Endpoint specification, protocol-dependent measurement properties and clinical interpretation of gait and balance assessments.
Define the construct and the estimand
Specify the construct, target population and decision threshold before selecting instrumentation. Relative reliability depends partly on between-person heterogeneity; absolute agreement and measurement error determine whether a difference is interpretable at the individual level. Clinical validity concerns the intended concept and context of use, while responsiveness concerns change. These properties require distinct evidence. V3 evaluation framework.
Distance/time, expressed in m/s; distinguish usual, maximal and free-living speed.
Reporting specification. Define the timed segment, static versus moving start, acceleration exclusion, instruction and assistive-device condition. For free-living summaries, specify bout eligibility and whether aggregation weights time, steps or bouts.
Interpretation. Speed is an integrated performance outcome with limited mechanistic specificity. A longitudinal difference requires comparable acquisition and an error estimate appropriate to the protocol.
Step frequency, conventionally steps/min; cadence and mean step length jointly determine forward speed.
Reporting specification. Identify the event detector and observation interval. Avoid pooling non-steady transitions with steady-state walking unless the endpoint intentionally includes them. Report cadence with spatial progression to distinguish frequency-driven from length-driven pace changes.
Interpretation. A faster cadence is not inherently better: interpretation depends on speed, limb length, task and compensatory strategy.
Spatial progression between successive same-foot contact events.
Reporting specification. Specify the event convention, laboratory or inferred displacement, units and any normalisation to anthropometry. Distinguish left and right estimates from pooled means and report the algorithm’s performance in the relevant speed range.
Interpretation. A spatial estimate derived from inertial signals is model-dependent. Changes in foot motion or attachment can affect analytical performance independently of clinical recovery.
The two portions of a walking cycle during which both feet contact the ground.
Reporting specification. Report whether the outcome is total or phase-specific double support, absolute duration or percentage of stride, and how contact events are identified. Stratify or adjust for speed where appropriate to the question.
Interpretation. Longer double support may characterise a cautious support strategy, but it is not a direct measure of dynamic stability or a stand-alone fall-risk estimate.
A defined bilateral relationship for a specified spatial or temporal parameter.
Reporting specification. Publish the equation, denominator, sign convention and aggregation. Signed differences preserve direction; absolute ratios discard it. Analyse limb-specific trajectories alongside the index.
Interpretation. Improved symmetry can arise from recovery of one limb or deterioration of the other. The same index can therefore represent clinically different trajectories.
Within-series dispersion, commonly a standard deviation or coefficient of variation for step or stride measures.
Reporting specification. Report sample length, bout composition, outlier handling and whether turns or transitions are excluded. Separate within-bout from between-bout variation; estimate precision for the chosen statistic.
Interpretation. Mixed speeds, event-detection errors and nonstationary recording can inflate dispersion. Variability is not interchangeable with asymmetry, complexity or dynamic stability.
Accumulated locomotor exposure, expressed as steps, walking time, distance or bout distributions.
Reporting specification. Define valid observation time, nonwear classification, minimum bout length and aggregation across days. Specify how incomplete days enter the denominator and examine whether missingness depends on symptoms or activity.
Interpretation. The construct is enacted mobility, influenced by capacity and opportunity. A reduced total should be interpreted with coverage and bout structure rather than assumed to indicate physiological decline.
Related but distinct constructs spanning reorientation, transitions and postural control.
Reporting specification. For turning, define angle, event boundaries and angular-velocity processing. For balance, specify task, support surface and the actual sway or performance outcome. Validate each construct against an appropriate reference.
Interpretation. A single mobility total can conceal component-specific deficits. Turning performance and quiet-standing sway should not be pooled as equivalent measures of stability.
Select an instrument by construct coverage, population-specific measurement properties and the decision it will inform. The illustrations identify the task; the linked protocols define administration and scoring. Preserve protocol fidelity when interpreting longitudinal change.
A protocolised measure of short-distance walking speed, with comfortable and fast conditions representing different performance targets. Acceleration strategy and timed segment are part of the test definition, not interchangeable setup details.
Protocol specification. Document the specific protocol, static or moving start, timed distance, number of trials and aggregation. Record footwear, assistance and device use; preserve the same acquisition conditions for reassessment.
