been reviewed in detail. 1 With recent changes in performing PCI, especially with the introduction of intracoronary stents, older data may not adequately reflect contemporary levels of risk. The present study examines the risk of in-hospital mortality following PCI in octogenarians in the stent era using a multiinstitutional database. ••• From January 1998 to June 1999, 2,968 patients 80 years of age underwent PCI in 17 of the National Cardiovascular Network of hospitals; all of these patients were included in this study. The National Cardiovascular Network is an alliance of hospitals that performs a minimum of 500 PCI procedures and 500 coronary artery bypass procedures annually. National Cardiovascular Network physicians meet or exceed American College of Cardiology volume requirements. All PCI procedures were performed using the standard techniques employed during the period of time in which the procedure took place. PCI procedures included balloon dilations, stent implantation, and the use of other new devices. Patients were considered to have diabetes based on a physician’ s diagnosis (nonstressed blood sugar 140 mg/dl on 2 occasions, although this could not reasonably be verified), or if they had received oral hypoglycemic agents or insulin, or were on diet therapy. Patients were considered to have narrowing of the coronary arteries if they had 70% diameter luminal narrowing in the left anterior descending, left circumflex, or right coronary arteries, or of a major branch. Patients had left main disease if there was 70% diameter luminal narrowing in the left main coronary artery. Patients experienced angiographic success if there was a decrease in postprocedure diameter stenosis to 50% and a decrease in percent stenosis by 20%. They were diagnosed with myocardial infarction if new Q waves were seen after the procedure. The variables defined by patient history were hypertension, severity of angina, and prior myocardial infarction. Angina was defined by the Canadian Cardiovascular Society Classification. Heart failure was defined by the New York Heart Association criteria. Demographic, clinical, angiographic, and procedural data, including complications, were recorded at the clinical sites and entered into a local computerized database. Data were sent to the data coordinating center at Emory semi-annually. All fields were defined in a data dictionary. Data are expressed as proportions or as mean SD. Categorical data were compared by chi-square and continuous by analysis of variance. Missing covariate data were imputed using an S-Plus (Mathsoft, Seattle, Washington) procedure that performs multiple imputation using relations among all variables to predict the missing observations. Multivariate correlates of discrete end points were analyzed with logistic regression. Potential nonlinear effects of each of the continuous predictor variables were checked using restricted cubic splines. Statistical modeling and testing were performed using S-Plus. The ability of the models to discriminate among patients with respect to their outcomes was measured using the c-index, which is equivalent to the area under the receiver-operating characteristic curve. Bootstrap validation and calibration of a singly imputed dataset were carried out using the “ validate”and “ calibrate”functions in the “ Design”library of S-Plus statistical functions. There were 2,968 patients with 105 deaths (3.54%). Demographic and clinical characteristics of the 2,485 patients with (13%) and without (87%) an acute myocardial infarction are listed in Table 1. Mean age was 84 years, most of the patients were women, and diabetes was common. Most of the patients had systemic hypertension, a minority had heart failure, and approximately half had unstable angina. Previous coronary artery bypass graft surgeries and PCIs were more common in patients without acute myocardial infarction. Comorbidities of cerebrovascular disease, peripheral vascular disease, renal insuf ficiency, and chronic obstructive pulmonary disease were each noted in a minority of patients. Angiographic characteristics and outcomes according to presentation with and without an acute myocardial infarction are listed in Table 2. Multivessel disease was noted in most of the patients. The mean ejection fraction was 50% and left ventricular end-diastolic pressure was 19 7 mm Hg. Most of the cases underwent elective procedures and stent usage was common. The need for coronary artery bypass grafts, and the occurrence of Q-wave myocardial infarction, and stroke occurred in a small minority of patients. The unadjusted relative risk of death for patients with an
Read full abstract