Abstract 089: A Simple Algorithm Identifies Hypertensive Patients Who Benefit From Intensive Blood Pressure Lowering

Abstract

Background: Large randomized trials have provided inconsistent evidence regarding the benefit of intensive BP lowering in hypertensive patients. Identifying which patients derive a higher net benefit is essential in informing clinical decision-making. Objectives: To assess whether stratification by cardiovascular disease (CVD) risk will identify patients with a more favorable risk/benefit profile for intensive BP lowering. Methods: We used patient-level data from two trials that tested intensive vs. standard BP lowering: SPRINT and ACCORD. Within SPRINT, we selected a subset of patients at extremes of major adverse cardiovascular event (MACE) rates to develop a decision-tree using recursive partitioning modeling. We then validated its predictive effects in the remaining ‘intermediate’ SPRINT subset (n=8,357) and externally in ACCORD (n=2,258). Results: Recursive partitioning produced a three-variable decision-tree model consisting of age≥74 years, urinary albumin/creatinine ratio (UACR) ≥34, and history of clinical CVD. It classified 48.6% of SPRINT and 55.3% of ACCORD patients as “high-risk”. Compared with standard treatment, intensive BP lowering was associated with lower rates of MACE in this high-risk population in both SPRINT cross-validation data (HR=0.66, 95% CI 0.52-0.85) and ACCORD (HR=0.67, 95% CI 0.50-0.90), but not in the remaining low-risk patients (SPRINT: HR=0.83, 95% CI 0.56-1.25; ACCORD: HR=1.09, 95% CI 0.64-1.83). Additionally, intensive BP lowering did not confer an excess risk of serious adverse events in the high-risk group. Conclusions: A simple risk prediction model consisting of age, UACR, and clinical CVD history successfully identified a subset of hypertensive patients who derived a more favorable risk/benefit profile for intensive BP lowering.