Effect of amiodarone ± diltiazem ± beta blocker on frequency of atrial fibrillation, length of hospitalization, and hospital costs after coronary artery bypass grafting

Abstract

C linical pathways are management plans that provide a sequence of steps in the care process to achieve a certain goal with optimal efficiency. Many cardiac surgical programs have implemented clinical pathways for postoperative atrial fibrillation (AF), yet little data exist as to the efficacy of such clinical pathways. This study assesses the impact of a clinical pathway on the incidence of postoperative AF, length of stay (LOS), and cost. • • • A clinical pathway addressing prophylaxis with oral amiodarone, rate and rhythm control, cardioversion, and anticoagulation was developed and accepted by all clinical services involved in the care of cardiac surgery patients (Figure 1). Amiodarone was administered either preor postoperatively, at the discretion of the attending physician. The entry criteria consisted of patients who underwent coronary artery bypass grafting (CABG) and who had preoperative normal sinus rhythm. The exclusion criteria consisted of a contraindication to amiodarone or chronic antiarrhythmic drug therapy. Data collection and patient enrollment occurred at 1 large community hospital. This study was in accordance with the regulations of the local human studies policy of the hospital. A retrospective sample of consecutive primary patients who underwent CABG from January 1998 to April 1998 was reviewed and served as a historical control group. Following the development of the clinical pathway, a 1-month trial period was implemented to ensure familiarity with the pathway. From September 1998 to January 1999, consecutive patients who underwent CABG were prospectively treated under the clinical pathway. Demographic and clinical information were recorded on standardized data forms by research nurses. The main end points of the investigation were the incidence of postoperative AF (postoperative, at discharge, or at clinical follow-up), postoperative LOS, and hospital costs. Postoperative AF was coded as present if it persisted for 4 hours or required treatment of any kind. Rate control, if needed, was achieved primarily with intravenous diltiazem and less often with a blocker. Hospital costs were derived from charges incurred from the operation to discharge. These charges were multiplied by a factor of 0.52 to produce the estimated actual hospital costs, which were then subject to statistical analysis. The final costs are presented as a cost index with 1.0 representing the control group. These cost data do not represent “true costs,” but are reasonable estimates for comparative purposes. Costs in the pathway group were discounted by 3% to reflect changes in the cost of health care over the average of 8 months that separated the different groups. Statistical analyses compared means and SDs between historical controls and the clinical pathway groups. Medians were presented with an interquartile range (IQR) (25th to 75th percentiles). Chi-square contingency tests were performed to compare univariate associations with historical or pathway cases. Student’s t tests or Wilcoxon rank-sum tests (2-tailed) were applied as appropriate; the latter method was applied to contrast highly skewed distributions. No corrections for multiple comparisons were made. A Bonferroni-type correction for pairwise comparisons was applied as appropriate. Statistical significance was set at p 0.05. A total of 229 consecutive patients (102 in the control group and 127 in the pathway group) who underwent CABG met the eligibility requirements for the study. One patient in the control group died and was excluded from analysis. The baseline demographic and clinical information for the control and pathway groups are listed in Table 1. Although mean age was similar across the groups, there were slightly more patients aged 65 to 74 years in the control group (p 0.04). There was a higher rate of prior myocardial infarction in the pathway group, but no other significant differences between the 2 groups were found. Postoperative -blocker utilization was similar. The mode and rate of utilization of amiodarone prophylaxis is noted in Table 1. The implementation of oral amiodarone prophylaxis as part of the clinical pathway resulted in a significant reduction in the incidence of postoperative From the Division of Cardiology, Evanston Northwestern Healthcare, Evanston, Illinois; Toledo Hospital, Toledo, Ohio; Division of General Pediatrics and Cardiovascular Division, University of Michigan, Ann Arbor, Michigan; and Cardiovascular Division, Washington University, St. Louis, Missouri. Dr. Kim was supported by grants from A Company-wide Commitment to Outcomes Research and Development (ACCORD), Hoechst Marion Roussel, Kansas City, Missouri, and St. Jude Medical, Sylmar, California. Dr. Eagle was supported by a grant from Hoechst Marion Roussel, Kansas City, Kansas. Dr. Kim’s address is: Cardiac Electrophysiology, Burch 300, Evanston Hospital, 2650 Ridge Ave., Evanston, Illinois 60201. E-mail: mhkim@northwestern. edu. Manuscript received October 19, 2001; revised manuscript received and accepted January 28, 2002.