Gait assessment and prognosis after total hip arthroplasty

Executive assessment

Gait assessment after total hip arthroplasty (THA) should distinguish five questions: can the person walk safely now; how much walking capacity has changed; how the movement is accomplished; how much walking occurs in ordinary life; and which later outcome is being forecast. A walking speed, a hip angle, a step count and a probability of falling are different outputs. Evidence supporting one does not validate the others.

Standardized usual and fast walking tests are a practical foundation, supplemented by a sustained walk when endurance and symptom tolerance matter. The strongest interpretive requirement is postoperative timing. A same-hour repeat in a discharge-ready patient on day 2, a stable preoperative retest and a test 5 years after surgery have very different error structures. Aids, restrictions, surgical indication, contralateral disease and turning rules must accompany the score. The OARSI framework identifies useful activities; it does not supply universal postoperative thresholds. [1–7]

The original evidence contains clinically consequential qualifications. A 2024 day-2 study provides direct THA error estimates for usual/maximum 10-m walking, but enrolled only discharge-ready patients and used crutches for walking while testing TUG without crutches. Its ICC confidence bounds are much less reassuring than the high point estimates alone. A late-THA 50-foot walk study reports a small error estimate but describes an out-and-back course in its methods and a straight course in its appendix. A 2026 important-change study exists, but its abstract reports poor discrimination of perceived improvement by 10MWT and 2MWT, despite presenting several MCID estimates. These sources support careful measurement, not automatic green/red response labels. [4–8]

Prospective gait evidence is endpoint-specific. Preoperative speed and TUG can help explain inpatient recovery under particular pathways. Postoperative gait abnormality has been associated with subsequent falls in a selected female cohort. Other prospective evidence found no association between preoperative maximum speed and falls, while a frequently cited small gait/falls study measured gait after the year of falls it tried to explain. These findings are compatible once timing, ascertainment and outcome are separated. None licenses a universal speed threshold for discharge, future falls or participation. [9–19]

Instrumented and portable methods add valuable information about movement strategy and daily performance. However, agreement is variable-specific, and use of crutches can particularly affect detection of gait-cycle phases. A recent single-thigh-IMU hip-ROM model still depended partly on optically measured walking speed and had wide limits of agreement. Smartphone recovery curves are informative descriptions of an observed cohort, but a group becoming statistically indistinguishable from its own preoperative mean is not an individual recovery criterion. [20–25]

For a rehabilitation tool, the most defensible first product is a protocol-aware record of capacity, assistance, symptoms and movement features, with transparent source-specific uncertainty. Patient-level prognosis should be a separate, locally validated function with a defined outcome, prediction time and horizon, missing-data policy and calibration assessment.

Scope and evidence approach

The principal population is adults undergoing primary elective THA for hip osteoarthritis. Preoperative assessments are included when they establish an individual's baseline or predict a later postoperative outcome. Fracture-related THA, hemiarthroplasty, revision, resurfacing, bilateral procedures and other indications are not merged with that population. Mixed hip/knee studies remain explicitly mixed. The report concerns assessment and prognosis rather than comparative surgical or exercise effectiveness.

This is a critical narrative review with an auditable search and source-access record. Three exact native PubMed streams addressed measurement properties, prognosis/recovery and protocol/biomechanics/technology, supplemented by Scopus, original-source searching and bounded reference/citation chasing. All exact PubMed result sets were reconciled through official NCBI ESearch and EFetch because connector pages overlapped. The appendix reports executed queries and coverage, with original-source access limits. Retrieval counts are not eligible-study counts, and this work does not claim exhaustive dual-reviewer screening, registration or a formal systematic review.

Population stage and construct

Define the operation and the patient

THA is a procedure, not a homogeneous functional state. Record the indication, side, primary versus revision status, approach, dates, weight-bearing instructions and current movement precautions. An early posterior-approach cohort allowed unrestricted weight bearing cannot establish the safety or measurement properties of the same protocol after a fracture, intraoperative complication or restricted reconstruction. Use the treating team's current restrictions, rather than inferring precautions from an approach name alone.

