Applying digital phenotypes and epigenetics to understand clinical outcomes in acute hypoxemic respiratory failure without acute respiratory distress syndrome (NAHRF) - PROJECT SUMMARY Each year approximately 1 million adults receive invasive mechanical ventilation in the United States for acute hypoxemic respiratory failure (AHRF), and 30-40% of these patients die in the hospital. Most research has focused on acute respiratory distress syndrome (ARDS), a severe form of AHRF that affects one quarter of mechanically ventilated adults. Despite over 50 years of research, few pharmacotherapies for ARDS have shown benefit in clinical trials and, although more common than ARDS with comparable mortality rates, there are no targeted therapies for non-ARDS AHRF (NAHRF). These failures result, at least in part, from a “one- size-fits-all” approach to treatment. This K23 proposal aims to advance management of critically ill adults with NAHRF toward a phenotype-driven, individualized approach that will 1) Improve clinical outcomes through more precise application of existing and future therapies and 2) Inform future research by enabling phenotype- based clinical trial screening and enrollment strategies. The Specific Aims are: 1) Develop clinical prediction models for mortality in adults with NAHRF; 2) Apply machine learning to detect subphenotypes of NAHRF associated with differential clinical outcomes; and 3) Use sequencing-based whole methylome data to better understand the biological mechanisms contributing to these outcomes. With these Specific Aims and a career development plan designed to fill gaps in her training, Dr. Arnold will become an independent clinical investigator. The career development plan includes didactics, applied learning, and training with her mentors to: 1) Learn advanced modeling methods for risk stratification and digital phenotyping; 2) Acquire expertise in the discovery and application of novel biological markers; and 3) Obtain the experience and leadership skills needed to conduct prospective clinical studies in the acute care setting. Dr. Arnold has assembled a multidisciplinary team of mentors and advisors with extensive track records of research success, mentorship, and collaboration in clinical and translational research. Supportive and well-resourced with internationally recognized leaders in clinical prediction model derivation and validation, clinical trials, emergency care research, and bioinformatics, UC Davis has a track record of excellence in patient-oriented research and provides an ideal environment to support Dr. Arnold’s development into a national leader in precision-guided approaches to NAHRF.
