Mobility over time

Proposed approach

Aggregate quality-controlled movement features over predefined observation periods. Track availability alongside estimates so absence of data cannot be mistaken for inactivity.

Evaluation

Study repeatability before interpreting a change as meaningful. Analyse room layout, direction and observation coverage.

Limitations

Mobility features are not diagnoses of frailty. External studies cannot establish Hakilix clinical validity.

Measurement before interpretation

A longitudinal summary describes observations over time. Before interpreting a trend, a research protocol should define which movements are eligible, how observation windows are selected and how much coverage is required. A shorter observed walking duration could reflect a sensor outage, a different routine or time spent outside the room.

Proposed features to investigate

Candidate features include the distribution of observed movement velocities, duration of identifiable transitions and consistency of repeated room transits. Each feature needs a definition, a quality threshold and an explanation of its dependence on sensor geometry. Features would be evaluated as research measurements rather than diagnostic scores.

Within-person comparisons

An initial reference period could be used to explore individual variability. The protocol should explain how that period is chosen, when a reference becomes stale and whether changes in furniture or sensor position require recalibration. A missing reference must not be silently replaced with an unrelated population average.

Reporting uncertainty

Summaries should show the amount of usable data, missing intervals and uncertainty around any estimated change. Repeated measurements need an analysis that accounts for their dependence. Prospective study design and appropriate statistical review would be needed before investigating associations with clinical outcomes.