AI Use Case on Adjusting Targets by Propensity to Prescribe

Impact

Next Best Action triggers proactively identified HematologyOncology HCPs with high propensity to prescribe target brand

Optimal targeting with non-personal promotion (NPP) resulted in 25 new prescribers’ worth $472k within one month of operalization of Next Best Action recommendation

Background

Despite large volume of digital promotion efforts by agencies, less than 1/3rd of targeted HCPs were prescribing the target brand. Brand team sought to identify HCPs with highest likelihood of prescribing in future. Then determine the Next Best Action in terms of NPP to maximize future Rx volume.

Objectives

Increase target brand Rx by targeting accounts with the right NPP tactics.

Identify drivers of adoption and growth from amongst volumetric and promotional factors as well as static account attributes

For already prescribing accounts assign a growth likelihood score. For non-prescribing accounts assign an adoption likelihood score. Segment accounts into High-Med-Low based on scoring

Solution

A rolling 12-month window of time was defined for predicting future Rx volumes for each HCP. With a focus on simulating impact of NPP tactics, personal promotions were included as control variables. NPP activity was modelled for each agency: BioPharm, Epocrates, Medscape, Watzan, and others.

A decision-tree based approach yielded accurate P2Rx (propensity to prescribe) scores for 7500+ HCPs within Hematology-Oncology. A simulation model was used to generate Next Best Action recommendations for NPP at a vendor/agency level.

ai-ml-pic2