SWISSTEP / KNOWLEDGE LIBRARY
Gait monitoring: questions and answers
58 searchable answers for clinicians, researchers and teams implementing movement measurement.
Find practical answers about measurement, protocols, clinical interpretation and deployment. Search a term such as “asymmetry”, “Parkinson”, “missing data” or “reimbursement”.
58 questions
★★★★★What is the difference between gait analysis and gait monitoring?
Gait analysis characterises walking mechanics or performance during a defined observation. Monitoring repeats observations over time. A monitoring programme also needs a sampling plan, valid-data rules and an agreed interpretation of change.
★★★★★Which gait measure should be selected first?
Start with the clinical or research question. Speed may suit a standardised mobility endpoint; asymmetry, turning or walking volume may address different questions. Choose the quantity before choosing the device.
★★★★★What is the difference between validity and reliability?
Validity concerns whether a measure represents the intended quantity. Reliability concerns repeatability under specified conditions. A consistently biased measurement can be reliable without agreeing with a reference.
★★★★★Why is correlation insufficient for validation?
Correlation measures association. Two methods can correlate strongly while differing systematically. Examine bias, absolute errors and limits of agreement for the intended range of measurements.
★★★★★What does clinical validation establish?
It links a measurement to the clinical concept in its stated context of use. It does not, on its own, establish the benefit of a treatment or monitoring service.
★★★★★How should missing data be handled?
Separate nonwear, technical loss and missing assessment from observed inactivity. Pre-specify inclusion and analysis rules. Report coverage so a low movement total is not mistaken for a complete day.
★★★★★Can monitoring identify an unsuccessful surgical recovery?
It may reveal an unexpected trajectory that prompts review. Gait data alone cannot determine why recovery is slow or diagnose infection, implant failure or another complication.
★★★★★Is slow walking alone sufficient to assess fall risk?
No. It is one observation. Assess fall history, balance, strength, medicines and other relevant factors rather than treating one walking parameter as a complete evaluation.
★★★★★Can a wearable diagnose Parkinson’s disease?
A device may contribute measurements to research or clinical assessment, but a validation study of gait outputs does not establish a stand-alone diagnostic indication. Examine the precise intended use and evidence.
★★★★★Can walking data diagnose dementia?
Associations between gait and later cognitive decline support research and broader assessment. They do not establish a wearable-only diagnostic test for dementia.
★★★★★What should a dual-task walking protocol specify?
Describe the secondary task, instructions, task priority and scoring of both walking and secondary-task performance. Changing task difficulty can change the interpretation of the comparison.
★★★★★Does a CE mark guarantee reimbursement?
No. Market-access requirements and payer decisions are separate. European coverage pathways vary by country and by the service or product being proposed.
★★★★★Is gait monitoring automatically eligible for US remote-monitoring payment?
No. Eligibility depends on the service, device, condition, practitioner and current payer rules. Verify the date-of-service requirements and documented work before billing.
★★★★☆Are step length and stride length the same?
No. Step length spans successive contacts of opposite feet; stride length spans successive contacts of the same foot. A stride contains two steps during alternating walking. Confirm the event convention used by the system.
★★★★☆How is walking speed calculated?
Speed is distance divided by elapsed time. State the units, timed distance and start procedure. A value derived from short indoor bouts is not automatically comparable with a standardised straight walk.
★★★★☆What is cadence?
Cadence is the number of steps per unit time, usually steps per minute. Do not report strides per minute as steps per minute. Cadence and step length together contribute to forward speed.
★★★★☆What does double-support time mean?
It is the time during which both feet contact the ground. State whether the value is seconds or a percentage and whether the denominator is a stride or another interval. Interpretation depends partly on speed.
★★★★☆What is gait asymmetry?
It describes a difference between corresponding left and right measurements. Report the parameter and exact formula. A signed difference preserves direction; an absolute score may conceal which limb changed.
★★★★☆What does stride-to-stride variability capture?
