Hospital mobility scale: Measurement properties in neurocritical patients in the intensive care unit (ICU)

Objective: To evaluate the measurement properties of the hospital mobility scale (HMS) in neurological patients admitted to the intensive care unit. Design: Measurement properties study. Setting: Neurological intensive care unit. Participants: About 101 neurocritical patients. Intervention: The HMS and the ICU mobility scale were applied at intensive care unit admission. The HMS was reapplied 4 hours later and again at discharge. Main Measures: Inter- and intra-rater reliability were analysed using the intraclass correlation coefficient; measurement error using the standard error of measurement; internal consistency using Cronbach's alpha; construct validity through correlation with the ICU mobility scale total score; floor and ceiling effects using frequency analyses; and responsiveness through effect size (internal), receiver operating characteristic analysis, and correlation with change in ICU mobility scale (external). Results: The HMS demonstrated excellent inter-rater and intra-rater reliability (intraclass correlation coefficient = 0.966 [95%CI 0.946-0.978]; p < 0.001 and 0.955 [95%CI 0.928-0.970]; p < 0.001, respectively), with very good measurement error (standard error of measurement = 5%). Construct validity showed a moderate negative correlation with the ICU mobility scale (r = -0.72; p < 0.001). Internal responsiveness was moderate (effect size = 0.59). External responsiveness showed a moderate correlation with change in ICU mobility scale (r = -0.52; p < 0.001) and a small area under the receiver operating characteristic curve (0.67; 95%CI 0.53-0.81). No floor or ceiling effects were observed. Conclusions: The HMS appears to be a valid and reliable tool for assessing functional mobility in neurological intensive care unit patients; however, its ability to detect short-term changes in mobility over time is limited.