Mobility Measures in Older Cancer Survivors: An Examination of Reliability, Validity, and Minimal Detectable Change

Abstract

Background: Cancer and its treatment contribute to multiple sequelae affecting physical function, including difficulty in balance and walking. Although clinical tests should be used to assess fall risks in older cancer survivors, the literature describing the validity, reliability, measurement error, and minimal detectable change (MDC) of mobility measures is lacking. Objective: The purpose of this study was to describe the reliability, validity, standard error of the measurement (SEM), and minimal detectable change at 95% confidence (MDC95) of mobility measures in older cancer survivors. Methods: Fifty cancer survivors (breast, prostate, lung, or colorectal) aged 65+ years participated. Measures of gait speed (GS) (usual, fast, and dual-task) and mobility (Timed Up and Go [TUG], TUG-cognitive, and TUG-manual) were performed once and then repeated 2 weeks later. Test-retest reliability was assessed using intraclass correlation coefficient (ICC2,1). MDC95 and SEM were calculated. Construct validity was examined using Pearson's correlation coefficient with 6-item Activities-specific Balance Confidence scale and Short Physical Performance Battery performance. Results: Test-retest reliability was excellent for all mobility measures, with ICC values ranging from 0.89 to 0.94 for GS and 0.95 to 0.98 for TUG measures. Construct validity was present for TUG (r = −0.53 to −0.73) and GS (r = 0.29 to 0.5) measures. Low SEM values demonstrated excellent reproducibility. To be considered real change, usual GS should be 0.14 m/s or more and TUG 2.49 seconds or more. Conclusions: Measures of GS and TUG have excellent reliability and construct validity and should be considered for use in older cancer survivors. Further study is indicated to establish the psychometric values of mobility measures by cancer type.