Building Equitable and Feasible Clinical Decision Support Tools for Population Health Management of People Living with Chronic Lung Disease (CLD CDSS)

People living with a chronic lung disease (CLD) require precision delivery of care, focused on actionable risk mechanisms. Prior work on stratifying risk in this population has focused almost exclusively on risks associated with 30-day hospital readmission. This approach is limited because risks are not always the actual reason a patient is hospitalized, and therefore may not be useful when deploying interventions to help reduce risk. Furthermore, the importance of reducing hospital admissions for community-dwelling people living with CLD has been overlooked.

There is an opportunity for artificial intelligence (AI) to help address this issue, however, not enough research has been done to understand the best ways to deploy AI tools and how to assess how well they perform when integrated in real clinical workflows.

To address this knowledge gap, our team will create an AI-based clinical decision support system to promote population health management for patients with CLD. This AI tool will use electronic health record data to identify high-risk individuals with CLD who may benefit from five common frailty- or depression-focused interventions (including pulmonary rehab, physical therapy, depression screening, nutrition screening, and a fall-reduction program).

Then, our team will conduct a pilot feasibility study, where the AI clinical decision support system will be embedded in a clinical workflow with pulmonologists and their patients with CLD. AI will predict patient risk and make referral recommendations for appropriate patients. Pulmonologists will then review these referral recommendations. If in agreement that a patient is appropriate for a specific referral, pulmonologists will work with the study team to implement the referral recommendation for their patient. The goal is to assess the feasibility and acceptability of this AI tool in a real-world clinical workflow. Those participating will complete a survey and interview to assess their feelings and thoughts about this process involving the AI tool.

National Heart, Lung, and Blood Institute (NHLBI)