Record contralateral hip and knee symptoms and prior replacements. The nonoperated limb is not necessarily a healthy reference. Studies excluding bilateral disease often provide cleaner biomechanical comparisons while excluding an important clinical subgroup. Conversely, pooling unilateral and bilateral replacements can conceal different compensatory possibilities. Analgesia, dizziness, fatigue, footwear and assistive devices can affect the observation even if the implant and radiograph are unchanged.

Preoperative, inpatient, early outpatient, 3–6-month and at-least-12-month assessments are useful organizing categories, but exact elapsed days remain necessary. A day-2 enhanced-recovery patient and a 3-week inpatient in an older rehabilitation pathway are not equivalent stages simply because both are described as early postoperative. Shibuya's cohort averaged 13.5 days in hospital, while Oosting's short-stay endpoint was discharge within 36 hours. Their thresholds answer different service questions. [9, 11]

Separate capacity mechanics and performance

A clinic walk measures capacity under a particular instruction. Maximum speed is a challenge to available reserve; usual speed is the pace selected for that context. A 6MWT additionally samples sustained effort, turning, rests, symptoms and motivation. TUG combines chair rise, straight walking, turning and sitting. Its total time cannot identify which phase limits performance.

Kinematics and kinetics characterize movement, not merely completion. A person may regain speed through altered pelvic/trunk strategy while hip excursion or loading remains different from controls. Gasparutto's instrumented TUG study illustrates the value of phase separation: different component tasks recovered differently at 6 months, with substantial between-patient variability. Its task used a 47-cm armchair, permitted arm support and analyzed the fastest of the first three trials; this is not equivalent to every clinical TUG implementation. [26]

Free-living walking is another construct. Step counts depend on opportunity, behavior, wear time, sensor carriage and algorithm rules as well as ability. In Jeldi's 30-person longitudinal study, pain, self-reported function and 6MWT improved without a corresponding change in activity-volume measures. The original abstract supports that dissociation; the complete author manuscript was not accessible here. Capacity gains should therefore not be displayed as proof that daily activity or participation increased. [25]

Table 1 Select the walking assessment for the question

Clinical walking protocols

Short distance walking

Select usual speed, maximum safe speed or both according to the question. Keep instruction, start method, timed segment, acceleration/deceleration space, direction, surface, footwear and aggregation constant. A central 10 m timed within a 14- or 16-m course differs from timing the entire course from standing. One maximum trial differs from a mean of two. Aids are part of the measurement condition, not a nuisance variable to remove when interpreting change.

The OARSI 40-m fast-paced test uses four 10-m lengths with the specified handling of turns. Kennedy's fast self-paced test timed two 20-m lengths with the turn excluded and a particular safe-fast instruction. Both cover 40 m, yet they should not be treated as the same protocol. A time-domain MDC cannot be converted into a single constant speed-domain MDC by simply replacing its unit: speed equals distance divided by time, so the corresponding speed change depends on baseline time. [1, 2]

The 50-foot walk illustrates why the original methods matter. Unver and colleagues explicitly described 25 feet out, a 180-degree turn and 25 feet back; their appendix instead describes a straight 50-foot walkway timed from the start cue to the trunk crossing its end. The paper's reliability numbers should retain that unresolved course ambiguity. Choosing one version for practice is reasonable, but calling its error threshold validated by this paper requires caution. [4]

Sustained walking and composite mobility

Use 2MWT or 6MWT when sustained walking is clinically relevant, and preserve corridor/circuit length, encouragement, rests and aid. Two minutes cannot be converted to six minutes by multiplication: fatigue, symptom escalation, pacing and turns may be nonlinear. In early recovery, inability or an unsafe attempt is a result. Do not assign a fabricated distance or exclude such participants without reporting the reason.

