Abstract

BACKGROUND: Analyses of clinical trials find that an intensive systolic blood pressure (SBP) target of less than 120 mm Hg is cost-effective compared with a target of less than 140 mm Hg for patients at high cardiovascular disease risk. However, guidelines from the American College of Cardiology and American Heart Association recommend a target of less than 130 mm Hg, citing blood pressure measurement error in routine practice. OBJECTIVE: To evaluate the effect of measurement error on the cost-effectiveness of intensive SBP targets. DESIGN: Microsimulation model varying SBP measurement error. DATA SOURCES: SPRINT (Systolic Blood Pressure Intervention Trial) data and published literature. TARGET POPULATION: Patients at high cardiovascular risk. TIME HORIZON: Lifetime. PERSPECTIVE: Health care sector. INTERVENTION: SBP targets of less than 120 mm Hg, less than 130 mm Hg, and less than 140 mm Hg. OUTCOME MEASURES: Incremental cost-effectiveness ratios (ICERs). RESULTS OF BASE-CASE ANALYSIS: With research-grade SBP measurement (mean error, 0 mm Hg), the ICER for the target of less than 120 mm Hg versus less than 130 mm Hg was $24 400 per quality-adjusted life-year (QALY). With average measurement error (mean error, 7.3 mm Hg in the <120-mm Hg target), the ICER increased to $42 000 per QALY. RESULTS OF SENSITIVITY ANALYSIS: The ICER for the target of less than 120 mm Hg was greater than $100 000 per QALY in scenarios with high error (mean error, ≥14.6 mm Hg in the <120-mm Hg target), when an inflection point for increasing risk for cardiovascular disease (CVD) was at or above 116 mm Hg, and in scenarios with a medication-taking disutility of at least 0.003 per antihypertensive medication. LIMITATION: Uncertainty in the relationship between low treated SBP (for example, <115 mm Hg) and cardiovascular risk. CONCLUSION: For SPRINT-eligible patients at high cardiovascular risk without diabetes or prior stroke, a target of less than 120 mm Hg seems cost-effective across most settings with SBP measurement error. In scenarios with high error and an increase in CVD risk at low SBPs, a target of less than 130 mm Hg may become cost-effective. PRIMARY FUNDING SOURCE: National Science Foundation and National Institute of Neurological Disorders and Stroke.