It describes dispersion across repeated cycles. Report the statistic, number of strides and selection rules. Turns, mixed walking speeds and event-detection errors can affect the result.
★★★★☆Why can mean values hide important changes?
Two recordings can have the same average but different distributions, outliers or bout composition. Inspect repeatability and the spread of observations before concluding that a person’s walking is unchanged.
★★★★☆Is a daily step count enough to describe mobility?
It describes an amount of detected activity, not the full mechanics or quality of walking. It does not by itself explain speed, asymmetry, balance, symptoms or whether the recording covered the whole day.
★★★★☆What is analytical validation?
It evaluates whether an algorithm accurately derives the intended measurement from sensor data. It is distinct from establishing clinical relevance or demonstrating that using the measure benefits patients.
★★★★☆What does V3+ add?
V3+ explicitly includes usability validation: whether intended users can operate the technology as required. This addresses real-world barriers that accurate sensing alone cannot resolve.
★★★★☆Can a laboratory validation be applied to home monitoring?
Only with evidence that the relevant conditions are covered. Home recordings introduce variable placement, terrain, turns, short bouts and missingness. Examine performance in the intended deployment setting.
★★★★☆How many strides are needed?
There is no universal minimum for every metric. Specify the outcome, desired precision and bout conditions. A count adequate for mean speed may be inadequate for a stable variability estimate.
★★★★☆Should turns be included in a walking summary?
Define this before analysis. Turning and straight walking are different tasks. Either report them separately or state how their inclusion affects the endpoint and its interpretation.
★★★★☆What should be recorded about sensor placement?
Record the body site, side, orientation, attachment and any placement changes. Use the configuration covered by the validation evidence. A device name alone does not define the measurement system.
★★★★☆Can different software versions be pooled?
Check whether processing, event detection or output definitions changed. Preserve version provenance and assess comparability before combining observations into a longitudinal endpoint.
★★★★☆What does minimum detectable change mean?
It concerns change exceeding measurement error under stated conditions. It is different from a patient-important change. Use estimates matching the population, protocol and confidence convention.
★★★★☆Is 0.1 m/s always a meaningful improvement?
No. Perera and colleagues proposed approximately 0.05 m/s and 0.10 m/s as initial small and substantial change estimates in studied older-adult datasets. These are context-dependent estimates, not universal diagnostic thresholds.
★★★★☆What belongs in a gait-data report?
Include the metric definition, units, task, acquisition conditions, usable sample, processing version and uncertainty. Show the comparator and time period used to support an interpretation.
★★★★☆Can a smartphone replace a motion laboratory?
For some validated outputs and conditions it may be a practical alternative. Detailed joint mechanics, forces and complex tasks can require different instrumentation. Read metric-specific agreement and limitations.
★★★★☆What has OneStep’s validation research examined?
Christensen and colleagues studied several walking conditions in healthy adults and postoperative feasibility in a small arthroplasty group. The paper reports parameter-dependent validity and reliability rather than one universal accuracy value.
★★★★☆Do smartphone gait measures replace patient-reported outcomes?
No. Associations with functional categories show related information, not interchangeability. Symptoms, confidence and perceived function remain important outcomes alongside objective movement measures.
★★★★☆Why can gait speed improve while walking volume falls?
Capacity and behaviour are different constructs. A patient may walk faster in a test but take fewer trips because of symptoms, opportunity or confidence. Also check recording coverage before interpreting the difference.
★★★★☆How should walking aids be handled during reassessment?
Record the aid and assistance and aim for comparable testing conditions. If the aid changes, report that change explicitly rather than attributing the entire measurement difference to recovery.
★★★★☆Can a gait score predict the exact time of a fall?
No. A risk estimate is not a prediction of an individual event’s timing. Falls assessment considers multiple domains and should connect findings to appropriate preventive action.
★★★★☆What does the Timed Up and Go measure?
It samples functional mobility through rising, walking, turning and sitting. Total time combines those components, so qualitative observation and component measures may explain differences hidden in the total.
