Tutorial- Monday at 11:15 AM EDT
Title: SMART Treatment Decisions: Predicting 10-Year Cardiovascular Event Risks & Assessing Treatment Thresholds in a UK Population
Presenters: Laura Gunn, UNC-Charlotte

Abstract:
Future vascular event risk varies broadly among those with atherosclerotic cardiovascular disease (ASCVD). Therefore, the ability to predict individual risks allows for more personalized approaches to risk management and treatment. This validation study assesses performance of the SMART model in predicting 10-year cardiovascular event risks. Population-level data from the Clinical Practice Research Datalink consists of adults registered with UK National Health Service primary care providers diagnosed with coronary, cerebrovascular, peripheral, and/or aortic ASCVD. Risk factors include demographics, medical history, and clinical measurements, while the outcome is first post cohort-entry occurrence of non-fatal myocardial infarction, non-fatal stroke, or cardiovascular death. The model is tested against a range of potential treatment thresholds, and compared to treat-none and the current treat-all benchmark used in medical practice. A net gain from using the SMART model for treatment discrimination is found across a range of reasonable thresholds for treatment. This gain is relevant for both medical and financial reasons. From a clinical standpoint, treat-all solutions lead to unnecessary overtreatment of lower-risk cases. From a financial standpoint, maintaining the treat-all status quo would reach a funding barrier as more expensive treatment options become available. Therefore, adequate identification of those at most risk allows for delivery of increasingly advanced treatment options only to those above a clinical risk threshold. The calibration and discrimination achieved by the SMART model was not dissimilar to the reported performance within an internal validation, based on a limited Dutch cohort 100 times smaller than those in our study. This demonstration of translational medicine to different populations even under incomplete covariate information is compelling. Our study shows that the model slightly under-predicted risk among lower risk groups, but clinical utility was apparent across potential treatment thresholds and best among those most at risk. Results remained consistent across sensitivity analyses. The SMART model has clinical utility in the context of routine UK primary care-based secondary prevention of cardiovascular disease.

Bio:
Dr. Laura H. Gunn received her PhD in Statistics and Decision Sciences from Duke University, during which time she also held a research training fellowship in Biostatistics with the National Institute of Environmental Health Sciences. Laura is currently Associate Professor of Public Health Sciences, Director of Biostatistics Core, and Affiliate Faculty in the School of Data Science at the University of North Carolina at Charlotte, as well as Honorary Research Fellow at Imperial College London’s (ICL) School of Public Health (SPH) within the Faculty of Medicine. Prior, Laura was Associate Professor of Public Health in Biostatistics, Department Chair of Health Sciences, and founding Program Director of Public Health at Stetson University. Additional prior positions include Associate Director of the Global eHealth Unit within ICL’s SPH. During this time, she served as Lead Biostatistician for Research Design Service London at ICL. Laura was also Biostatistics Director and Interim Associate Dean of Georgia Southern University’s Jiann-Ping Hsu College of Public Health. Her research accounts for over 60 peer-reviewed journal articles, book chapters, and technical reports, as well as serving as PI or co-I on funded grants and contracts totaling over $6.5 million, including National Institutes of Health (US) and National Institute for Health Research (UK) grants. She has also co-authored/co-presented over 175 invited and contributed regional, national, and international presentations. Laura previously served as BASS Chair (2007-11) and Co-Chair (2005-06), and she currently serves on the BASS Program Committee (2004-11; 2018-Present).