Validation of the de Morton mobility index (DEMMI) among acute stroke patients aged 65 years: a register study

Purpose: To validate the de Morton Mobility Index (DEMMI) in older (>= 65 years) patients with acute stroke in a hospital setting within the first week after stroke onset. Materials and methods: In the Danish National Database of Geriatrics, we identified 4,176 patients with acute stroke (>= 65 years). Floor and ceiling effects of DEMMI were investigated. Furthermore, convergent validity was investigated by correlations between DEMMI and the Barthel Index using Spearman's rho. Known-groups validity was tested by comparing DEMMI scores for different groups (with/without dementia, depression, comorbidity, and walking aids), and unidimensionality of DEMMI was evaluated by Mokken scale analysis. Results: A floor effect was identified with 22.1% of the patients scoring 0 on DEMMI on admission. Both convergent and known-groups validity were confirmed for DEMMI. Patients who were bedbound had a lower DEMMI score (median [IQR]: 0 [0;0]) than patients without any walking aid (median [IQR]: 62 [33;74]). Furthermore, Mokken scale analysis identified unidimensionality with overall fit to the model (Loevinger H 0.88 (p < 0.0001)). Conclusion: DEMMI is a valid instrument for use in patients with acute stroke (>= 65 years) in a hospital setting within the first week after stroke onset.