Background
Relational decision-making and problem-solving are powerful activities occurring daily across all healthcare contexts. Their empowering potential can be difficult to judge and is seldom fully exploited, and they can even be perceived as disempowering. This calls for assessment of the extent to which healthcare users perceive their interactions with healthcare providers as empowering. We developed a 36-item draft version of the generic EMPOWER-UP questionnaire deemed to have good face and content validity by healthcare users and researchers with expertise across multiple healthcare contexts. This article reports the psychometric properties of the Danish and English versions of the questionnaire.
Methods
A cross-sectional study was carried out among 526 adults diagnosed with diabetes, cancer, or severe mental illness in Denmark, United Kingdom, and Australia with recent (≤ 6 months) interaction with a healthcare provider. We evaluated structural validity, cross-cultural validity, and measurement invariance using confirmatory and exploratory factor analyses, item response theory modelling, and differential item functioning. Additionally, we evaluated empirical reliability and internal consistency.
Results
The final 24-item Danish and English versions of EMPOWER-UP has a three-factor second-order structure. The three factors of empowering mechanisms (14 items), disempowering mechanisms (6 items), and empowering insight (4 items) are connected through the second-order factor, realizing empowerment. Partial credit model calibration supported the use of summed subscale item scores, enhancing usability in clinical practice settings. Differential item functioning analysis demonstrated invariance across subgroups. Empirical reliability of the subscales was estimated at 0.85–0.95, and Cronbach’s alpha was 0.90–0.96.
Conclusions
The 24-item EMPOWER-UP questionnaire measuring empowerment in relational decision-making and problem-solving showed strong psychometric properties among adults with cancer, diabetes, and severe mental illness. Further evaluation in other contexts and populations is needed, including evaluation of construct validity, responsiveness to change, and test-retest reliability.