C-D2-01: Outcomes Associated With Thiazide Combinations for Second-line Treatment of Hypertension in Older Patients

Abstract

Context: Consensus guidelines recommend thiazides for first-line treatment in uncomplicated hypertension. Most patients require two or more drugs and little is known about the relative efficacy and outcomes of second-line treatments including thiazide. Objective: To compare differences in blood pressure, renal function and medical outcomes between commonly used two drug regimens, each including thiazide. Design: Retrospective cohort using data from a longitudinal electronic medical record. Setting: Large group medical practice. Patients: All patients >60 years treated for hypertension between 2001 and 2006. Main Outcome Measures: Changes in blood pressure and renal function, incident medical events, and time to failure indicated by an event or regimen change. Results: Of 47,419 patients, 6,534 received second-line therapy including thiazide for > 3 months. Thiazides + angiotensin converting enzyme inhibitors (ACE-I), and thiazides + potassium sparing diuretics (PS) were associated with the greatest reductions in systolic pressure, and thiazides + ACE-I with the greatest reductions in diastolic. However, thiazide combinations with PS and ACE-I were associated with the greatest declines in estimated glomerular filtration rate, and these and thiazides + angiotensin receptor blockers (ARB) with increased incidences of renal disease. ACE-I + thiazide were also associated with high incidences of stroke and diabetes. Thiazide + Betablocker (BB) were associated with high rates of cardiovascular disease while thiazide + antiadrenergics (AA) were associated with low event rates overall. Thiazide + Calcium channel blockers (CCB) had no consistent pattern overall, but performed somewhat better in the oldest patients. Thiazides + BB were associated with low composite event rates and the longest time before regimen change. Conclusions: The present study suggests disjunction between blood pressure reduction and the events that treatment is intended to prevent. Thiazide with BB was associated with the best composite outcomes, and thiazide with ACE-I with the poorest. These findings must be interpreted with caution because of potential confounding by indication. Exploration in more robust datasets is warranted.