Interpretation. Interpret individual change against absolute repeatability and population-specific meaningful-change estimates. A between-session difference smaller than measurement error cannot be rescued by a statistically significant group effect.
A composite functional-mobility task integrating sit-to-stand, a 3-m walk, turning, return and controlled sitting. Total duration can conceal different contributions from transfer mechanics, straight walking and reorientation.
Protocol specification. Specify chair configuration, instructions, aid, footwear and trial handling. Instrumented decompositions require explicit event boundaries and component-level validation in addition to agreement for total duration.
Interpretation. Use component and qualitative findings to explain the total. Its value in falls assessment is contextual; a universal cutoff ignores population spectrum and the multifactorial nature of falls.
A standardised sustained-walking assessment with distance as the principal endpoint. The observed performance integrates physiological capacity, symptoms and pacing over a fixed interval.
Protocol specification. Retain the applicable protocol for course layout, standardised encouragement, rests, monitoring and stopping criteria. Document assistance and relevant treatment conditions. Learning and turning frequency can influence repeatability.
Interpretation. Use diagnosis- and protocol-specific estimates of responsiveness. Free-living accumulated distance is a different estimand and cannot substitute for a controlled six-minute performance endpoint.
Ten tasks challenge postural stability during walking under changing demands, including speed changes, head movements, obstacles and stairs. The assessment samples adaptability beyond steady-state pace.
Protocol specification. Apply the published 0–3 item criteria, yielding a maximum of 30. Retain item-level scores to identify which demands contribute to the total and use the defined administration conditions.
Interpretation. The total is ordinal. Evaluate ceiling effects and population-specific measurement properties before treating a numerical change as an equal-interval improvement; instrumented features may explain performance but do not automatically reproduce the clinical score.
Fourteen functional tasks sample balance during transfers, standing and progressively challenging activities. The scale addresses performance under specified tasks rather than continuous exposure to environmental hazards.
Protocol specification. Use the published item criteria and maximum total of 56. Record assistance and deviations. An item-level profile can help distinguish task-specific limitation from a similar total obtained through different failures.
Interpretation. Consider ceiling effects in higher-functioning populations and whether the intended construct is adequately challenged. Static and functional task performance do not exhaust the demands of dynamic locomotor stability.
Repeated sit-to-stand performance integrates lower-limb force production, balance and task execution under a fixed time constraint. It complements gait measures by challenging a different functional demand.
Protocol specification. Follow CDC instructions for chair setup, arm position and completed-stand counting. Record inability to complete the standard task rather than silently substituting an assisted version into the same reference comparison.
Interpretation. Chair height, upper-limb assistance and movement strategy change the mechanical demand. Interpret repeated counts under comparable conditions and in the appropriate population context.
Progressively narrower and unilateral support conditions challenge standing postural control. Performance provides a structured clinical observation complementary to gait and strength assessment.
Protocol specification. Apply the CDC sequence: feet together, semi-tandem, tandem and single-leg stance, with the specified timing and safety procedures. Document achieved position, hold duration and any protocol deviation.
Interpretation. The result describes task performance rather than a direct physiological decomposition of balance. Combine it with clinical findings and other falls-assessment domains; sensor-derived sway is a separate endpoint requiring its own validation.
Specify the clinical question, target population and action the result could support.
Confirm that the measurement and reference method match that question.
Review absolute errors, repeatability and uncertainty for the relevant parameter.
Check acquisition conditions, sensor placement and hardware/software versions.
Examine valid wear, excluded observations and the handling of missing data.
Separate laboratory performance from free-living performance.
Assess patient burden, accessibility and usability in the intended population.
Review funding, manufacturer involvement and external replication.
Pre-specify analysis and decision thresholds before testing clinical benefit.
V3 and V3+ provide complementary evaluation frameworks.
How to use the ratings and topic labels
Stars indicate editorial relevance to clinicians and researchers using this library: 5 = foundational or broadly applicable; 4 = directly useful for a defined question; 3 = specialist or emerging application; 2 = indirect relevance; 1 = background context. They are not journal impact factors, evidence-certainty grades, product comparisons or estimates of clinical benefit. Lists place higher-relevance entries first; chronological steps and test protocols retain their required sequence. Topic labels use the same colors throughout the library, with text identifying every subject.