Assessing a novel way to measure three common rehabilitation outcome measures using a custom mobile phone application

Clinicians often use thirty-second-sit (chair)-to-stand (30CST), timed-up-and-go (TUG), and the five-times-sit-to-stand (5xSTS) since these outcome measures (OMs) are sensitive for strength, balance and mobility.

The purpose of this study was to validate a custom smart phone application (App) that can remotely assess the 30CST, TUG, and 5xSTS.

Thirty-one healthy adults (range: 22-55 y; 54.6-106.8 kg; 160-185 cm; 19 females) participated in this cross-sectional study. Each participant performed the 30CST, TUG, and 5xSTS at a slow and normal speed. They performed each OMs twice while the App collected their performance data using both an iOS and Android phone. The gold standard of each test was the average of the silent count of two investigators for the 30CST and the time recorded by two investigators using stopwatches for the TUG and 5xSTS. Investigators analyzed the data using Intraclass Correlation coefficients (ICC), Pearson R coefficients, Signed Rank Tests, and Wilcoxon Rank-Sum Tests.

A significant correlation was observed between the performances recorded by the phones and the direct observation gold standard for all three OMs (r > 0.97). For 30CST, no significant mean count differences were found for the following comparisons: between phones, within phone types, or within phone-by-speed levels. (P-values range 0.06-1.00). While a statistically significant difference was found in all of the time comparisons when performing TUG and 5xSTS (p < 0.0001) except for the between phone comparison with TUG (p = 0.27). For TUG and 5xSTS, the time difference was less than a second when compared to the gold standard and ICCs showed moderate to strong agreement when comparing the phone application to the gold standard (ICCs range 0.60-0.99). These data suggested that the App could validly measure performance of these OMs.