PROJECT ABSTRACT Nonalcoholic fatty liver disease (NAFLD) is a public health crisis affecting one third of the US adult population. NAFLD is an obesity related disease with serious health and quality of life consequences including type 2 diabetes, cirrhosis, liver cancer, and cardiovascular disease. Despite serious threats to health and wellbeing, NAFLD has not received the same attention as other cardiometabolic diseases like type 2 diabetes. In addition, NAFLD is often asymptomatic and patients are frequently diagnosed when hepatic steatosis is reported as an incidental finding on imaging studies or a patient develops advanced liver disease. Our prior work found hepatic steatosis reported in just under 10% of written radiology reports. Unfortunately, lack of awareness often leads to hepatic steatosis being overlooked by providers and unreported to patients leading to delays in diagnosis and referral to specialty care. Our recent survey revealed only 13% of providers always reported this finding to patients. This project, Improving Diagnostic Safety through STeatosis Identification, Risk stratification, and Referral pathway in the ED (STIRRED), proposes to improve diagnostic safety by ensuring that patients receive timely notification of the new finding and referral to follow up care. To date, we have developed the building blocks for the STIRRED CDSS by employing natural language processing and machine learning to identify hepatic steatosis in written emergency department (ED) radiology reports and combine it with common clinical data and comorbidities in the electronic health record to create an individualized risk profile and follow up recommendations for patients with suspected (NAFLD). Providing risk stratified appropriate referrals for patients from the ED will increase diagnosis of NAFLD, enable those deemed high risk for advanced liver disease to receive specialist referral and begin the important process of cancer screening and intervention to improve steatosis. Communicating this incidental imaging finding and initiating referral from the ED will enable linkage to care to improve health outcomes and reverse or slow disease progression for those at all risk levels. While the groundwork for this intervention has been laid, for STIRRED to be successful, it must be both refined further and implemented thoughtfully. Through Aim 1 and 2, we will seek to: 1) determine the barriers, facilitators, and additional strategies to implement the STIRRED CDSS, and 2) conduct usability testing of the STIRRED CDSS. Upon finalizing the optimized STIRRED CDSS and implementation plan, we seek to: 3) evaluate STIRRED across a large health system. This work will be led by a team of experienced researchers with expertise in clinical trials, risk communication, patient communication, diagnostic uncertainty, hepatic steatosis and NAFLD, incidental findings, implementation science, and health services research. Completion of these aims will result in tools to improve care transitions and diagnostic safety for patients with hepatic steatosis and suspected NAFLD. Findings will inform future work to implement STIRRED in EDs nationally and develop ED-based interventions for patients with suspected NAFLD.

PROJECT NARRATIVE Hepatic steatosis is a common radiographic “incidental finding” that is overlooked and underreported to patients. We developed a clinical decision support system using machine learning and natural language processing that will prompt reporting to patients and provide ED clinicians risk stratified follow-up care recommendations. Data on both the implementation and effectiveness of our intervention resulting from this trial will inform future use with a goal of ultimately improving diagnostic safety and outcomes for patients with hepatic steatosis.