| Session Title: Methodological Innovations: When There are Holes in Your Data |
| Multipaper Session 209 to be held in Panzacola Section H2 on Thursday, Nov 12, 9:15 AM to 10:45 AM |
| Sponsored by the Quantitative Methods: Theory and Design TIG |
| Chair(s): |
| Melinda Davis, University of Arizona, mfd@email.arizona.edu |
| Abstract: The proposed session will develop several new approaches in measurement and analysis. Each presentation has both methodological and analytical aspects that are expected to be of interest practicing evaluators and methodologists alike. The first paper is by Mende Davis and Katalin Scherer, who will discuss the development and uses of a disease severity measure from billing and claims records. The disease severity measure was developed using Rasch modeling, and can be used to track disease progression over time. In the second paper, Michael Menke discusses a new approach to meta-analysis using Bayesian decision analysis that is designed to take into account missing data into account. The approach is comparable to classic deductive reasoning riddles. |
| Estimating Disease Severity in Muscular Dystrophy |
| Melinda Davis, University of Arizona, mfd@email.arizona.edu |
| Katalin Scherer, University of Arizona, kscherer@email.arizona.edu |
| A variety of scales and measures have been used to measure disease severity for Duchenne muscular dystrophy. However, the majority of these scales require direct clinical assessment. We will present a method to estimate disease severity using existing data. We used six years of administrative (billing and claims) data for children with muscular dystrophy. Rasch modeling was used to estimate the severity of people and items at the same time. The pilot muscular dystrophy scale had a reliability of .80, and the majority of the billing and claims items fit the measure. Billing and claims measures may be used to identify individuals whose chronic condition is not under control, to identify individuals who may be at risk for severe medical events, and to predict future need for care. |
| What Do You Do When You Don't Have all Your Marbles? |
| J Michael Menke, University of Arizona, menke@email.arizona.edu |
| What do you do when you don't have all your marbles? When data are synthesized to help form a health policy decision, there are often many 'holes'. These holes need not be imputed. Instead studies and meta-analyses can be combined to compare treatments where they have not been compared before, using decision analysis. |