Since 1953 PV over 460 products of adverse drug reactions. One of the major challenges that drug safety ex- perts face when conduction pharmacovigilance (PV) is the many disparate evidence sources they have to con- tend with when identifying, prioritizing, and evaluating PV signals. Fewer patients would experience serious harm from drug therapies if PV experts had access to user friendly tools to help them more efficiently search, retrieve, evaluate, and synthesize drug safety evidence across all relevant sources. It is currently very difficult for experts to accurately synthesize evidence across multiple sources because each source has non- intuitive strengths and limitations, uses differing terminologies, and is available in different data formats. The goal of this Phase I SBIR project is to develop the first pharmacovigilance signal generation, prioritization, and validation platform to integrate the major pharmacovigilance evidence sources into a user-friendly application based on scalable cloud-based web services. The project team will use a systematic scientific approach to build a new working prototype for a commercial drug safety research platform called the Common Evidence Model for Advanced PHarmacOvigilance REsearch (Cemaphore):
Aim 1 : Evaluate a novel automated PV signal prioritization procedure across all the evidence sources in Cemaphore: We will develop and validate three novel algorithms that, together, utilize the growing number of evidence sources within the Cemaphore to assess PV signal strength, novelty, and public health impact.
Aim 2 : Test the effect on PV signal prioritization of including evidence from predication-based seman- tic indexing (PSI) of Structured Product Labeling: Prior work has focused on PSI of only the indexed scien- tific literature (i.e., MEDLINE). We will generate semantic predications using SPLs for all prescription drugs, integrate this novel source into Cemaphore, and then test if the novel source improves PV signal generation.
Aim 3 : Conduct formative usability testing of a working prototype web UI for PV signal prioritization and validation built with web services: We will develop a Cemaphore web UI that provides a single unified view across harmonized evidence sources and publish web services to encourage integration and innovation in the wider eco-system of safety/PV applications. The services will provide a secure, performant, cloud host- ed, REST API to access Cemaphore-generated PV signals for hypothesis generation, prioritized PV evidence metrics, and positive/negative controls for empirical calibration of observational study estimates. We will en- gage PV experts in think aloud tests of a functional user interface that integrates these services into PV signal investigation workflows. This SBIR Phase I project will demonstrate the technical merit and feasibility of Cemaphore for advancing ac- tive post-market risk identification and analysis. The project will contribute to public health by greatly lowering the significant barriers that drug safety professionals face when investigating medication safety signals.
The goal of this Phase I SBIR project is to develop the first pharmacovigilance signal generation, prioritization, and validation platform to integrate the major pharmacovigilance evidence sources into a user-friendly applica- tion based on scalable cloud-based web services. The project team will use a systematic scientific approach to build a new working prototype for a commercial drug safety research platform called the Common Evidence Model for Advanced PHarmacOvigilance REsearch (Cemaphore). The project will contribute to public health by greatly lowering the significant barriers that drug safety professionals face when investigating medication safe- ty signals.
