1 . Patients managed invasively were younger, had less cognitive impairment and a lower Charlston age-comorbidity index, and were more likely to be independent with mobility and activities of daily living.
| Conservative Strategy (N=56) | Invasive Strategy (N=45) | P- value | |
|---|---|---|---|
| Age (years), mean ± SD | 90.0±3.4 | 87.7±3.4 | <0.01 |
| Men | 23 (41%) | 14 (31%) | 0.32 |
| Cardiology Admission | 17 (30%) | 43 (96%) | <0.01 |
| Diabetes Mellitus | 14 (25%) | 10 (22%) | 0.74 |
| Hypertension | 42 (75%) | 34 (76%) | 0.95 |
| Hypercholesterolemia | 21 (38%) | 16 (36%) | 0.84 |
| Previous Myocardial Infarction | 15 (27%) | 11 (24%) | 0.79 |
| Smoking History | 16 (29%) | 11 (24%) | 0.64 |
| Congestive Cardiac Failure | 9 (16%) | 9 (20%) | 0.61 |
| Peripheral Vascular Disease | 14 (25%) | 10 (22%) | 0.74 |
| Stroke | 15 (27%) | 6 (13%) | 0.09 |
| Obstructive Sleep apnoea | 6 (11%) | 1 (2%) | 0.09 |
| Rheumatoid Arthritis | 4 (7%) | 3 (7%) | 0.93 |
| Previous percutaneous coronary intervention | 4 (7%) | 8 (18%) | 0.10 |
| Previous Bypass Surgery | 1 (2%) | 0 (0%) | 0.37 |
| Cancer | 8 (14%) | 5 (11%) | 0.64 |
| Cognitive Impairment | 24 (43%) | 6 (13%) | <0.01 |
| Charlson Age-Comorbidity Index (mean ± SD) | 7.5±2.5 | 6.1±1.9 | <0.01 |
| Independent with Activities of Daily Living | 15 (27%) | 26 (58%) | <0.01 |
| Independent Living | 20 (36%) | 26 (58%) | 0.03 |
| Low Level Care | 11 (20%) | 1 (2%) | <0.01 |
| High Level Care | 5 (9%) | 2 (4%) | 0.38 |
| Mobility | |||
| 15 (27%) | 27 (60%) | <0.01 |
| 9 (16%) | 5 (11%) | 0.47 |
| 30 (54%) | 13 (29%) | 0.01 |
| 2 (4%) | 0 (0%) | 0.20 |
At presentation, conservatively managed patients exhibited higher risk features such as Killip class ≥2, anterior ST-elevation, and poorer kidney function ( Table 2 ).
| Conservative Strategy (N=56) | Invasive Strategy (N=45) | P- value | |
|---|---|---|---|
| Rhythm | |||
| 43 (77%) | 39 (87%) | 0.21 |
| 12 (21%) | 6 (13%) | 0.29 |
| 1 (2%) | 0 (0%) | 0.38 |
| Killip Class ≥2 | 25 (45%) | 11 (24%) | 0.04 |
| Cardiogenic Shock | 11 (20%) | 6 (13%) | 0.40 |
| Intra-aortic Balloon Pump | 0 (0%) | 1 (2%) | 0.26 |
| Out-of-Hospital Cardiac Arrest | 2 (4%) | 2 (4%) | 0.82 |
| Inotropes | 3 (5%) | 5 (11%) | 0.29 |
| ECG changes: | |||
| 41 (75%) | 19 (42%) | <0.01 |
| 2 (4%) | 4 (9%) | 0.26 |
| 11 (20%) | 20 (44%) | <0.01 |
| 1 (2%) | 0 (0%) | 0.37 |
| 0 (0%) | 2 (4%) | 0.11 |
| Heart rate (beats/minute) | 86±23 | 78±17 | 0.05 |
| Systolic Blood Pressure (mmHg) | 120±25 | 130±26 | 0.67 |
| Diastolic Blood Pressure (mmHg) | 67±15 | 69±16 | 0.47 |
| Urea (mg/dL) | 13±9 | 9±4 | 0.05 |
| Creatinine (mmol/L) | 127±72 | 100±34 | 0.03 |
| Haemoglobin (mmol/L) | 119±19 | 123±16 | 0.24 |
None of the patients received thrombolytic therapy. Guideline-directed medical therapy was significantly better in patients managed invasively ( Table 3 ). Patients who were managed invasively also had significantly shorter length of stay with a mean (±SEM) of 6 ± 1 days (vs 12 ± 2 days; p <0.01).
| Conservative Strategy (N=56) | Invasive Strategy (N=45) | P-value | |
|---|---|---|---|
| Aspirin | 42 (75%) | 45 (100%) | <0.01 |
| Dual antiplatelet therapy | 26 (46%) | 41 (91%) | <0.01 |
| Anticoagulation | 39 (70%) | 43 (96%) | <0.01 |
| Beta-blocker | 36 (64%) | 43 (96%) | <0.01 |
| Statin | 36 (66%) | 43 (96%) | <0.01 |
| ACE inhibitor | 31 (55%) | 43 (96%) | <0.01 |
| Cardiology Review | 43 (77%) | 45 (100%) | <0.01 |
The conservatively managed cohort had significantly higher in-hospital, 30-day, 12-month, and long-term mortality, as presented in Table 4 and Figure 1 . Cox proportional hazard modeling found invasive management was an independent predictor of lower long-term mortality with a HR of 0.31 (95% CI 0.18 to 0.55; p ≤0.01). Cardiogenic shock was the only other independent predictors of long-term mortality (HR 3.60, 95% CI 1.72 to 50.86; p ≤0.01; Figure 2 ).
