Proarrhythmic Medicines and Primary Cardiac Arrest

BACKGROUND:

The original grant from 1990 to 1995 was funded because analyses of clinical trial subgroups had raised the concern that, in patients with high blood pressure, diuretic therapy may increase the risk of sudden cardiac death. Given the size of the hypertensive population in the United States, the prevalence of diuretic therapy for hypertension, and the persistent concerns regarding the relation of diuretic therapy to the risk of primary cardiac arrest (PCA), the study proved to be of particular interest to clinicians, epidemiologists, public policy makers, and the general public.

Beginning in 1995 when the grant was renewed, unexpected findings from the Cardiac Arrhythmia Suppression Trial--an adverse effect on mortality of two antiarrhythmic drug therapies--had heightened concerns that drug therapies other than diuretics may increase the risk of primary cardiac arrest.

DESIGN NARRATIVE:

The original study beginning in 1990 was population-based with a case-control design. Using the community-based surveillance system for out-of-hospital primary cardiac arrest in Seattle and King County, Washington, all cases of primary cardiac arrest (PCA) were identified which had occurred among 18,000 pharmacologically-treated hypertensive patients receiving care at Group Health Cooperative (GHC) from 1977-1993. Approximately 180 cases were identified. Controls were obtained from a random sample of GHC enrollees with pharmacologically-treated hypertension, matched to cases at a ratio of 3 to 1, according to age, gender, and year of occurrence of PCA. The computerized pharmacy data base of GHC allowed ascertainment of patterns of exposure to specific antihypertensive drug therapy in an identical fashion for both cases and controls. Medical records were reviewed to gather information about potential confounding factors and effect modifiers, such as severity of hypertension. Data analysis, using stratification and logistic regression, determined whether use of diuretics increased the risk of PCA compared to use of other antihypertensive agents; whether the risk of PCA depended upon the dose of diuretic therapy; and whether electrocardiographic abnormalities modified the risk of PCA associated with diuretics.

The study was renewed in 1995 to determine whether treatment with antidepressant, anticonvulsant, and antiarrhythmic drug therapies having the potential for proarrhythmia increased the risk of primary cardiac arrest. The study was a population-based case-control study nested within a cohort of patients who received medical care at a large pre-paid Health Care Plan in Seattle, Washington. Cases were patients who had a primary cardiac arrest between 1977 to 1994. Controls were a stratified random sample of patients, frequency-matched to cases by age, gender, calendar-year, and known heart disease. Treatment with drugs was assessed through a computerized pharmacy database. Ambulatory-care medical records were reviewed to assess clinical characteristics, including the indication for therapy, the severity of heart disease, co-existing morbidity, and other risk factors. For both antidepressant and anticonvulsant drugs, analyses were stratified by known heart disease, because the risk of treatment might be particularly large among patients with known heart disease. For antiarrhythmic drugs, analyses were restricted by a single, current indication for the therapy--maintenance of sinus rhythm among patients with chronic atrial fibrillation; and, by the availability of a prior echocardiogram, in order to control for the type and severity of underlying heart disease. After adjustment for potential confounders, the investigators estimated the relative safety of: 1) drugs within the same therapeutic class; and, 2) the dosage schedule for specific drugs. In addition, they determined if concurrent treatment with other drugs that altered cardiac conduction or morbidity that altered drug disposition influenced the risk among patients treated with a drug therapy.