Atrial Fibrillation in a Multiethnic Inpatient Population of a Large Public Hospital: METHODS Study
Population
Records of patients with electrocardiogram (ECG) diagnosis of AF were examined from LAC+USC Medical Center’s Marquette ECG database for the 1999 calendar year. This ECG database included records of both inpatients and outpatients. The automatically interpreted ECGs generated for that calendar year were previously reviewed, corrected when necessary, and confirmed by the hospital’s cardiologists. Of 80,021 over-read ECGs in the Marquette database, 3,955 had a diagnosis of AF, comprising 904 distinct subjects. Since the number of ECGs per subject was variable, the first ECG manifesting AF during the hospitalization was used in the study. The hospital’s discharge record database was queried for patients with a diagnosis of AF during the calendar year 1999. The discharge record of the hospital admission that was matched to the index ECG was analyzed. A total of 737 subjects were matched; the remaining 167 unmatched subjects had only outpatient ECGs and were excluded from the study. Seventeen hospitalized subjects did not have an admission period that included the date of the index ECG. The first ECG manifesting AF was paired with the closest hospital admission for each of these subjects.
Clinical Variables
Patient age, sex, and race (Caucasian, Hispanic, African-American, Asian, or other) were ascertained from ECG and hospital discharge database. Specific associated ECG diagnoses [myocardial infarction, left ventricular hypertrophy (LVH), left bundle branch block, right bundle branch block, complete heart block, and nonspecific ST- and T-wave changes] from the hospital ECG database were recorded for each subject with AF from the index ECG. Myocardial infarction included definite and probable cases from ECG over-readers verification. Bundle branch block included both incomplete and complete blocks. Additional ECG parameters (ventricular rate and corrected QT duration) were obtained from the same ECG for each subject. Automated calculations and interpretations were made by the Marquette system software, and over-readers used LVH criteria of Sokolow and Lyons as modified by Romhilt and Estes. Tachycardia was defined as ventricular rate >100. Corrected QT (Bazett’s formula) duration was defined as prolonged if it was >0.44. online pharmacy without a prescription
Hospital Variables: The International Classification of Diseases, Ninth Revision, version 15 [ICD-9] was used to define hospital discharge diagnosis groups: hypertension, diagnosed heart failure, diabetes, coronary heart disease other than acute myocardial infarction, acute myocardial infarction, valvular diseases, rheumatic heart disease, cardiomyopathy, cerebral vascular accidents, and cerebral embolism.
Statistical Methods
Means were calculated for continuous variables and comparisons were made using independent-samples T test. Proportions with 95% confidence intervals were calculated for categorical variables, and comparisons were made using %2 test. Logistic regression was used to calculate odds ratios (OR, with 95% confidence intervals) for pairwise comparisons between Hispanic vs. Caucasian, African-American vs. Caucasian, and Asian vs. Caucasian subjects. The other category was included in the total database but not used for the group comparisons. Statistical significance was considered for p values <0.05. cialis canadian pharmacy
