Abstract T MP100: Improving Encounter Frequency, Physician Prescribing, or Patient Adherence - What Works Best for Controlling Blood Pressure? Results From a Validated Micro-Simulation Model

Abstract

Background: Only half of hypertensive adults achieve blood pressure (BP) control in the United States. Methods: We developed the BP Control Model to simulate physician- and patient-level processes relevant in achieving BP control. We validated the model by simulating the intervention arm of a recent multicenter clinical trial and used the validated model to examine the effects of isolated improvements in three modifiable processes on BP control. Data from national surveys, cohort studies, and trials were used to parameterize the model. We simulated 5,000 hypothetical adult patients with uncontrolled hypertension (systolic BP≥ 140) using probability sampling of participants from the 2009-2010 National Health and Nutrition Examination Survey. We modeled 50% improvements and ideal scenarios for each process parameter. Outcome: We reported outcomes in terms of BP control (% with SBP < 140 mmHg), and average change in BP at 52 weeks. Results: In our validation analysis, the model-predicted BP control was similar to what was achieved in the VIPER-BP clinical trial (63.5% vs. 63.8%). In our base case scenario, 24% of the NHANES-derived cohort achieved BP control at 52 weeks. In scenarios with 50% improvements, the model predicted small increases in BP control, but substantially larger and more variable effects when processes were idealized (Table). Control was reached in 77% with ideal treatment intensification, 44% with ideal encounter frequency, only 32% with ideal adherence, and 97% when all three processes were idealized. Conclusion: While improving patient adherence to medications would improve BP control, healthcare systems can achieve similar or greater success by focusing on increasing the frequency of clinical encounters and improving physicians’ prescribing behavior. The BP Control Model can be used to predict how much improvement to expect from such interventions.