Reducing ceiling effects in the Working Alliance Inventory- Rehabilitation Dutch Version

Abstract

ABSTRACT Introduction: The Working Alliance Inventory Rehabilitation Dutch Version (WAI-ReD) is a valid instrument for measuring therapeutic alliance within rehabilitation. A limitation of the WAI-ReD is its ceiling effects for all domains, in more than 15% of the patients. These effects affect the responsiveness of an instrument. Objective: Determine if modifying response scales, by changing labels and utilizing visual analogue scales of the WAI-ReD will reduce ceiling effects. Methods: The primary outcomes were the percentage of ceiling effects and the median scores on the total- and domain scores of the WAI-ReD and the two modified versions. As secondary outcomes, the construct validity and internal consistency of the modified versions were explored. The items of the original WAI-ReD are graded on a balanced Likert scale, the modified versions have a; (1) Likert scale with Positive-Packed labels (WAI-ReDPP); or (2) Visual Analogue Scale (WAI-ReDVAS). Adult patients from the Department of Rehabilitation of the University Medical Center Groningen and two physiotherapy practices were recruited. This study had a cross-sectional design. Participants randomly received one of the versions of the WAI-ReD. Results: No ceiling effects were present in the total- or domain scores of the modified WAI-ReDVAS. Significantly (p≤ 0.05) fewer ceiling effects were found in the domains “Bond” and “Goal” of the WAI-ReDVAS compared to the WAI-ReD and WAI-RedPP. Ceiling effects were still present on the domains “Bond” and “Goal” of the WAI-ReDPP. Median outcomes of the modified versions were higher than the median outcome of the original version, but not significantly. The correlations (Pearson’s r) between the WAI-ReDVAS and similar validated instruments were high (r ≥ 0.70) . The internal consistency of the total- and domain scores of both modified versions were high (0.70 ≤ r < 0.95). Conclusion: Utilizing visual analogue scales is an effective method to prevent ceiling effects on the total score and individual domains of the WAI-ReD. The clinimetric properties of the WAI-ReDVAS are superior compared to the WAI-ReD and WAI-ReDPP. The WAI-ReDVAS is therefore the recommended scale for measuring therapeutic alliances within rehabilitation. Keywords: Rehabilitation, therapeutic alliance, ceiling effect.