Kennedy's mixed arthroplasty cohort used a 46-m rectangular circuit for 6MWT with standardized encouragement. Among 119 eligible at the first postoperative occasion, 33 could not complete the 6MWT; this is far more informative than reporting the mean among completers alone. Its preoperative eligibility also selected people able to perform the tasks. Both features constrain generalization to the most disabled patients. [2]

For TUG, specify chair height, arm use, pace, turning point, aid and start/stop events. A total time may be useful for screening mobility while remaining nonspecific about gait. An apparent gait improvement after changing from a walker to a cane also incorporates reduced device handling. Prefer reporting both versions when clinically useful rather than pretending that the measurement condition is unchanged.

Reliability and measurement error

What an error estimate can answer

An ICC describes relative ordering in the studied sample. Heterogeneous participants can generate a high ICC despite material absolute differences. SEM, limits of agreement and MDC/SDC answer related but different questions in original units. A same-trial device/manual comparison primarily samples measurement disagreement; repeated performance adds biological and behavioral variability; a later-day repeat adds further variation. None is interchangeable with another solely because the same test name appears.

MDC is not an important-change threshold. Under its model it estimates how large a repeated-score difference must be to exceed expected measurement error at a chosen confidence level. It does not establish that the patient values that difference, that the difference is caused by treatment, or that future disability has changed. Confidence level, individual versus group use and the source of SEM should remain visible.

Direct THA clinical evidence

Temporiti and colleagues studied 35 patients on postoperative day 2 who had already met discharge-readiness criteria, including walking at least 100 m and negotiating stairs with crutches. All had primary unilateral THA for advanced OA, a posterolateral approach and no weight-bearing restrictions. The same assessor retested after 1 hour. Walking used crutches over the middle 10 m of a 14-m course and averaged two trials; TUG used no crutches and retained the best of two after familiarization. [5]

Reported maximum-speed 10MWT ICCs were 0.94 indoors and 0.91 outdoors, with MDC95 0.13 and 0.16 m/s. Usual-speed MDC95 was 0.16 m/s in both environments. These are unusually stage-relevant estimates, but not evidence for unrestricted use in all day-2 patients. The indoor usual-speed ICC confidence interval was 0.13–0.96, and table means improved from 0.83 to 0.92 m/s at retest. The corresponding precision and possible practice/recovery effects should accompany the point estimate. Some demographic counts and indoor/outdoor statistical details are internally inconsistent in the published text; the central protocol and error estimates are reported with that qualification. [5]

The study's mean outdoor advantage was smaller than its individual MDC. That does not, by itself, show that a statistically significant group difference is random noise. Group-mean inference and individual repeated-score uncertainty are different analyses. For clinical follow-up, the appropriate consequence is to standardize environment or interpret an environment change explicitly, rather than silently applying an individual threshold to a group comparison.

Unver 2015 supplies direct late-THA data: 37 primary recipients, at least 1 year after surgery and on average 5.3 years, retested after 1 hour without aids. The 50-foot time ICC was 0.98, SEM 0.3 s and SRD95 0.8 s as reported. Those small errors describe a stable, relatively high-functioning sample and have the unresolved course ambiguity already discussed. They cannot be used as day-to-day error during acute recovery. [4]

Two additional THA studies reported excellent same-day reliability in their original abstracts. Unver 2013, n = 34, reported 6MWT ICC 0.96, SEM 3.67 m and SRD95 10.17 m after a 1-hour rest. Yuksel 2021, n = 37, reported 2MWT ICC 0.96, SEM 6.37 m and MDC95 17.56 m, and TUG MDC95 1.62 s. Complete methods were unavailable, so stage, course and selection restrictions remain acquisition gaps. The studies do not justify choosing whichever published error is smallest. [6, 7]

Preoperative and mixed joint estimates

Tolk's 90-person primary-THA cohort was assessed before surgery and at 12 months, but the reliability substudy was preoperative: 30 people rested only 30 minutes between tests. The original supplement reports 40-m fast-paced speed ICC 0.94 (95% CI 0.88–0.97), SEM 0.08 m/s and SDC 0.22 m/s. The SDC label should be retained; its magnitude is consistent with the conventional individual 95% formula, but the table does not separately label the confidence level. This is not direct evidence of postoperative between-day error. [3]

