Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


College of Health Professions

First Advisor

Kit N. Simpson

Second Advisor

Annie N. Simpson

Third Advisor

Paul J. Nietert

Fourth Advisor

J. Dunc Williams

Fifth Advisor

Zoher F. Kapasi


Cost and cost savings have become an important focus for health policy administrators. However, there are missing pieces in our approach to cost analysis; there is no consensus on multivariable methods, no indicators of minimally acceptable values, and no specification of process costing. In this dissertation, I propose to fill the gaps in the literature by 1) identifying which methods are appropriate for large claims data, 2) examine existing methods to establish minimally important difference (MID) in health outcomes to identify MID in costs, and 3) determine differences in sick visit clinic costs using a modified micro-costing method. Most models that were compared to the generalized linear models Gamma distribution with log link found it to be the superior model in both simulated data and real administrative data. We recommend that in cases where acceptable anchors are not available to establish an MID, both the Delphi and the distribution-method of MID for costs be explored for convergence. Our micro-costing approach is feasible to use under virtual working conditions; requires minimal provider time; and generates detailed cost estimates that have “face validity” with providers and are relevant for economic evaluation.


All rights reserved. Copyright is held by the author.