1. Seizure. 2012 Jul;21(6):437-43. doi: 10.1016/j.seizure.2012.04.009. Epub 2012 May

Using Rasch measurement theory to assess three depression scales among adults
with epilepsy.

Walker ER(1), Engelhard G Jr, Thompson NJ.

Author information: 
(1)Rollins School of Public Health, Emory University, 1518 Clifton Rd., Atlanta, 
GA 30322, USA. ereisin@emory.edu

People with chronic conditions, such as epilepsy, are at a high risk for
depression; however depression is often under-recognized and undertreated.
Depression scales, including one specific to people with epilepsy, have been used
for screening in this population, although none have been assessed with Rasch
measurement theory. This study used Rasch analyses in order to evaluate and
compare the psychometric properties of the modified Beck Depression Inventory,
the Patient Health Questionnaire, and the Neurological Disorders Depression
Inventory for Epilepsy in a sample of people with epilepsy and depression who
participated in an intervention designed to reduce depressive symptoms. A
secondary purpose was to assess item functioning across time. The sample includes
44 participants in the Project UPLIFT program who completed the assessments
before and after taking part in the intervention. Results of the Rasch analysis
indicate that the three depression scales functioned as intended. There was good 
overall targeting between the items and the sample, acceptable model-data fit,
and good reliability of separation for persons, items, and time. The participants
experienced a significant decrease in depressive symptoms from pretest to
posttest. This study illustrates the value of using model-based measurement with 
the Rasch model to combine items across the three depression scales. It also
demonstrates an approach for analyzing and evaluating the results of small scale 
intervention programs, such as the UPLIFT program.

Copyright © 2012 British Epilepsy Association. Published by Elsevier Ltd. All
rights reserved.

DOI: 10.1016/j.seizure.2012.04.009 
PMID: 22608975  [Indexed for MEDLINE]