The reliability and agreement of scores in a novel balance measure for older adults: Specific Training According to BaLance Evaluation (STABLE)

Abstract

Purpose

Falls are the leading cause of injury and premature death among community dwelling elderly but can be prevented through comprehensive balance rehabilitation which ideally targets the patient’s specific needs. In this study, we evaluate the reliability and agreement of six novel clinical measures of different balance domains which applied in a patient-specific balance profile guides exercise prescription in balance rehabilitation.

Materials and methods

The intra-rater reliability and agreement of the six measures were evaluated on the same day in six different cohorts of elderly with balance disability (n = 65–100). Further, the inter-day intra-rater and inter-rater and test–retest reliability and agreement of the measures and the balance profile were evaluated (n = 100).

Results

The intra-day intra-rater reliability and agreement was moderate to excellent (ICC_2.1 = 0.525–0.968, with SDC% = 6.5–284.9%) but poor to good for the inter-day conditions (ICC_2.1 = 0.123–0.832, with SDC% = 6.6–229.2%). The reliability of classifying the lower domain in the balance profile was fair, with kappa = 0.56 (95%CI 0.36–0.76).

Conclusions

Five of the six measures may reliably be applied to measure balance disability and to guide rehabilitation.

Implications for rehabilitation

• Balance disability is the primary reason for accidental falls among elderly but can be prevented through comprehensive individualized balance rehabilitation.

• Specific Training According to BaLance Evaluation (STABLE) is a novel approach for designing effective balance exercises based on clinical measurements.

• Five of the six measures are reliable when applied in a patient-specific balance profile to guide rehabilitation following the STABLE approach.

Keywords:

• Fall prevention
• balance assessment
• elderly adults
• rehabilitation exercises
• reliability analysis
• rater agreement

Acknowledgements

We wish to thank the physiotherapists Andreas Thenning Andersen, Lise-Lotte Sangild, Marte Bjerketvedt, Rasmus Jørgensen, Rikke Steen Krawcyk, Sanni Topholm Janum, and Vibeke Svendsen for contributing to the data acquisition.

Disclosure statement

The authors report there are no competing interests to declare.

Additional information

Funding

This work was supported by the grant from the Danish Physical Therapists Research Grant and from Johannes Fogs Fond.