Kennedy's mixed sample contained 69 hips and 81 knees. Although 21 people contributed preoperative reliability observations, only 17 met the study's stability criterion based on LEFS. Median time from first to third assessment was 178 days. The reported 6MWT MDC90 of 61.34 m and fast-walk-time MDC90 of 4.04 s are useful historical benchmarks with that design attached. They are neither THA-only nor directly postoperative estimates. Dobson's 2017 mixed hip/knee OA study excluded prior replacements altogether. Its values should not be relabeled THA error. [2, 27]

The wide separation between 10.17 m and 61.34 m for apparently the same 6MWT is therefore not resolved by averaging the values. Different patients, stability intervals, protocols, statistical formulations and confidence levels may explain differences. A practical tool should show the matching source and its limitations, or state that an applicable threshold has not been established.

Table 2 Clinical error estimates with their evidence boundaries

Validity responsiveness and importance

Construct validity and recovery are not synonyms

Tolk's prespecified hypothesis analysis did not confirm construct validity or responsiveness of the OARSI core tests against the chosen PROM, pain and strength comparators. For the 40-m walk, 6 of 17 baseline hypotheses and 4 of 8 change hypotheses were confirmed; correlation between speed change and the global-change anchor was 0.28. This challenges broad claims that timed performance represents every aspect of physical functioning. It does not show that a stopwatch cannot measure time accurately. [3]

Interpret both sides of the disagreement. PROMs and observed capacity overlap imperfectly, so low correlations are not automatically measurement failure. Nevertheless, a failed prespecified validation exercise should not be dismissed merely because the test has face validity or is recommended. State the construct more narrowly and test appropriate hypotheses. Improvement in a mean after surgery, or a large standardized effect, is sensitivity to an observed treatment period; it is not alone proof of longitudinal validity against a relevant external criterion.

The recovered supplement also limits anchor interpretation. Of 77 global-change responses, 67 described much or very much improvement, five little improvement, one no change and four much worse. This distribution provides little information about the boundary between minimal improvement and stability. Supplementary wording refers to a knee operation despite a hip cohort, and describes an OHS scoring direction inconsistent with the displayed improvement. These reporting issues warrant care; they do not authorize inventing a corrected MIC or discarding all of the study's observations.

New acute important change evidence

De Leo and colleagues' 2026 study prevents a blanket statement that THA-specific MCID research is absent. The original abstract describes 100 inpatients tested on postoperative days 1 and 3 and a 15-point global rating of functional change after the second session. Estimates varied substantially across within-patient, between-patient, ROC and distribution methods: TUG 3.7–9.0 s, 10MWT 0.06–0.25 m/s and 2MWT 3.3–33.5 m. TUG discrimination was AUC 0.73 (95% CI 0.62–0.82), versus 0.64 and 0.55 for 10MWT and 2MWT. [8]

These ranges should not be collapsed into a universal preferred number. Full-text anchor grouping, anchor-change correlation, exact thresholds, uncertainty and test protocols were unavailable. Distribution-based numbers estimate statistical magnitude, not patient importance by themselves. Weak ROC discrimination further cautions against an individual meaningful-improvement badge for walking. An acute day-1-to-day-3 result also cannot establish important change at 6 months.

For repeated care, combine observed change with confidence in the measurement, symptom and task context, and the patient's own valued activities. If error and importance thresholds differ, explain the distinction rather than selecting the more convenient one. Absence of an applicable MIC is an uncertainty to communicate, not a reason to replace it with a knee/TKA or neurological threshold.

