Forensic investigation using data science helps pharmacy chain uncover pricing error
Learn how Baker Tilly’s and Forensic, Litigation and Valuation Services teams helped a company investigate rebate discrepancies from their supplier.
Client background
The client is a regional pharmacy chain with 25 locations throughout three states.
The business challenge
After multiple months of rebating being less than expected, the client needed a full audit of the 340B rebate process from their supplier. They needed help analyzing the data of their buying behavior and why they weren’t receiving accurate drug credit and rebate amounts from their supplier. They also needed an analysis of their generic compliance ratio (GCR), or what percentage generic drugs they were selling compared to name brand, and how they were being compensated for those.
Strategy and solution
Baker Tilly's digital team analyzed multiple years of the client and their supplier’s data, including purchasing history, the pricing catalog from the supplier and how their drug credits and rebate calculations were being applied to their account. Amazon Athena and AWS Lake Formation were used to create a foundation for database storage and queries, then we used Amazon SageMaker for statistical testing and anomaly detection and Amazon QuickSight for data visualizations and dashboarding.
This project was in collaboration with Baker Tilly’s Forensic, Litigation and Valuation Services team. They provided the forensic analysis and background into a forensic investigation, while the Baker Tilly digital team supported the data science approach by helping facilitate how the analytic questions were framed and creating data requests to make sure the client and the supplier were providing the right data sets from the correct sources during the correct timeframes. Through this approach, we created evidenced based conclusions (that would answer the research questions we helped formulate) to develop an audit report that documented our procedures, the data sources considered, test results and our findings.
As a result of this analysis, Baker Tilly found a $75,000/year error and recommended to the client that they start using more current pricing to calculate the credit amounts given relative to the 340B rebate, which gave them a boost in their rebate amounts. We also confirmed the pricing being utilized by the supplier was compliant with their price catalog, which provided assurance to the client that they were being charged the proper amount. Any anomalies in pricing were explained through our independent verification. Additionally, the review of the auto substitution process confirmed the process was working correctly and the client was receiving the correct credits and billing pricing.