Traditional and Rasch psychometric analyses of the Quality of Life in Adult Cancer Survivors (QLACS) questionnaire in shorter-term cancer survivors 15 months post-diagnosis

Objective: The aim of this paper is to provide new insights into the psychometrics of the Quality of Life in Adult Cancer Survivors (QLACS) questionnaire, originally developed for longer-term survivors 5 + years post-diagnosis. Specifically, to examine the classic psychometric properties of QLACS in a sample of shorter-term survivors, and to undertake Rasch analysis to explore the extent to which the Generic and Cancer-Specific summary scales (and separately-analysed Benefits of cancer domain) are unidimensional, with linear measurement properties and no differential item functioning (DIF). Methods: Patients with potentially curable breast, colorectal or prostate cancer completed QLACS 15 months post-diagnosis (N = 407). Score distributions, floor and ceiling effects, internal reliability, and feasibility (completion time and missing data) were examined. Rasch analysis included examination of item fit, DIF and unidimensionality. Results: The QLACS domains and summary scales had very similar score distributions and classic psychometric properties (no ceiling effects, majority no floor effects, acceptable reliability) to those found in development work with longer-term survivors. Median completion time was 10 min and total missing data 23%. The Generic summary scale contained several misfitting items and exhibited multidimensionality. The Cancer-Specific summary scale and Benefits domain showed fit to the Rasch model and demonstrated unidimensionality and no DIF, with just one or no item modifications respectively. Conclusion: QLACS demonstrates similarly good classic psychometric properties among shorter-term as among longer-term survivors, and has good feasibility. The Cancer-Specific summary scale and Benefits domain showed an impressive degree of fit to the Rasch model, although the validity of computing the Generic summary score was not supported. 2014 Published by Elsevier Inc.