Table 3 Important change is not the same as error

Instrumented and technology supported assessment

Laboratory gait and variable specific reproducibility

Three-dimensional motion analysis adds joint and segment information unavailable from speed alone, but marker placement, anatomical models, soft-tissue artifact, event detection and walking speed influence results. Zügner's study examined 20 THA patients 1–2 years after surgery, 20 hip-OA patients and 20 controls. Two experienced observers reapplied markers within 2 hours; one randomly selected trial from six recorded trials was analyzed. Measurement scatter differed by variable and population. This is evidence about that observer/model design, not proof that all laboratory gait outputs are equally precise. [20]

Peak angle and range of motion can behave differently because a systematic angular offset affects the peak more than the excursion. An aid, a change in pace or an altered start can change true gait mechanics as well as error. For longitudinal interpretation, compare both the achieved speed and the biomechanical feature. Whether to adjust for speed depends on the causal question: speed can be a confounder, part of recovery itself, or a mediator. Reporting only a speed-adjusted result may remove a meaningful component of functional recovery.

Wearables IMUs and cameras

Bravi's mixed recent-THA/TKA study compared a single trunk IMU with optoelectronic analysis in 20 patients using crutches and 10 controls. Its original abstract reports substantially weaker correlations for stance/swing/support phases than for general spatiotemporal variables. Correlation alone is insufficient for interchangeability, but even those results argue against assigning one validation label to every sensor output. The hip/knee mix and crutch dependence also restrict THA-specific claims. [21]

A 2025 single-thigh-IMU study analyzed 18 patients preoperatively, 12 at 3 months and eight at 1 year, plus seven controls. The best multiple linear regression model had mean absolute error about 4.2 degrees and limits-of-agreement width 26.8 degrees for sagittal hip ROM. Critically, walking speed among its input features came from the optical reference system. Its classification labels were thirds of the observed ROM distribution, rather than established clinical severity thresholds. Internal grouped cross-validation does not establish external patient-level performance, and the described grouping of subject-side data deserves scrutiny before assuming that both sides of an individual were always independent of training data. [22]

Thus, a low sensor count does not necessarily mean a fully independent single-sensor workflow. Before implementation, validate every input that will be substituted, preserve subject-level separation across visits and limbs, and report absolute agreement across severity, aids and stages. A classifier trained on local thirds should not be displayed as normal versus abnormal hip function without an independent clinical interpretation.

Step-App provides another instructive distinction. Lippi's study assessed 42 patients, enrolled 28 and analyzed 25 completers, including 20 hip and five knee recipients; four procedures followed hip fracture. Measurements began around 12 days after surgery and compared app and clinician scoring during the same performances, before and after rehabilitation. The app's 10-m timing also required start/end commands. High same-trial walking correlations and exact observed chair-count agreement under supervision cannot be described as independent test-retest reliability or validated unsupervised home measurement in elective OA THA. [23]

The targeted literature did not establish a universal markerless-video criterion-validity or individual-change threshold for gait in primary THA. Group differences measured with a camera do not validate the camera. A future video pathway should specify camera position, frame rate, calibration, occlusion handling, clothing, aids, event definitions and rejection rules, and compare the actual claimed variable against an appropriate reference in the intended patients.

Free living trajectories

Fary and colleagues followed smartphone-derived gait metrics in 612 unilateral THA recipients. The denominators differed by output, and days contributing to weekly averages were particularly variable for asymmetry. The study defined recovery as a weekly group value no longer significantly inferior to the preoperative value under its multiple-testing threshold. Reported recovery weeks therefore depend on sample size, missingness, variance, baseline and the chosen significance rule. Failure to reject a difference is not equivalence, and return to an impaired preoperative level is not necessarily adequate function. [24]

These data are valuable for describing the shape and heterogeneity of observed recovery. The authors explicitly note that the smartphone gait algorithm had not been validated specifically in THA, and that aids or loose clothing could affect data capture. The trajectories do not supply an individual expected deadline or a validated complication alert. A dashboard should display valid recording days, eligible walking bouts, phone carriage and algorithm version. Near-normal clinic speed with low daily activity should prompt discussion about goals, opportunity and symptoms, rather than automatically indicating either successful participation or a technical error.

