Patients scheduled to undergo isolated CABG from December 2003 through August 2005 were screened for participation in this prospective, randomized trial. Inclusion criteria were: age >50 years and treatment with β-blockers (atenolol, propranolol, or metoprolol) for a target heart rate of about 60–70 beats/min, at least 1 week before surgery. Those included had to give informed, written consent. One hundred patients entered the study: 50 patients were randomly assigned to the ascorbic acid group (those who were to receive combination therapy with β-blockers and ascorbic acid), and the other 50 were assigned to the control group (those who were to receive β-blocker therapy alone). The mean age of the population was 60.19±7.14 years, and 67% were men. Baseline and perioperative characteristics were similar in the ascorbic acid and control groups ().
TABLE I. Baseline and Perioperative Characteristics of Patients
Excluded from the study were patients who had a history of atrial fibrillation, medication with class I and III antiarrhythmic agents or digoxin, a permanent or temporary pacemaker, any degree of atrioventricular block or bradycardia with a heart rate of <50 beats/min, end-stage renal disease, severe pulmonary disease (pneumonia or chronic obstructive pulmonary disease), or severe hepatic disease (cirrhosis or fulminant hepatitis).
Echocardiography (Hewlett-Packard, Image Point HX; now a part of Philips Medical; Bothell, Wash) was performed before surgery by a single investigator in a blinded fashion. The 2-dimensional echocardiogram was used to obtain measurement of diastolic left atrial diameter at the level of the aortic root by means of electronic calipers. Left ventricular diameters at end-systole and end-diastole were measured via M-mode echocardiography at the tip of the mitral valve. Left ventricular ejection fractions were estimated by study of the apical 2- and 4-chamber views.
Hypertension was defined as a systolic blood pressure ≥140 mmHg or a diastolic blood pressure ≥90 mmHg, or as a history of treatment for hypertension. Hyper-cholesterolemia was defined as low-density lipoprotein levels >100 mg/dL, and all patients with hypercholester-olemia were prescribed lovastatin (≥20 mg/day) at least 1 week before surgery. Cigarette smoking was defined as the consumption of at least 3 cigarettes per day during the preceding 3 months.
Patients in the ascorbic acid group were given 2 g of effervescent ascorbic acid tablets (Hakim Pharmaceutical Co.; Tehran, Iran) on the night before surgery, followed by 1-g doses twice daily for 5 days after surgery. Before surgery, patients in both the control and ascorbic acid groups were treated with β-blockers (atenolol, propranolol, or metoprolol), and after surgery, the β-blockers were continued in both groups. Other medications were prescribed according to clinical indications in both groups.
Patients underwent CABG on standard cardiopulmonary bypass with myocardial protection provided by the perfusion of blood cardioplegic solution. Patients were initially monitored in the cardiothoracic intensive care unit (ICU) before their transfer to a nursing ward. Holter monitoring (Novacor, Vista Access; Rueil-Malmaison, France) was performed for 4 days after the transfer of patients to nursing wards. All of the Holter recordings were examined by a single investigator who had been blinded to patients' group assignments. In addition to the Holter recordings, standard 12-lead electrocardiography was recorded on the day of surgery and on postoperative days 1 and 5.
The primary endpoints in our study were the detection of an episode of atrial fibrillation lasting >10 minutes or the requirement for urgent intervention due to atrial fibrillation (profound symptoms of atrial fibrillation or of hemodynamically unstable situations due to atrial fibrillation).
Our hospital's ethics committee approved our study.
Clinical data were evaluated using SPSS version 11.5 statistical software (SPSS Inc.; Chicago, Ill). Continuous variables are expressed as mean±standard deviation.
The univariate association between qualitative vari-ables was evaluated by the χ2 test or Fisher's exact test, if needed. Student's t test was used for 2-group comparisons of continuous parametric data; the Mann-Whitney test was used for nonparametric 2-group comparison.
Binary logistic regression analysis with the backward Wald method was performed to determine the independent predictors of atrial fibrillation. Independent variables for the multivariate model were selected from those identified in a review of the literature. These variables were as follows: age, sex, and left ventricular ejection fraction <0.30. Also included were variables that had an association of P <0.1 with atrial fibrillation in the univariate model; ascorbic acid medication could therefore be included in the model. P values of <0.05 were considered statistically significant.
Sample size was derived from the assumptions that the baseline incidence of post-CABG atrial fibrillation in the control population (treated with β-blockers) would be 25% to 30%9
and that ascorbic acid, as an adjunct to β-blockers, could result in a 20% decrease in atrial fibrillation.12
With an α error of 0.05, a β error of 0.2, and a power of 80%, approximately 50 patients per group were needed.