Randomized Trial Comparing Diltiazem and Metoprolol For Atrial Fibrillation Rate Control

If either an initial DT or MET strategy are shown to be more effective in obtaining rate and rhythm control, decreasing length of stay in the ED, and decreasing admissions and readmissions, there is a potential for: 1.) Major health cost savings; 2.) Improved bed utilization for our hospital, which is frequently at capacity and unable to accept transfers from outlying hospitals; and, 3.) Education of ED and non-ED health care providers of optimal AF/FL medication options.

Sample

Preliminary information obtained from CAMC's Data Warehouse from July 1, 2011 to June 30, 2012 revealed 370 patients were seen in the Emergency Department (ED) at either CAMC Memorial or General Hospitals with a diagnosis of AF/FL, without any other acute co-morbidity. The investigators will identify which Hospital the patient was enrolled.

The investigators will enroll a total of 150 patients based on the following:

Prior data indicate that the conversion to a normal sinus rhythm by 8 hours is approximately 20%. If the investigators wish to detect a 25% change between medications then the investigators would need to include a total of 150 patients (75 in each cohort) to reject the null hypothesis that the conversion rates are equal between the 2 medications with probability (power) 0.8 and Type I error \< 0.05. The investigators will use a continuity-corrected chi-squared statistic or Fisher's exact test to test our hypothesis. Power analysis performed using GPower Version 3.

Procedures

Potential study patients will be identified by an ED nurse or a member of the study team using the inclusion and exclusion criteria. ED nurses will be informed of the study by charge nurses who have been educated about the protocol by the research team. The purpose, content and logistics of the study will also be described to ED physicians participating in the study. Research assistants, when available, will be notified by cell phone or pager to see patient in the ED within 20 minutes. The attending physician or resident will be informed of the potential study patient. The attending physician or resident will give the research associate/member of the study team permission to discuss the study with the patient. If the patient is interested in participating in the study, the research associate/member of the study team will discuss the details of the study and review the consent form. It is likely the onset of treatment with DT or MET will be delayed 30 to 40 minutes due to the consenting process and study enrollment. The attending physician must agree to this delay.

At this point the attending physician will discuss the study with the patient and answer questions. The physicians or resident initially talking with the patients about being in the trial will have Collaborative Institutional Training Initiative (CITI) training. General education to the ED staff will be made concerning aims and objectives of the strategy study. If the patient is still interested and willing to participate, the informed consent and Health Insurance Portability and Accountability Act (HIPAA) privacy statement will be signed.

Following consent, patients will be randomized to either IV diltiazem (DT) or IV metoprolol (MET). Sealed envelopes containing the randomized treatment protocol will be maintained on the ED premises to be pulled consecutively after a patient has been consented to be in the study. Inside the envelope, the specific treatment plan will be revealed. Treatment with either DT or MET will be initiated using standard CAMC protocol. Deviations from the randomly assigned medication regimen would be at the attending physician's discretion pending vital signs response. If the study drug is stopped or changed to a different type of medication, the patient's participation will continue and medications will be documented with the analysis of results being the intent to treat.

The same physicians rotate at both General and Memorial Hospitals. The investigators will start at the Memorial Hospital, but after experience with enrollment the investigators may enroll patients from both Hospitals.

Data Analysis

Basic descriptive statistics including means and standard deviations for continuous variables and proportions and frequencies for categorical variables will be used to describe the patient sample and disease characteristics.

To determine if a medication regimen (MET vs DT) is noted by descriptive analysis to be associated with primary aim 1 for treating AF/FL in the ED or during hospitalization (Primary Aim-1), the investigators will describe the proportion of patients with rhythm control (defined as normal sinus rhythm) by time of ED discharge using Chi-square analysis with continuity correction. Using the same statistical tests, the investigators will describe the proportion of patients with normal sinus rhythm at time of hospital discharge, in patients admitted to CAMC. Possible confounders, including patient characteristics, co-morbid conditions, ED events, or deviation from medicine regimen on achieving NSR will be examined by comparing patients with the confounder to those without the confounder of interest using Chi-square with continuity correction analyses if data is categorical, t-test or Mann-Whitney U analysis if data is continuous. In addition, the investigators will examine the effect of medication regimen on rate control (defined as heart rate \< 100 bpm) (Primary Aim-2) at 2, 4, 6, 8 hours in the identical manner as described for rhythm control including the examination of confounding variables.

The proportion of patients receiving CAMC hospital admission versus discharged to home from the ED will be described between the two medication regimens using Chi-square analysis with continuity correction to determine if one regimen resulted in a greater percentage of home discharges directly from the ED (Secondary Aim-1). The effect of possible confounders including patient characteristics, ED events, achieving rate and/or rhythm control in the ED or deviation from medication regimen will be examined by describing patients with the confounder to those without the confounder of interest on the need for hospital admission using Chi-square with continuity correction analyses if data is categorical, t-test or Mann-Whitney U analysis, if data is continuous. Those confounders that are shown to associate with ED discharge to home will be included in a multivariate analysis (binomial logistic regression) to examine if medication regimen associates with ED discharge to home in the presence of confounders.

To determine if a medication regimen (MET vs DT) associated with decreased financial costs (Secondary Aim-2), ED costs will be examined between the 2 medication regimens in all patients using t-tests, or Mann-Whitney U analysis. In addition, the total cost of hospitalization (ED + hospital costs) of patients admitted to the CAMC hospital from the ED will be examined between the medication regimens. Cost will also be examined in relationship to length of stay by examining cost /length of stay between the two treatment groups using the same statistical tests.

The need for readmission and atrial fibrillation reoccurrence as reported by the phone questionnaire (Secondary Aim-3) will be compared between the two medication regimens using Chi-square with continuity correction analyses. The investigators will also determine which rate control medication at time of phone interview associates with freedom from atrial fibrillation reoccurrence using chi-square analysis. Lastly, the time between discharge and phone survey study completion (maximum: 6 months) will be determined and the investigators will examine the effect of this lag time (4- 6 months) on the percentage of patients requiring readmission and/or having a reoccurrence of atrial fibrillation via chi square analysis.

For all applicable analyses described above, Fisher's exact tests will be substituted for Chi square analysis when appropriate. For all comparisons, a p level of \< 0.05 will be used to determine statistical significance. The statistical package Statistical Analysis Software (SAS) (Version 9.3) will be used to analyze data.

Confidentiality

Minimal patient identification information will be collected. The patient account number and telephone number must be collected. Subjects will also be assigned a study number for data analysis; account numbers will not be included in the data analysis. Personal health information will be available only to the study investigators. All collected data will be stored in a locked filing cabinet in the Emergency Department until it can be delivered to CAMC Outcomes Research. Electronic data will be stored on a password protected computer. At the conclusion of the study, all data collection sheets will be shredded and electronic data destroyed.