Table 4 Technology claims and validation gaps

Prognosis separate the endpoint and the horizon

Inpatient walking and length of stay

Shibuya's retrospective single-center cohort included 317 of 336 primary unilateral THA patients after excluding preoperative nonwalkers and perioperative complications. Preoperative usual and maximal speeds were timed over the middle 10 m of a 16-m course with necessary aids. The event was achieving 50 m without human assistance, allowing a walking aid. Both speeds retained adjusted associations with time to this event. Maximum-speed AUC for recovery by day 5 was 0.70, similar to a clinical model. [9]

This is genuinely temporally ordered evidence, but selected uncomplicated walkers and a local discharge pathway define its target. Speed tertiles were descriptive strata, not externally validated clinical cutoffs. Excluding postoperative complications from a model intended for preoperative planning is particularly important: at prediction time, those future exclusions are unknown. The model should not be advertised as forecasting recovery for all scheduled THA patients.

Oosting's 2016 prospective cohort supports the added prognostic information of performance testing for delayed inpatient independence; complete original methods were not recovered, so its published thresholds remain contextual rather than implementation-ready. In the 2021 cohort of 1,559 anterior-approach elective primary THAs, adding TUG improved the reported AUC from 0.75 to 0.77. A TUG cutoff of 9.7 s had 79% sensitivity but only 48% specificity for short stay. The authors themselves cautioned against using it as an absolute selection criterion. [10, 11]

Discharge is influenced by medical recovery, home support and service organization. A mobility model cannot determine medical fitness for discharge. Similarly, a statistically significant predictor may add little decision value if a base model already discriminates well. External validation should compare discrimination, calibration, net benefit and consequences of false reassurance or unnecessary prolonged admission.

Later walking and participation

Nankaku's retrospective study included 204 unilateral THA patients, classified at 6 months as independently ambulatory (118) or using a cane (86), based solely on self-report. Its original abstract reports a preoperative TUG cutoff of 10 s, sensitivity 76.7%, specificity 93.2% and AUC 0.93. This is a potentially useful development finding, but it predicts reported aid use in that cohort, rather than all aspects of independent community mobility. Stepwise selection and an in-sample cutoff warrant optimism concerns, and the full methods and external validation were unavailable. [12]

Kamimura studied 48 Japanese women with severe unilateral OA and asymptomatic contralateral hips; full weight bearing began only on day 14. Its TUG-based good/nongood classification borrowed 13.5 s from an older-adult falls source. The study did not validate that value against future falls after THA. Different preoperative factors were selected at different follow-up times, with small samples and stepwise model development. [13]

Kawano's retrospective 321-person clinical prediction rule used preoperative characteristics and a 3-week 10MWT to classify 1-year UCLA activity. This is a later self-reported activity outcome, not objectively measured habitual walking. Its abstract reports high in-sample probabilities when several dichotomized factors are present; without full model appraisal and external validation, these should not become personalized probability estimates. [19]

A 2025 study related discharge gait to 2-year Forgotten Joint Score. Only 44 patients contributed final data from 313 surgical patients, and the analysis was restricted to women able to meet the study's postoperative walking conditions. Its later outcome concerns joint awareness, not directly falls, work or endurance. Variable selection and ROC thresholds in a small responder subset make this hypothesis-generating longitudinal evidence. [18]

The explainable-machine-learning study of preoperative gait subpopulations and postoperative gait classification demonstrates heterogeneity and potential research utility. Training classifiers to distinguish current gait patterns, then describing postoperative shifts, is not equivalent to validating a model that predicts an unseen patient's future participation. Similarly, the 2026 cumulative-loading biomarker abstract does not establish enough detail about predictor timing to treat its associations with 26-week function as proven prospective prognosis. Both warrant further appraisal before clinical deployment. [28, 29]

