Cost-Effectiveness of Intensive versus Standard Blood-Pressure Control

Abstract

Background: In the Systolic Blood Pressure Intervention Trial (SPRINT), adults at high risk for cardiovascular disease who received intensive systolic blood-pressure control (target, <120 mm Hg) had significantly lower rates of death and cardiovascular disease events than did those who received standard control (target, <140 mm Hg). On the basis of these data, we wanted to determine the lifetime health benefits and health care costs associated with intensive control versus standard control.

Methods: We used a microsimulation model to apply SPRINT treatment effects and health care costs from national sources to a hypothetical cohort of SPRINT-eligible adults. The model projected lifetime costs of treatment and monitoring in patients with hypertension, cardiovascular disease events and subsequent treatment costs, treatment-related risks of serious adverse events and subsequent costs, and quality-adjusted life-years (QALYs) for intensive control versus standard control of systolic blood pressure.

Results: We determined that the mean number of QALYs would be 0.27 higher among patients who received intensive control than among those who received standard control and would cost approximately $47,000 more per QALY gained if there were a reduction in adherence and treatment effects after 5 years; the cost would be approximately $28,000 more per QALY gained if the treatment effects persisted for the remaining lifetime of the patient. Most simulation results indicated that intensive treatment would be cost-effective (51 to 79% below the willingness-to-pay threshold of $50,000 per QALY and 76 to 93% below the threshold of $100,000 per QALY), regardless of whether treatment effects were reduced after 5 years or persisted for the remaining lifetime.

Conclusions: In this simulation study, intensive systolic blood-pressure control prevented cardiovascular disease events and prolonged life and did so at levels below common willingness-to-pay thresholds per QALY, regardless of whether benefits were reduced after 5 years or persisted for the patient's remaining lifetime. (Funded by the National Heart, Lung, and Blood Institute and others; SPRINT ClinicalTrials.gov number, NCT01206062 .).

Figure 1. Structure of the SPRINT Simulation Model

Shown is the microsimulation model used to estimate costs, clinical outcomes, and quality-adjusted life-years of intensive control of systolic blood pressure in adults who were eligible to participate in the Systolic Blood Pressure Intervention Trial (SPRINT). (A complete list of the eligibility criteria for participation in SPRINT is provided in the Methods section in the Supplementary Appendix.) Panel A shows the two interventions — intensive control and standard control of systolic blood pressure — and health states of the patients, and Panel B shows the three categories of subsequent clinical events: cardiovascular disease (CVD) events, serious adverse events, and death from causes other than cardiovascular disease. The blue square indicates the decision node, the point at which a treatment strategy is chosen; the purple encircled letter “M” indicates the Markov node, with branches indicating the health states in transition every 6 months; the green circle indicates the chance node, after which there is a probability of the occurrence of each event; and the red triangle indicates the terminal node, the end of a pathway within a 6-month cycle. ACS denotes acute coronary syndrome, and MI myocardial infarction.

Figure 2. Incidence Rate Ratios for the Primary Outcome and Incremental Direct Medical Costs for Intensive versus Standard Control, According to Four Scenarios for Medication Adherence and Treatment Effect

Panel A shows incidence rate ratios for the SPRINT primary outcome (the first occurrence of myocardial infarction, acute coronary syndrome not resulting in myocardial infarction, stroke, heart failure, or death from cardiovascular causes) for intensive control versus standard control of systolic blood pressure during the simulation over different time periods. The results are shown according to the four post-trial persistence-of-treatment-effect scenarios: base case (i.e., reduced adherence to the medication regimen and treatment effects after 5 years until total nonadherence and no treatment effects at 15 years), worst case (i.e., nonadherence and no treatment effects after 5 years), best case until 15 years (i.e., in-trial adherence and persistence of treatment effects for 15 years), and lifetime best case (i.e., lifetime in-trial adherence and persistence of treatment effects). Although the assumptions and input were identical for all four scenarios for the first 5 years of the simulation, there were small differences in the incidence rate ratios for cardiovascular disease events for the period from 0 to 5 years that reflect the role of chance in the microsimulation approach. The I bars indicate 95% confidence intervals. Panel B shows the range of mean cumulative incremental direct medical costs of intensive control versus standard control of systolic blood pressure, according to the expenditure — including costs associated with serious adverse events, treatment, background health care for the treatment of noncardiovascular diseases, chronic cardiovascular disease (CVD), or CVD event — in the four post-trial persistence-of-treatment-effect scenarios over time.

Figure 3. Probability of Cost-Effectiveness of Intensive versus Standard Blood-Pressure Control

Shown is the probability of the cost-effectiveness of intensive control of systolic blood pressure, as compared with standard control, according to a range of willingness-to-pay thresholds (the cost in dollars per quality-adjusted life-year [QALY]). The curves represent the four post-trial persistence-of-treatment-effect scenarios. The curves were generated from the results of the probabilistic analysis in which the model was run 1000 times with the use of random draws for all model measurements to capture joint uncertainty in the model results.