PROJECT SUMMARY/ABSTRACT There is a profound need to improve venous thromboembolism [VTE] treatment and prevention, particularly in patients with inflammatory bowel disease [IBD]. This Pathway to Independence Award application is submitted by a pharmacoepidemiologist committed to improving VTE prevention and treatment in populations who are often excluded or vastly under-represented in randomized clinical trials [RCTs]. Of the 3 million Americans affected by IBD, ~1–7% will develop VTE, a significant source of morbidity and mortality, in both the outpatient and inpatient settings. While attention has been paid to the role of anticoagulants for VTE prevention in IBD patients, there are key barriers to prescribing anticoagulants in this population including: (1) absence of data on the effectiveness of anticoagulants for the prevention of VTE in IBD patients; (2) fear of bleeding complications; and (3) lack of evidence on the safety of anticoagulants in high-risk subgroups such as patients with active disease flares or postpartum IBD patients, who are at an increased risk of adverse complications. The overarching goal of this research is to evaluate the use, effectiveness, and safety of anticoagulants in IBD patients in the inpatient and outpatient settings. The applicant will achieve the proposed aims of this K99/R00 award under the guidance of established researchers who span the disciplines of epidemiology, biostatistics, gastroenterology, and hematology at the University of Pennsylvania. First, the applicant will use a combination of machine learning and advanced methodological techniques to examine predictors of use, effectiveness, and safety of anticoagulants in the inpatient setting (K99 phase). In the R00 phase, the applicant will use linked electronic health records/claims database to examine the comparative effectiveness and safety of direct oral anticoagulants, warfarin, and heparins in IBD patients in the outpatient setting. The research proposed in this K99/R00 is critical since it provides timely needed information on the risks and benefits of anticoagulants in IBD patients, data that will not be forthcoming from large RCTs in the foreseeable future. The proposed career development and training goals will provide the applicant with training in predictive modeling, time-varying adjustment, and machine learning. The successful completion of the proposed training, infrastructure, and institutional support at the University of Pennsylvania will guarantee the applicant's success and transition to scientific independence.

PROJECT NARRATIVE Randomized clinical trials of anticoagulants vastly under-represented patients with inflammatory bowel disease [IBD] despite their increased risk of venous thromboembolism [VTE]. Using inpatient data, outpatient data, machine learning algorithms, and data linkage, the applicant will evaluate the risks and benefits of anticoagulants in IBD patients.