Falls

Ikutomo's prospective report assessed gait abnormality at 3 weeks and observed falls during the postoperative year. Of a 286-patient inception cohort, 162 women were included; the abstract reports a falls incidence of 31.5% and adjusted hazard ratio 2.91 (95% CI 1.55–5.48) for gait abnormality. This supports gait observation as a candidate risk marker in the studied group, but missing original details of the abnormality definition, selection and ascertainment preclude reproducing the rule here. The abstract does not explain how falls before the three-week assessment were handled, so the exact alignment of index time and risk window still requires confirmation. The hazard ratio is not an individual's probability of falling. [14]

In Ninomiya's prospective cohort, 157 patients were analyzed and 32 fell. Preoperative maximum speed was virtually the same in eventual fallers and nonfallers, and one-leg stance did not distinguish them. Hip abductor strength and previous falls were more informative. Its selection for short-stay home discharge and predominance of women matter, as do limitations of reported odds-ratio coding. These findings do not contradict a postoperative gait-abnormality association: the index task, timing and cohort differ. [15]

The Australian elective-THR cohort followed 167 adults aged at least 60 using calendars and monthly calls and recorded 67 fallers (reported as 42%; the abstract does not explain the percentage denominator). This is stronger prospective ascertainment than asking patients once to remember the preceding year; it does not establish a particular speed cutoff. Its results also show why seemingly good surgical recovery should not be equated with absence of falls risk. [16]

By contrast, Lin and colleagues assessed 12 THA patients around 1 year after surgery and related current strength, fear of falling and treadmill gait to recalled falls in the preceding year. The reported high explained variance is contemporaneous/retrospective association with serious small-sample overfitting risk. It is not prospective evidence that a new assessment predicts the next year's falls, and its lack of gait correlation is not proof that gait never matters. [17]

Table 5 Prognostic evidence must retain its outcome and horizon

Practical implementation and research priorities

Start with an explicit question and one reproducible protocol. Record exact timing from surgery, procedure/indication, precautions, pain, fatigue, aid, assistance, footwear, course and trial aggregation. Keep noncompletion and its reason. Add sustained walking when it answers a separate question, and add instrumented assessment only for outputs that may change interpretation or treatment planning.

Show absolute scores and change before any category. Attach an error estimate only when the population, stage, protocol and confidence level reasonably match; label it as reported or derived and distinguish within-session from between-day evidence. Do not display an MDC as an MIC. A protocol change should be visible rather than hidden inside a trend line.

For movement analysis, preserve speed, side and source of measurement. A symmetry ratio cannot replace the two limb values, and an apparently normal ratio can arise from bilateral limitation. Report device agreement separately from reproducibility and clinically important change. Use clinically meaningful quality-control failures instead of forcing an estimate from every recording.

For prognosis, require a defined future endpoint and independent validation. A useful development program would prospectively test protocol-specific gait measures across modern pathways, retain complications and nonwalkers, collect repeated falls with calendars, and measure participation rather than infer it from clinic capacity. Models should assess calibration across sex, age, bilateral disease, aids and surgical approach, and evaluate benefit relative to simpler clinical information.

The overall conclusion is that gait testing after THA is already useful for transparent assessment. The evidence is strongest when claims remain close to the observed task and source population. Precision about stage, protocol and temporal ordering is more valuable than a universal threshold assembled from heterogeneous studies.

Primary study characteristics

Original studies and explicitly labeled contextual sources are grouped by question. Population, postoperative stage, protocol, endpoint, source access and analysis design constrain interpretation. Publications from the same cohort are not independent replications.

Framework and contextual evidence

Dobson F 2013

Study and population [1] Consensus framework. International consensus for hip/knee OA and joint replacement

Protocol and timing Activity-based test selection using evidence, feasibility and expert consensus

Principal findings 30-second chair stand, 40-m fast-paced walk and stair climb form the minimum set; TUG and 6MWT are additional

Interpretive limits Consensus does not supply universal postoperative error, importance or prediction thresholds

Source examined Original article body retrieved

Dobson F 2017

Study and population [27] Direct measurement evidence; nonoperative boundary. 51 stable hip and/or knee OA participants; prior replacement excluded

