Understanding the CaseVue Score
Since its beginning, CaseVue has been carefully designed to render the complex simple. The complex formulas that drive the scoring engines in CaseVue, are designed for ease of use and interpretation by all kinds of users. Caseworkers, case managers, business office mangers, and executives alike have access to a real time score delivered to their desktop computers.
The score is the result of a multi-layered algorithm or formula configured to suit the business process needs of our customers. The predictive analytics driving the scoring process in our base engines, is derived from millions of previous cases and scores that predict the likelihood and level of risk in a particular situation. In the case of eligibility for government programs, the score, from 1 to 100, ranks the likelihood that a case is eligible to receive assistance and assist users with their determination decisions.What data is in the score?
Each CaseVue scoring engine is different. Government assistance programs (Medicaid, LTC, Food Stamps, SCHIP, TANF) each have unique data sets, that reflect the separate rules and distinct business processes used in determining eligibility. AutoGov owns and maintains a “base engine” for each of these programs. Common factors do exist among the various scoring engines, but there are factors that are unique to a specific program. Generally, the type of information common across programs relates to identity, residency, citizenship, income and assets.
What does the algorithm look like?
A typical CaseVue scoring engine consists of more than 400 lines of code. The more complex the program, the more lines of code that make up the scoring engine. Each case scored passes (with its data) through the scoring engine algorithm. A typical case is scored in less than a second. The CaseVue scoring engines and their exact contents are the intellectual property of AutoGov and are considered trade secrets.
Where does the data come from?
CaseVue fuses data from a variety of data sources to provide real time scores to users. Basic case data often originates from the case management system of record. AutoGov often supplements this information by fusing reputable third party data available from its strategic business partners. In addition, AutoGov holds Agreements with specialty data providers for use in unique circumstances.
What is the distinction between rule adherence and risk identification?
Almost all eligibility systems have “rules engines” that help facilitate a case as it moves through the process of determining eligibility. This fact is especially true with state based eligibility systems. Program rules are very specific and have certain thresholds – rules - designed to support a business process that results in a YES or NO decision. While CaseVue’s scoring engines consider the rules as a critical component of the engines, they are not necessarily part of a risk score in and of themselves. The predictive factors weigh risk to the group or person required to make a decision. For example, a state may have a rule that limits total assets to $2,000 in long term care eligibility. This same case may be risky with only $1,700.
How does a risk score affect timeliness and efficiency?
Identifying risk at any point in time during the life of a case or application as it is being work is critical to addressing timeliness and accuracy issues. Eligibiity determination for example is a dynamic process. Cases are in motion. They change and become more or less risky based on the substance of the case, the amount of information in the case and the amount of time that the case has been in process before a final decision. Generally as cases age their risk level increases, particularly if critical pieces of information remain “missing. CaseVue’s scoring engines are designed to encourage usage that ensures that the level of risk is measured as a case is in motion.