JMIR Publications

ABSTRACT

Background:

Objective measures of physical function in older adults are widely used to predict health outcomes such as disability, institutionalization, and mortality. App-based clinical tests allow users to assess their own physical function and have objective tracking of changes over time by use of their smartphones. We developed three smartphone apps with instrumented versions of the ‘Timed Up and Go’ (Self-TUG), ‘Standing tandem’ (Self-Tandem) and ‘Five times sit-to-stand’ (Self-STS), respectively. Results from these tests can potentially guide interventions remotely, and provide more detailed prognostic information about the participants’ physical performance for the users themselves and for therapists and other health care personnel.

Objective:

Develop and test the usability of three smartphone app-based self-tests of physical function using an iterative design.

Methods:

The apps were tested in three iterations; the first and second in a lab-setting, and the third in a separate home-based study. Participants were healthy adults between 60 to 80 years of age. Assessors observed while participants self-administered the tests without any guidance. Errors were recorded and usability problems defined. Problems were addressed in each subsequent iteration. Perceived usability in the home-based setting was assessed by use of the System Usability Scale (SUS), the User Experience Questionnaire (UEQ) and semistructured interviews.

Results:

Seven usability problems were identified in the first iteration, where 22% and 27.5% were able to correctly perform the Self-TUG and the Self-Tandem, respectively. In the second iteration, errors caused by the problems identified in the first iteration was drastically reduced, where 83.1% and 75.8% of the participants correctly performed the Self-TUG and Self-Tandem, respectively. A first version of Self-STS was also tested in this iteration, with a completion rate of 30.1%. For the third usability test, the seven usability problems initially identified were further improved. Testing the apps in a home-setting gave rise to some new usability problems, and for Self-TUG and Self-STS, the rate of correctly performed trials was slightly reduced from the second version, while for Self-Tandem it increased. Mean score on the SUS was 77.63 ± 16.1, and 80-95% of the participants reported the highest or second highest positive rating on all items in the UEQ.

Conclusions:

The study results suggest that the apps have the potential to be offered as a solution for self-testing of physical function in seniors in an non-supervised home-based setting. The participants reported a high degree of ease of use. Evaluating the usability in a home-setting allowed us to identify new usability problems that could affect the validity of the tests. These usability issues are not easily found in the lab-setting, indicating that if possible, app usability should be evaluated in both settings. Before being made available to end-users, the apps require further improvements and validation.