A Data Science Central Community
INSOFE's prior authorization models predict the outcome with a 93% recall and accuracy.
Doctors want to prescribe medicines that do not need further authorization by insurance companies so that patients can pick them up immediately. The churn in cases where prior authorization is needed is very high.
However, the policies for each insurance company are different, dynamic and not well stated and hence it is impossible to keep up.
INSOFE developed a big data solution that looks at millions of past data to predict correctly whether prior authorization would be required for the prescription and got exceptional accuracy on unseen data.
Efforts are now underway to engineer the solution.