SBIR Phase I: A Clinical Decision Making tool to improve diagnosis, management and research in rare and genetic disease

The broader impact of this Small Business Innovation Research (SBIR) Phase I project is to potentially improve the diagnosis and management of patients with rare diseases by developing an Electronic Health System integrated artificial intelligence Clinical Decision Support Tool. 1 in 10 people are affected by a rare disease worldwide, half of them are children, and 30% of them will die within the first 5-years of their life due to their disease. On average, it takes 12-15-years from the onset of symptoms to be diagnosed with one of the >10,000 currently known rare and genetic diseases, much longer for patients who reside in rural and underserved communities.

Patients with a rare disease are seen by all medical specialties, but it is not possible for any physician, not even a specialist, to be and remain an expert in the over 10,000 currently known rare diseases, leading to preventable adverse patient outcomes. It costs approximately $28,000 more a year to treat a patient with a rare disease in comparison to a patient with a common chronic disease. 70% of this excess medical cost is carried by governmental single payors such as the Center for Medicare and Medicaid Services.

This Small Business Innovation Research (SBIR) Phase I project aims to develop an Electronic Health Record (EHR) integrated artificial intelligence system that can predict rare diseases in undiagnosed patients based on their patient data alone and give evidence-based, personalized treatment recommendations of already diagnosed patients relevant to the department specialty. With improved and earlier precision management this system can reduce diagnostic delays and prevent adverse outcomes while leading to significant cost savings per patient of up to $28,000 a year, totaling nearly $1 Billion dollars of direct medical cost savings in the US alone per year.

The project utilizes diverse EHR data from various institutions across the US enriched by published data sources such as NIH databases to create predictive algorithms for undiagnosed patients and evidence-based management algorithms for already diagnosed patients using virtual pooling technology; This eliminates the need for patient-level data sharing across institutions and enables wide scalability to any rare disease. This point-of-care EHR-integrated app can be used in any setting worldwide with any patient population as it continuously self-updates locally and globally through bidirectional algorithm sharing.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.