We were retained by a client in the healthcare industry that wanted to understand exact reasons behind patient leakage and the necessary actions to capture lost revenue
We analyzed large amounts of data and delivered a detailed report by completing the following tasks:
- Cleaned and transformed large datasets, including electronic medical records (EMRs) and insurance claims data
- Analyzed a very large number of input variables to calculate feature weights and importance in influencing patients’ decisions to leave the client for another healthcare provider
- Applied several machine learning methods, such as gradient boosting, deep learning, and random forest among others, selecting the best method based on accuracy in predicting patients’ decisions
- Generated several dashboards, graphs, and diagrams that visualized data patterns and insights that depicted ways to capture lost revenue
The client was very pleased with the actionable insights we provided to capture lost revenue. The strategy and implementation defined by our recommended actions predicted that 30% of lost revenue could be retrieved within six months.