Revascularization Versus Medical Therapy in Patients Aged 80 Years and Older With Acute Myocardial Infarction

Derek Q. Phan, MD; Ara H. Rostomian, MD, MBA, MPH; Franz Schweis, MD; Joanie Chung, MPH; Bryan Lin, MS; Ray Zadegan, MD; Ming-Sum Lee, MD, PhD


J Am Geriatr Soc. 2020;68(11):2525-2533. 

In This Article


Study Population

This is a retrospective observational study of all patients aged 80 years or older who underwent cardiac catheterization at the Regional Cardiac Catheterization Laboratory at the Kaiser Permanente Los Angeles Medical Center. This is a tertiary cardiac referral center serving all medical centers within the Kaiser Permanente Southern California (KPSC) Healthcare System, of which all patients are insured under the same insurance plan. All patients were evaluated utilizing a Heart Team approach, including senior cardiothoracic surgeons, general cardiologists, and interventional cardiologists. The agreed on treatment recommendations/options are offered to the patients who ultimately make the final decision. In general, coronary anatomy complexity, comorbidities, frailty, dementia, and other pertinent patient characteristics, in addition to patients' goals of care and preferences, are taken into consideration when developing treatment recommendations/options. All patients undergoing invasive coronary angiography (ICA) between March 2009 and March 2019 were evaluated (Figure 1). The following patients were excluded: younger than 80 years, not a current Kaiser Permanente health plan enrollee at time of ICA (index date), less than 1-year enrollment before ICA, presence of significant concomitant valvular disease (defined as moderate or greater in any valve location), endocarditis, and presence of subclavian stenosis. Only patients with troponin level greater than 99th percentile of the upper limit of normal (which was defined as >0.04 ng/mL at the laboratories used for this study) were included. Patients with nonobstructive coronary artery disease (CAD) (defined as absence of ≥50% stenosis in any coronary vessel) after undergoing ICA were excluded from the analysis. A total of 1,433 patients with obstructive CAD were ultimately included. This study was approved by the Institutional Review Board of KPSC.

Figure 1.

Flowchart of patient cohort derivation. CABG, coronary artery bypass grafting; ICA, invasive coronary angiography; PCI, percutaneous coronary intervention.

Clinical Characteristics

Baseline demographics were obtained using KPSC administrative records. Medical comorbidities were obtained using International Classification of Diseases, Ninth Revision (ICD-9), and International Classification of Diseases, Tenth Revision (ICD-10), codes and confirmed with manual review of electronic health records. Manual review of ICA reports was performed to determine presence of obstructive CAD, left main involvement, and presence of chronic total occlusion (CTO) in any vessel. Presence of CTO was defined by the treating physician at the time of ICA. Manual review of electronic medical records was performed to determine patients' clinical presentation (ST-segment–elevation myocardial infarction or NSTEMI) and left ventricular ejection fraction (LVEF; as assessed by echocardiography and left ventriculogram at index hospitalization).

Revascularization Versus Medical Therapy Groups

Patients were grouped by treatment received: PCI, CABG, or medical therapy alone. They were then grouped into two categories: revascularization (PCI or CABG) and medical therapy alone. Manual chart review was performed to determine the rationale behind the choice of medical therapy versus revascularization.


The following outcomes were evaluated: all-cause mortality, nonfatal myocardial infarction (MI), and need for repeated revascularization. All-cause mortality was obtained from the California State Death Master Files and Kaiser Permanente Healthcare Plan database. Nonfatal MI was defined as (1) any hospitalization or emergency room visit with the primary encounter diagnosis of AMI as identified by ICD-9 and ICD-10 codes or (2) need for ICA for a troponin elevation greater than 99th percentile of the upper limit of normal. Need for repeated revascularization was defined as any PCI or CABG in any vessel occurring after the index date and after completion of all staged PCI procedures (determined at the discretion of the physician at the time).

Statistical Analysis

Comparison between revascularization and medical therapy groups was performed. Normal distribution of continuous variables was determined by the Shapiro-Wilk test. Analysis of variance, independent samples t-test, Wilcoxon rank-sum test, or Kruskal-Wallis test was performed where appropriate for continuous variables (expressed as mean ± standard deviation for parametric and median (interquartile range (IQR)) for nonparametric tests). The chi-square test was performed for categorical variables (expressed as number (percentage)).

Univariate and multivariate Cox proportional hazards regression analysis was performed to determine association of revascularization with outcomes. The Kaplan-Meier method was used to estimate outcome event rates, and log-rank was used to compare differences between groups. Subgroup analysis for all-cause mortality was performed and displayed in a forest plot. A two-sided P ≤ .05 was used as the cutoff for statistical significance.

To control for confounders, both groups were balanced using inverse probability of treatment weighting (IPTW). The propensity score for revascularization was estimated by a logistical regression model including 29 covariates that covered demographics, clinical characteristics, and baseline medication use. Assessment of balance between groups was determined by independent t-test and chi-square test for continuous and categorical variables, respectively. Covariates were determined to be well balanced if the P > .05. Statistical analysis was performed using R version 3.6.3.