The Problem
The U.S. Food and Drug Administration (FDA) utilizes several systems to collect adverse event reports. For example, the FDA Adverse Event Reporting System (FAERS) is a database that contains adverse event reports, medication error reports, and product quality complaints submitted following exposure to drugs and biological products. During the FAERS launch in 2015, the FDA issued an Open Data Challenge to encourage the development of computational algorithms for the automatic detection of drug adverse event anomalies.
The Solution
Infolytx analyzed both structured medical content FAERS and unstructured data from social media such as Twitter. We applied NL techniques to parse and normalize the data and then used ML models to build multiple predictive and actionable models. In addition, we created an API for the models to be accessible as a SaaS supporting multiple user bases and use cases.
The Outcome
We demonstrated the successful application of the SaaS solution by building several demonstration mobile apps for predicting previously unknown adverse reactions between two drugs.
- My Virtual Pillbox: Patients can query a proposed OTC purchase against their prescriptions to determine if any adverse reactions are reported.
- Prescription Checker: Physicians can query proposed prescriptions against a patient’s existing prescription to determine if any previously unknown adverse reactions are reported.