How can healthcare teams measure movement between appointments?

One-off assessment versus longitudinal monitoring

A single assessment describes a moment. It answers “how is this person moving today?” — useful at a baseline, a discharge decision or a study endpoint.

Longitudinal monitoring answers a different and often more clinically useful question: “is this person changing, in which direction, and how fast?” Because each measurement is compared with the same person’s earlier results, using the same method and protocol, small consistent differences become visible in a way they never are across separate clinic visits months apart.

• A one-off measure needs to be accurate in absolute terms to be interpretable against a norm.
• A repeated measure needs above all to be consistent, because the comparison is internal.
• Consistency comes from holding the protocol, device placement and setup steady between assessments.

Objective movement data

Objective measures give a numerical record of how someone moved during a defined task. Captured on everyday devices, they can be repeated frequently without a clinic visit.

• Gait measures — speed, cadence, step and stride length, gait timing.
• Joint measures — joint angles and range of motion for defined, visible movements.
• Functional measures — Timed Up and Go time, 30-second Sit-to-Stand repetitions.
• Consistency measures — how similar repetitions are within a set.

Their value in monitoring is comparability. A number recorded the same way each time can be plotted; a note describing how someone looked cannot.

Patient-reported outcome measures

PROMs capture what movement measurement cannot: pain, confidence, function in daily life, and whether the person feels they are getting better.

The two often diverge, and that divergence is informative. Measured mobility improving while reported pain worsens is a different clinical picture from both improving together. Collecting PROMs alongside objective measures, at the same points in time, makes that comparison possible rather than anecdotal.

Wearable activity context

An assessment shows what someone can do when asked. Wearable activity data shows what they actually did across the days in between.

• Activity volume and pattern over days and weeks.
• Whether an assessed improvement translates into more real-world activity.
• Whether a period of reduced activity precedes a decline in assessed movement.
• Context for interpreting an unexpectedly poor assessment — illness, a bad week, unusual demands.

Activity data is contextual rather than precise. It describes behaviour, not movement quality, and device algorithms differ. It is best read alongside assessed measures, not instead of them.

Monitoring recovery and change

1. 1Establish a baseline early, with the protocol and setup that will be repeated.
2. 2Choose a review interval that matches the expected rate of change — weekly during early rehabilitation, monthly for a slower condition.
3. 3Record what happened between assessments: interventions, procedures, changes in treatment, setbacks.
4. 4Look at the trend rather than reacting to a single result, which can reflect a bad day or an imperfect capture.
5. 5Decide in advance what change would be meaningful enough to act on.
6. 6Note when the protocol or setup changes, because that can shift results independently of the person.

Recording interventions alongside movement data is what turns a series of numbers into something interpretable — the measured change can be read against what was actually done.

Research follow-up

• Repeated objective measurement at defined study timepoints, without requiring site visits.
• Consistent protocols across participants and across sites.
• Reduced participant burden, which supports retention in longer follow-up.
• Structured, exportable data for analysis rather than transcribed observations.
• Movement measures collected alongside PROMs and activity context for the same participants.

Where a study depends on a specific measure, start by checking the validation for that measure, including the tested population and setting.

A generic monitoring workflow

1. 1An assessment is initiated by a clinician or researcher, scheduled or as part of a follow-up plan.
2. 2A secure browser link reaches the person, tied to them and time-limited.
3. 3A guided assessment runs on their own device, with permissions, a short setup step and clear instructions.
4. 4The capture is processed and quality-checked, with unusable captures flagged rather than reported as valid.
5. 5The result is added to that person’s longitudinal record, alongside earlier assessments, PROMs and activity context.
6. 6An authorised clinician or researcher reviews the trend and decides what, if anything, it means.

Practical constraints worth planning for

• Somebody has to look at the data — monitoring without a review process just accumulates records.
• More frequent measurement is not automatically better; match the interval to how fast change is expected.
• Assessments must be safe to perform unsupervised, and it should be clear when they are not.
• Consistency of setup between assessments matters as much as the measurement itself.
• Match each measure to the population and setting described in its validation evidence.
• Repeated collection of health data brings information-governance obligations around lawful basis, retention and the person’s rights.