How a biopharmaceutical company evaluated social media to monitor off-label drug use
Customer
Global biopharmaceutical company
Goal
Gather intelligence about off-label drug usage by screening patient comments in social media
Challenges
- Insights about off-label drug usage is difficult to collect for pharmeutical companies which do not have direct contact to patients
- In the context of pharmacovigilance, they need to monitor off-label use of their drugs in order to comply with regulations
- Patient records are subject to privacy restrictions and when available they do not allow to reconstruct which medication was prescribed for which condition
Solution
During the initial phase of the project, Cortical.io created a Pharma Intelligence prototype based on SemanticPro Classify & Automate to identify mentions of on- and off-label medication usage in a static set of Reddit posts. Leveraging Cortical.io meaning-based algorithms, the application automatically and accurately filters and classifies Reddit messages and summarizes the results. Cortical.io trained classifiers for each of the example drugs specified by the company using publicly available information.
This solution automatically
- reviews Reddit posts mentioning medication
- filters for trade or generic names of medication
- filters out off-topics and ambiguous posts
- classifies posts by medical conditions
- identifies on- versus off-label usage
See how the Pharma Intelligence solution works in this short demo
Results
- 2.2 million Reddit post reviewed
- 92-100% accuracy in classifying on-label versus off-label usage
The Cortical.io Difference
- SemanticPro Classify & Automate combines human-level accuracy with fast processing and high scalability
- The solution leverages Semantic Folding to accurately disambiguate posts mentioning multiple medication and filter out off-topic posts
- The solution is able to accurately process terms it has never seen before
