Choosing Models for Health Care Cost Analyses: Issues of Nonlinearity and Endogeneity
Article first published online: 23 APR 2012
© Health Research and Educational Trust
Health Services Research
Volume 47, Issue 6, pages 2377–2397, December 2012
How to Cite
Garrido, M. M., Deb, P., Burgess, J. F. and Penrod, J. D. (2012), Choosing Models for Health Care Cost Analyses: Issues of Nonlinearity and Endogeneity. Health Services Research, 47: 2377–2397. doi: 10.1111/j.1475-6773.2012.01414.x
- Issue published online: 12 NOV 2012
- Article first published online: 23 APR 2012
- Health Services Research and Development Service. Grant Numbers: IAD-06-060-2, REA 08-260
- nonlinear models;
- treatment effects;
- palliative care
To compare methods of analyzing endogenous treatment effect models for nonlinear outcomes and illustrate the impact of model specification on estimates of treatment effects such as health care costs.
Secondary data on cost and utilization for inpatients hospitalized in five Veterans Affairs acute care facilities in 2005–2006.
We compare results from analyses with full information maximum simulated likelihood (FIMSL); control function (CF) approaches employing different types and functional forms for the residuals, including the special case of two-stage residual inclusion; and two-stage least squares (2SLS). As an example, we examine the effect of an inpatient palliative care (PC) consultation on direct costs of care per day.
Data Collection/Extraction Methods
We analyzed data for 3,389 inpatients with one or more life-limiting diseases.
The distribution of average treatment effects on the treated and local average treatment effects of a PC consultation depended on model specification. CF and FIMSL estimates were more similar to each other than to 2SLS estimates. CF estimates were sensitive to choice and functional form of residual.
When modeling cost or other nonlinear data with endogeneity, one should be aware of the impact of model specification and treatment effect choice on results.