Blood Pressure and Kidney Function - SPRINT vs Electronic Health Record Data

点击翻译按钮获取中文摘要

Project Summary/Abstract Hypertension is prevalent in 68.5 million U.S. adults and is one of the leading chronic disease risk factors in the world. The Systolic Blood Pressure Intervention Trial (SPRINT) evaluated the effect of intensive (<120 mmHg) and standard (<140 mmHg) BP targets on cardiovascular events and all-cause mortality. Intensive BP lowering was associated with a 25% reduction in cardiovascular events and a 27% reduction in all-cause mortality. BP measurement in SPRINT followed guideline adherent protocols. Unfortunately, BP measurement in routine clinical practice is often not adherent to American Heart Association recommendations; inappropriate technique is associated with an overestimation of BP by approximately 5 to 15 mmHg. Therefore, the achieved BP in routine clinical practice during the treat to target phase of SPRINT is unknown. An additional limitation of SPRINT is that acute kidney injury (AKI) was only assessed in the inpatient setting and adjudication was based on billing codes and discharge summaries, not the gold standard for assessing AKI – creatinine values. Fortunately, SPRINT was one of the first large trials performed in the era of electronic health records (EHRs). The timing of the SPRINT trial provides a unique opportunity to link SPRINT data to EHR data to perform analyses that would be otherwise infeasible within either data source alone. The goals of the proposed study are to a) inform implementation of the SPRINT results by comparing contemporaneous BP in the routine clinic setting to those measured at SPRINT study visits, b) provide creatinine values to assess the effect of intensive and standard BP targets on rates of AKI and other renal outcomes during the trial and in long-term, post-trial follow-up, and c) inform the design of future pragmatic clinical trials by reporting on the concordance between EHR based and formal SPRINT based assessment of renal and cardiovascular outcomes. For the proposed study, SPRINT and EHR data (BP and creatinine values and ICD-9/10 codes) from a number of the larger SPRINT sites will be linked. The specific aims are to evaluate 1) the concordance between BP measured in the general clinic setting to BP measured using guideline recommended technique at SPRINT research visits; 2) the effect of intensive versus standard BP targets on renal and cardiovascular outcomes using SPRINT and EHR based creatinine values; and 3) the concordance between EHR based and formal SPRINT based assessment of renal and cardiovascular outcomes to inform the design of future pragmatic clinical trials. Our results will support physicians and clinic managers in their efforts to implement appropriate BP measurement protocols and will inform implementation of the SPRINT results. Finally, this study will be the first to link clinical trial data with EHR data on a large scale and will serve as an example for how these types of data can be merged to leverage the strengths of both data sources.