Rasch validation and comparison of the German versions of the Locomotor Capabilities Index-5 and Prosthetic Mobility Questionnaire 2.0 in lower-limb prosthesis users

Abstract

This study aimed to compare, through Rasch analysis, the psychometric properties of the Locomotor Capabilities Index (LCI-5) and Prosthetic Mobility Questionnaire (PMQ 2.0) in German lower-limb prosthesis users. The questionnaires were concurrently administered to a convenience sample of 98 consecutively recruited individuals with lower limb amputation (LLA) (male/female = 61/37; mean age 57 ± 14 years). LCI-5 showed disordered rating scale thresholds (one response option in three items required collapsing); local dependence between two items (resolved by creating a testlet); underfit of one item (‘Get up from the floor’); and presence of a second weak dimension. PMQ 2.0 showed a correctly functioning rating scale; good fit of the data to the model (apart from some overfit); local dependence between two items (absorbed by creating a testlet); and essential unidimensionality. At scale co-calibration onto a common interval-scaled metric, PMQ 2.0 was better targeted than LCI-5 (i.e. the extent of item difficulty was more appropriate for the sample) and its operational range allowed a more precise measurement of higher locomotor abilities. The correlation between LCI-5 and PMQ 2.0 scores was rho =  0.78. In conclusion, LCI-5 revealed some drawbacks, confirming a previous Rasch study; refinement of its rating scale and item selection seems therefore warranted. The PMQ 2.0 demonstrated good overall measurement quality, in line with previous Italian and Slovene studies. The operational range of the PMQ 2.0 makes it more suitable than LCI-5 for assessing people with high locomotor abilities.