Preparing Data for analytic solution

Step 1 (Done) Please review and revise if necessary.

Summary of Analytical Problem Requiring Risk Adjustment

I chose the topic of healthcare fraud because it is a serious problem costing billions of

dollars each year in the US healthcare system. Typical fraud committed by health

providers include:

• Double billing: Submitting multiple claims for the same service

• Phantom billing: Billing for a service visit or supplies the patient never received

• Unbundling: Submitting multiple bills for the same service

• Upcoding: Billing for a more expensive service than the patient actually received

The analytics solution focuses on identifying the anomalies to detect fraud. It includes detecting values that exceed standard deviation averages, besides an analysis of high and low values to detect abnormalities, which often indicate the likelihood of fraud. Another method is to group the data based on specific criteria such as the geographical location of events and other complex patterns not found in other ways.

It is often difficult to compare healthcare providers without adjusting for the conditions of the patients and other factors related to patient health. Risk adjustments are needed to arrive at a fairer comparison.

The conceptual steps of performing risk adjustment include creating a predictive model,

standardizing the patient data, and calculating an observed rate. The predictive model could

be obtained from open-sourced or commercial sources. Standardizing data means cleaning

the data so as to conform to specific requirements. By defining the numerators and the

denominators related to inpatient deaths and the number of patients in a specific population,

the observed rate can be calculated.