Rasch Analysis, Reliability, and Validity of the Turkish Version of the Lumbar Spine Instability Questionnaire

Abstract

Background/aim: Lumbar instability is an important condition that can be seen frequently in people with low back pain, affecting both the progression and the choice of appropriate exercise. The Lumbar Spine Instability Questionnaire (LSIQ) is a simple and low-cost tool for evaluating disturbed back stability in people with low back pain. The aim of this study is to develop the Turkish version of the LSIQ (LSIQ-T) and to evaluate its psychometric properties using the Rasch model.Materials and methods: One hundred participants with chronic low back pain completed the LSIQ-T. The LSIQ-T was repeated for 30 participants after 1 week to establish its test-retest reliability. While internal and external construct validity were investigated using Rasch analysis and the Spearman correlation coefficient, respectively, reliability was evaluated in terms of internal consistency by Cron-bach's alpha and the Person Separation Index (PSI).Results: All items of the LSIQ-T were found to fit the Rasch model (chi-square: 34.07 (df = 15), p = 0.0033). The internal construct validity was good, the overall mean item fit residual was 0 (SD: 0.765), and the mean person fit residual was 0.322 (SD: 1.123). Internal consistency reliability was low with a PSI of 0.63 although Cronbach's alpha was acceptable (0.68). When the test-retest reliability was examined via differential item functioning (DIF) by time, none of the items showed DIF.Conclusion: The LSIQ-T is a valid unidimensional scale for the Turkish population. Although the LSIQ-T had low internal consistency, it demonstrated unidimensionality and is appropriate for use. Therefore, the LSIQ-T can be used in clinical practice and scientific re-search.