Validation of the de Morton Mobility Index for measuring mobility related activities in Hungarian institutionalized...

In old age, decline in mobility has a significant impact on performing activities of daily living, participating in society, and on quality of life. Mobility changes should be measured and monitored using measures with appropriate psychometric properties. The psychometrically sound de Morton Mobility Index (DEMMI) assesses a broad range of mobility related activities. The aim of this study was to develop a valid and reliable, cross-culturally adapted Hungarian version of the DEMMI scale (HU-DEMMI) among institutionalized older people. Development of the HU-DEMMI was conducted using a cross-sectional study design with an overall of 170 participants. Validity was tested via construct (established by hypothesis testing) and structural validity (explored with structural equation modeling (SEM)). Reliability was assessed using three measurement properties: internal consistency (determined by Cronbach's alpha), test-retest reliability (by intraclass correlation coefficients (ICC)), and measurement error (standard error of measurement (SEM), minimally detectable change (MDC), limits of agreement (LoA)). Linguistic and cultural adaptation was realized successfully. As for construct validity, the HU-DEMMI showed a strong correlation with mobility related measures and a weak correlation with non-mobility measure. Significant differences in the DEMMI score were found across groups being at various levels of assistive device use. Internal consistency by Cronbach's alpha was 0.906; the dimensionality testing revealed no violation of unidimensionality. However, a three-dimensional structure uncovered a better fit. The inter- and intra-rater ICC of DEMMI scores were 0.981 and 0.989, respectively. The SEM for inter-rater reliability test was 2.924 and the MDC₉₀ was 6.803 points on the 100-point DEMMI scale. The 95% LoA showed a mean inter-rater difference of 0.2115 (- 11.252; 11.675), while a mean intra-rater difference of -1.0577 (- 9.809; 7.694). Four data points (7.69%) were identified as outliers in both cases. Maintaining institutionalized older people's functional capacity and delaying its decline requires effective interventions. Valid and reliable determination of mobility levels is necessary to address mobility decline in a personalized way. The psychometrically robust HU-DEMMI scale developed in this study would provide this and allow for future comparisons of Hungarian research results with those of other countries and for contribution to international, multicenter research.