As a clinical epidemiologist, Stephen Deppen, PhD, evaluates and seeks to improve the diagnosis, treatment and outcomes of various cancers. He is an Associate Professor of Thoracic Surgery, clinical epidemiologist and co-director of the Vanderbilt’s Thoracic Biorepository and the Early Detection Research Network’s Lung Group National Clinical Validation Center. Through these roles he has extensive experience in the validation and clinical evaluation of imaging, biomarkers, and diagnostic tests with a focus on the early detection of lung cancer.
Dr. Deppen researches the validation and measurement of clinical utility in biomarkers to diagnose indeterminate pulmonary nodules (IPNs), discovery of learning curves among medical device implants with machine learning tools and development of Veteran specific lung cancer screening models, which includes HIV as a risk factor. He has extensive experience in the validation and clinical evaluation of imaging, biomarkers, and diagnostic tests, cost effectiveness and implementation science with a focus on the diagnosis and treatment of cancer.
His modeling expertise includes geospatial, predictive risk, quantification of evidence and longitudinal modeling. This focus on issues relevant to the evaluation of indeterminate lung nodules has resulted in groundbreaking work that has changed current guidelines for nodule diagnosis and quantified the limitations of our imaging technology. Since the beginning of his research career, Dr. Deppen has worked directly with clinicians and scientists to bring evidence to the bedside and improve healthcare across a variety of diseases and their interventions.
His economics experience helped address questions regarding diagnostic imaging effectiveness, the costs of delirium, and is being applied to the cost effectiveness of strategies to improve HPV vaccination rates in a randomized implementation trial. He is a member of the Early Detection Research Network (EDRN) in the areas of epidemiology and biomarker validation and have been supported by both pilot and core funds from the network to investigate the diagnosis of granulomas caused by histoplasmosis and develop imaging based deep learning models to improve diagnostic low-dose lung cancer screenings.