Interadministrator Reliability of a Modified Instrumented Push and Release Test of Reactive Balance

Context: Traditional assessments of reactive balance require sophisticated instrumentation to ensure objective, highly repeatable paradigms. This instrumentation is clinically impractical. The Push and Release test (P&R) is a well-validated clinical test that examines reactive balance, and the application of wearable inertial measurement units (IMU) enables sensitive and objective assessment of this clinically feasible test. The P&R relies on administrator experience and may be susceptible to interadministration reliability concerns. The purpose of this study was to evaluate the interadministrator reliability of objective outcomes from an instrumented, modified version of the P&R test. Design: Crossover interadministrator design. Methods: Twenty healthy adults (20-35 y) completed the P&R in 4 directions with 2 different administrators. Measures quantified using IMUs included step latency, step length, and time to stability. Lean angle (LA) at release was used as a measure of administration consistency. The intraclass correlation coefficient (ICC) estimate was used to assess interadministrator reliability in each direction. To determine consistency of LA within and across administrators, we calculated the SDs for each rater by direction and the interadministrator reliability of LA using ICC. Results: Across individual directions, the ICC for agreement between raters ranged from .16 to .39 for step latency, from .52 to .62 for time to stability, and from .48 to .84 for step length. Summary metrics across all 4 directions produced higher ICC values. There was poor to moderate consistency in administration based on LA, but LA did not significantly affect any of the outcomes. Conclusion: The modified P&R yields moderate interadministrator reliability and high validity. Summary metrics over all 4 directions (the maximum step latency, the median time to stability, and the median step length) are likely more reliable than direction-specific scores. Variations in body size should also be considered when comparing populations.