# Teleassessments of Lower Limb Function in Adult Stroke Survivors: A Preliminary Study Evaluating Safety, Feasibility, and Validity for Telerehabilitation

## TL;DR

This study shows that certain tests for lower limb function can be safely and effectively done remotely for stroke survivors, supporting their use in telerehabilitation.

## Contribution

The study introduces evidence that specific teleassessments are valid and feasible for evaluating lower limb function in stroke survivors.

## Key findings

- The 2-minute step test (TMST) showed excellent validity (ICC = 0.971) and higher sensitivity and specificity for classifying walking speed.
- The 30-second chair stand test (30CST) and TMST were found to be safe and feasible for remote assessment via videoconferencing.
- The 5XSTS test had poor validity (ICC = 0.148) and low specificity for classifying walking speed.

## Abstract

Lower limb function teleassessments in adult stroke survivors are critical due to the rise of telerehabilitation. This preliminary study aims to establish the safety, feasibility, and validity of teleassessments, exploring their predictive capacity for walking speed.

In this cross-sectional study, 17 chronic stroke participants (11 males, 6 females; mean age: 63 ± 8) underwent lab-based and tele-based assessment sessions. Outcome measures included the 30-second chair stand test (30CST), five times sit-to-stand (5XSTS), and 2-minute step test (TMST). 10-Meter walk test categorized participants into slow versus fast (below/above 0.8 m/s). Absolute agreement and bias analyses determined teleassessment validity, while sensitivity and specificity analyses evaluated their effectiveness in classifying slow versus fast walkers.

Absolute agreements were good for 30CST (ICC = 0.885), poor for 5XSTS (ICC = 0.148), and excellent for TMST (ICC = 0.971). 30CST (p = 0.8), 5XSTS (p = 0.3), and TMST (p = 0.2) showed no systematic bias. Similarly, 30CST (R2 = 0.07, p = 0.25), 5XSTS (R2 = 0.04, p = 0.39), and TMST (R2 = 0.0008, p = 0.9) indicated no proportional bias. The TMST demonstrated higher sensitivity (80%) and specificity (50%) compared with 30CST (sensitivity 88% and specificity 12%) and 5XSTS (sensitivity 80% and specificity 25%).

30CST and TMST demonstrated adaptability for remote assessment via videoconferencing, establishing their safety, feasibility, and validity. These findings advocate for integrating teleassessments into adult stroke survivors’ telerehabilitation programs.