Scale Shortening and Decrease in Measurement Precision: Analysis of the Pain Self-Efficacy Questionnaire and Its Short Forms in an Italian-Speaking Population With Neck Pain Disorders

Abstract

Abstract Objective Short (2- and 4-item) forms of the Pain Self-Efficacy Questionnaire (PSEQ) have been proposed, but their measurement precision at the individual level is unclear. The purpose of this study was to analyze the Rasch psychometric characteristics of PSEQ and its 3 short forms (one 4-item and two 2-item versions) in an Italian-speaking population with neck pain (NP) disorders and compare their measurement precision at the individual level through calculation of the test information function (TIF). Methods Secondary analysis of data from a prospective single-group observational study was conducted. In 161 consecutive participants (mean age = 45 years [SD = 14]; 104 women) with NP disorders, a Rasch analysis was performed on each version of the PSEQ (full scale plus 3 short forms), and the TIF was calculated to examine the degree of measurement precision in estimating person ability over the whole measured construct (pain self-efficacy). Results In all versions of the PSEQ, the rating scale fulfilled the category functioning criteria, and all items showed an adequate fit to the Rasch model. The TIF showed a bell-shaped distribution of information, with an acceptable measurement precision (standard error <0.5) for persons with a wide range of ability; conversely, measurement precision was unacceptably low in each short form (particularly the two 2-item versions). Conclusions The results confirm and expand reports on the sound psychometric characteristics of PSEQ, showing for the first time, to our knowledge, its conditional precision in estimating pain self-efficacy measures in Italian individuals with NP disorders. The study cautions against use of the 3 PSEQ short forms for individual-level clinical decision-making. Impact Short scales are popular in rehabilitation settings largely because they can save assessment time and related costs. The psychometric characteristics of the 10-item PSEQ were confirmed and deepened, including its precision in estimating individual pain self-efficacy at different levels of this latent variable. On the other hand, low measurement precision of the 3 PSEQ short forms cautions against their use for individual judgments.