An item bank was created to improve the measurement of cancer-related fatigue

Lai, Jin-shei; Cella, David; Dineen, Kelly; Bode, Rita; Von Roenn, Jamie; Gershon, Richard C.; Shevrin, Daniel

doi:10.1016/j.jclinepi.2003.07.016

« Previous Next »Journal of Clinical Epidemiology
Volume 58, Issue 2 , Pages 190-197, February 2005

An item bank was created to improve the measurement of cancer-related fatigue

Jin-shei Lai
- Evanston Northwestern Healthcare, 1001 University Place, #100, Evanston, IL 60201, USA
- Northwestern University, Chicago, IL, USA
- Corresponding author. Tel.: 224-364-7312; fax: 847-570-8033.
David Cella
- Evanston Northwestern Healthcare, 1001 University Place, #100, Evanston, IL 60201, USA
- Northwestern University, Chicago, IL, USA
Kelly Dineen
- Private practice
Rita Bode
- Northwestern University, Chicago, IL, USA
- Rehabilitation Institute of Chicago, Chicago, IL, USA
Jamie Von Roenn
- Northwestern University, Chicago, IL, USA
Richard C. Gershon
- Evanston Northwestern Healthcare, 1001 University Place, #100, Evanston, IL 60201, USA
- Northwestern University, Chicago, IL, USA
Daniel Shevrin
- Evanston Northwestern Healthcare, 1001 University Place, #100, Evanston, IL 60201, USA

Accepted 3 July 2003.

Abstract

Objective

Cancer-related fatigue (CRF) is one of the most common unrelieved symptoms experienced by patients. CRF is underrecognized and undertreated due to a lack of clinically sensitive instruments that integrate easily into clinics. Modern computerized adaptive testing (CAT) can overcome these obstacles by enabling precise assessment of fatigue without requiring the administration of a large number of questions. A working item bank is essential for development of a CAT platform. The present report describes the building of an operational item bank for use in clinical settings with the ultimate goal of improving CRF identification and treatment.

Study design and setting

The sample included 301 cancer patients. Psychometric properties of items were examined by using Rasch analysis, an Item Response Theory (IRT) model.

Results and conclusion

The final bank includes 72 items. These 72 unidimensional items explained 57.5% of the variance, based on factor analysis results. Excellent internal consistency (α=0.99) and acceptable item–total correlation were found (range: 0.51–0.85). The 72 items covered a reasonable range of the fatigue continuum. No significant ceiling effects, floor effects, or gaps were found. A sample short form was created for demonstration purposes. The resulting bank is amenable to the development of a CAT platform.