REVIEWER: Allen Naidoo, Ph.D., director of Medical Informatics for BlueCross BlueShield of Tennessee (BCBST).

BACKGROUND: BCBST works diligently to provide members with affordable health care solutions and put better health at the center of all they do. It is the state’s oldest and largest not-for-profit health plan, serving nearly 3 million Tennesseans.

PLATFORMS: We’re running SAS predictive analytics on an AIX 5.2 operating system that sits on P570 IBM hardware.

PROBLEM SOLVED: We have access to health care claims data that dates back to 1996. SAS predictive analytics help with this large amount of data. We use SAS Enterprise Miner to answer business questions such as how to keep costs down while increasing revenue. One goal is to optimize revenue via diagnostic discovery by increasing the accuracy of diagnostic review. SAS Forecast Server helps us evaluate utilization management programs so we can better understand and respond to medical service program trends. It has also helped us evaluate cost avoidance.

PRODUCT FUNCTIONALITY: Efficiency is important to us. SAS Enterprise Miner streamlines the data mining process to create highly accurate predictive and descriptive models based on analysis of vast amounts of BCBST data. The unsurpassed scalability of SAS Forecast Server enables Medical Informatics to operate more efficiently by quickly and automatically producing reliable, statistically based forecasts.

STRENGTHS: The ease of using the graphical user interface (GUI) workspace in SAS Enterprise Miner gives us flexibility to revise the order in which we process and manipulate input data. We especially appreciate it when adjusting a model or testing a new approach by inserting new nodes. SAS Enterprise Miner provides efficient model comparison with a wide range of sophisticated modeling techniques such as regression, neural networks and decision trees. We find the organization of advanced analytic algorithms laid out as core tasks to be performed in data mining quite helpful: sampling, exploration, modification, modeling and assessment. In each step, we’re guided through an array of actions to perform. The rich Java client interface enables fast, maintenance-free distribution. SAS Forecast Server improves forecasting performance across all products. It automatically determines the most appropriate model from a broad and extensible model selection list, optimizing model parameters without requiring an analyst to engage in a time-consuming manual process. It also allows for automatic detection of seasonality and the creation of seasonally adjusted projected time series. It performs automatic hierarchical forecasting, allowing us to define the hierarchy, parameters and business rules through an interactive GUI. SAS Forecast Server reconciles up and down the hierarchy, preserving locked forecast values. It generates exception reports (outliers) based on sound statistical logic and business rules. It supports forecasting as an ongoing and repeated process that fits into our organization’s evaluation workflow related to care/disease/utilization management. It adds validity to ROI measurement and reporting by allowing for confidence intervals on the projected trend, ROI and overall utilization management program savings estimates.

WEAKNESSES: Our data transformation needs are large, and most of our data prep is done in SAS Enterprise Guide rather than SAS Enterprise Miner and SAS Forecast Server. The newest version of SAS Enterprise Miner addresses this with a new data input wizard and other features.

SELECTION CRITERIA: SAS Enterprise Miner has the most comprehensive set of models and the best methods for deploying score code available. It allowed for more customization than other products. SAS Forecast Server offers automatic hierarchical forecasting at a scale no other product delivers. With it, we are finding innovative ways to address care/disease/utilization management problems.

DELIVERABLES: SAS Enterprise Miner enables us to do accreditation analytics - predicting compliance and noncompliance using evidence-based guidelines to create lists for targeted outreach. SAS Enterprise Miner also helps us improve our diagnostic discovery efforts by identifying members with undocumented diagnoses. SAS Forecast Server gives us valid forecasts of utilization trends, enabling us to quantify the ROI of care/disease/utilization management intervention programs.

VENDOR SUPPORT: SAS support is superior at all levels. They work to understand our needs. We receive regular offers to help us further take advantage of the tools’ capabilities.

DOCUMENTATION: Documentation is complete and easy to understand.

SAS Enterprise Miner and SAS Forecast Server
SAS
SAS Campus Drive
Cary, NC 27513
(919) 677-8000
http://www.sas.com/

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access