★★★★☆How does the FGA differ from comfortable gait-speed testing?
The FGA samples stability during varied walking tasks and yields an ordinal score. A short gait-speed test measures pace under a more constrained protocol. They address complementary questions.
★★★★☆Does the Berg Balance Scale measure everyday fall exposure?
No. It measures performance on a set of functional balance tasks. Exposure to hazards and daily mobility require additional information.
★★★★☆Why add a chair-stand test to gait monitoring?
It adds a structured observation of functional lower-limb strength. Its task requirements differ from walking, helping broaden an older-adult falls assessment.
★★★★☆Can gait monitoring guide medication review in Parkinson’s disease?
Time-linked observations may help a clinician examine motor fluctuations. Record medication timing and context. Monitoring evidence does not authorise autonomous dose adjustment.
★★★★☆Why do hand position and phone location matter?
They change how body motion reaches the sensor. In Parkinson@Home, hand position affected wrist and pocket signals, illustrating why context and placement belong in validation.
★★★★☆Is freezing of gait the same as generally slow gait?
No. They are different phenomena and require appropriate definitions and measurement methods. Review freezing-specific event detection and reference annotations rather than substituting average speed.
★★★★☆Can gait reveal mood or mental state?
Movement is influenced by many physical, environmental and behavioural factors. Do not infer an individual’s mood or psychiatric diagnosis from gait changes without validated, context-specific evidence.
★★★★☆How do the 10-Meter and 6-Minute Walk Tests differ?
The first measures pace over a short defined distance; the second measures distance covered over a standardised time interval. They address different aspects of walking performance.
★★★★☆Is a home step total equivalent to a six-minute walk result?
No. One is accumulated detected activity; the other is a standardised test. Neither can be converted into the other without an appropriate validated method.
★★★★☆Can an inertial sensor directly measure ground-reaction force?
Its accelerometer and gyroscope measure motion, not a force-plate signal. A force estimate derived from motion is a model output requiring separate validation against an appropriate reference.
★★★★☆What does surface EMG add?
Surface EMG describes recorded muscle electrical activity and can complement kinematics. It does not directly supply joint force, and acquisition and processing choices affect interpretation.
★★★★☆When is an instrumented walkway useful?
It supports observed footfall and temporospatial measurement under controlled conditions. It does not by itself describe all-day mobility or detailed full-body kinetics.
★★★★☆Can a public dataset validate a commercial clinical claim?
It may support a defined analytical evaluation if its population and acquisition match the claim. It does not replace prospective clinical evidence, external validation or product-specific regulatory assessment.
★★★★☆How can train–test leakage occur in gait research?
Closely related observations from the same participant can enter both partitions and inflate apparent generalisation. Plan separation at the appropriate participant, site or time level and document it.
★★★★☆Does FDA clearance of a sensor cover every derived algorithm?
No. Assess the complete product and its intended use. A component’s authorisation cannot be assumed to cover a new clinical interpretation or service.
★★★★☆Where can I find reusable gait datasets?
PhysioNet and Mobilise-D are useful starting points. Review each dataset’s licence, access conditions, sensor configuration, cohort and citation requirements before reuse.
★★★★☆Is the PhysioNet Parkinson’s gait dataset smartphone data?
No. The linked collection contains under-foot force recordings from participants with Parkinson’s disease and controls. Match the data modality to the algorithm you intend to test.
★★★★☆How should provider claims be compared?
Compare the same endpoint, population, reference, task and error statistic. A broad accuracy claim or a long publication list is not a substitute for relevant metric-level evidence.
★★★★☆What do the stars in this library mean?
They indicate editorial relevance to the library’s professional readership. They are not a product-quality ranking, journal impact factor or graded estimate of treatment benefit.
★★★★☆Where should a clinical team start with OneStep?
Review its original validation and functional-association studies, then its clinical resources. Request evidence for the intended population, phone configuration and workflow before adopting a specific endpoint.
No matching questions. Try a broader term or clear the search.