Inter-Rater Reliability of the Balance-Based Torso-Weighting Method of Altering Balance and Gait

Background: Multiple sclerosis (MS) typically results in balance and gait deficits and increased risk of falls. One promising intervention that has resulted in balance and gait improvements in MS utilizes a Balance-Based Torso-Weighting (BBTW) method. With BBTW, small weights (less than 2% body weight) applied in different places on a vest-like garment counter the direction of balance loss when a clinician manually perturbs standing balance in various directions. Perturbations include nudges and rotational forces in six prescribed directions at the shoulders and pelvis. No reliability studies have yet been published to examine the inter-rater reliability of assessing direction of greatest balance loss, an essential component for placing the small weights.

Objectives: Examine the inter-rater reliability of the BBTW method with a focus on assessing direction of balance loss.

Methods: Four physical therapists with various amounts of BBTW training and clinical experience assessed direction of balance loss in 10 healthy volunteers to determine where to apply weights. Subjects were assessed by at least two therapists. Therapists assessed at least six subjects. Each subject had 30 minutes between sessions; all weighting sessions were videotaped and completed within a four-hour period. Six trained reviewers then viewed videos and scored balance loss following perturbations. At least four reviewers examined videos for each therapist. Score of balance loss was rated using a rubric where 0 indicated normal, quick balance response; 1-3 indicated progressively greater body movement and slower (but independent) recovery of upright orientation; 4 indicated that the subject would have fallen without clinician assist following a perturbation. Intra-class Correlation Coefficients (ICC) and measures of agreement were assessed across reviewers for each therapist.

Results: ICCs ranged from 0.57 to 0.78 within the four therapists. Agreement within one point for each perturbation scored on the rubric across five to six reviewers for each therapist ranged from 0.80 to 0.93. The highest ICCs and agreement values were consistently aligned with the most experienced therapist.

Conclusions: Assessing direction of balance loss has moderate to good inter-rater reliability and good to excellent agreement. These results support BBTW as a promising intervention. Clinicians and reviewers can be trained to provide reliable results although intermittent review of the scoring rubric after practice assessing balance loss is recommended.