There is to our knowledge few validated electronic tool in MS that measures the distance walked by a person with MS (PwMS). Utilising global positioning systems (GPS), WI-FI positioning and an in-built smartphone accelerometer to measure distance walked by a PwMS outdoors or indoors, could alleviate the uncertainty around using pedometers in those with gait disturbances, and is an attractive option.
To pilot an accurate measure of distance walked using a smartphone application (mSteps) to facilitate Expanded Disability Status Scale (EDSS) measurement, indoors and outdoors, using both an MS and a control cohort.
The pilot study recruited 25 PwMS and 10 controls. mSteps utilised the iPhone’s inbuilt accelerometer and GPS functionalities to calculate the distance walked and time taken, indoors and outdoors. Due to unpredictable weather the physician monitored walk took place indoors which was fitted with location beacons to allow for WI-FI indoor positioning. The control cohort did the same walk indoors and outdoors to validate the use of the GPS functionality.
The participant was instructed to walk 25 feet, without rest, whilst the study phone was attached to their arm using a runner’s arm band and study personnel walked alongside them with a trundle wheel. Measurements were taken at 3 separate time points within a 3-month period.
95% levels of agreement between app and trundle wheel (gold standard) were calculated using the Bland-Altman repeated measures analysis. Levels of agreement, app vs trundle, were calculated for indoor measurements on both PwMS and controls with additional app vs trundle outdoor measurements for controls only. The a priori defined clinically acceptable difference was 1.52m.
The 95% levels of agreement for indoor measurements on PwMS were -2.46 to 2.27m; and for controls were -2.02 to 2.71m. The 95% levels of agreement for outdoor measurements on controls were -0.45 to 0.43m.
The outdoor GPS functionality of mSteps is very accurate as shown by the 95% levels of agreements compared to the a priori clinically determined difference. The indoor WI-FI positioning function of mSteps however, was not accurate enough and shows that it is not reliable enough for further use. The control cohort showed the same inaccuracy indoors which eliminates the possibility that an uneven gait pattern in the MS cohort contributed to the error margin. A further validity study is being carried out, looking at a cohort of PwMS walking outdoors using mSteps and a trundle wheel.