Protocol and timing Independent raters within a session and same-rater repeat about a week later; standardized OARSI tests

Principal findings Useful nonoperative evidence for standardization and reproducibility

Interpretive limits Mixed joints and nonoperative stage; numerical original tables incompletely extracted; cannot be relabeled postoperative THA

Source examined Original article body retrieved; numerical original tables incomplete in extraction; nonoperative mixed OA context

Clinical measurement properties and important change

Kennedy DM 2005

Study and population [2] Direct measurement evidence; mixed-joint recovery context. 150 primary unilateral OA arthroplasties: 69 hips and 81 knees; 17 stable preoperative participants in reliability analysis

Protocol and timing Fast 2 × 20 m with turn excluded; 46-m rectangular 6MWT; regular aids; preoperative retests over long waiting-list intervals

Principal findings 6MWT ICC 0.94 and MDC90 61.34 m; fast-walk time ICC 0.91 and MDC90 4.04 s

Interpretive limits Mixed joints, not direct postoperative error; median 178-day first-to-third interval; 33 of 119 eligible patients could not complete the early 6MWT

Source examined Original article body retrieved

Tolk JJ 2019

Study and population [3] Direct measurement evidence; hypothesis-based validity/responsiveness. 90 unilateral symptomatic hip-OA candidates for primary THA; reliability subset n = 30; follow-up at 12 months

Protocol and timing Preoperative retest after 30 minutes; 43-cm chair; standardized OARSI tasks, with walking aids recorded

Principal findings 30CST ICC 0.86, SEM 0.99, SDC 2.7 stands; walk ICC 0.94, SEM 0.08 m/s, SDC 0.22 m/s; validity/change hypotheses not confirmed

Interpretive limits SDC confidence level not separately labeled; chair retest improved 0.8 stands; anchor responses concentrated in major improvement; supplementary wording/scoring discrepancies

Source examined Original article body and original supplementary PDF retrieved; supplementary Tables 2–4 visually checked

Unver B 2015

Study and population [4] Direct measurement evidence. 37 primary THA recipients, mean age 54.5 years; at least 1 year and mean 5.3 years postoperative; no aids

Protocol and timing 17-inch chair, arms crossed, practice stand; two occasions 1 hour apart; walking methods and appendix disagree about turning

Principal findings 30CST ICC 0.94, SEM 0.4 repetitions, SRD95 1.2 repetitions; 50FWT ICC 0.98, SEM 0.3 s, SRD95 0.8 s

Interpretive limits Selected stable late-stage sample; same-day evidence; no MIC; SEM derived from error mean square; walking course ambiguity unresolved

Source examined Original publisher PDF retrieved; Table II and appendix visually checked

Temporiti F 2024

Study and population [5] Direct measurement evidence. 35 primary unilateral OA THA patients ready for discharge on day 2; posterolateral approach; unrestricted weight bearing

Protocol and timing Middle 10 m of 14-m course with crutches, mean of two trials; TUG without crutches, best of two; same assessor after 1 hour

Principal findings Maximum 10MWT MDC95 0.13 m/s indoors and 0.16 outdoors; usual-speed MDC95 0.16 m/s in both; TUG 2.5 and 2.4 s

Interpretive limits Selected discharge-ready cohort; indoor usual-speed ICC CI 0.13–0.96; retest improvement; inconsistent demographic totals; individual MDC is not a test of a group mean difference

Source examined Original article body retrieved

Yuksel E 2021

Study and population [6] Direct measurement evidence, abstract-level. 37 primary THA recipients; exact stage not verified from the original full text

Protocol and timing Two same-day TUG and 2MWT occasions; detailed corridor, aid and aggregation rules unavailable

Principal findings ICC 0.96 for both; TUG SEM 0.59 s and MDC95 1.62 s; 2MWT SEM 6.37 m and MDC95 17.56 m

Interpretive limits Abstract-only numerical evidence; no MIC or validated future-outcome rule follows from reliability

Source examined Original abstract only; full article not retrieved

Unver B 2013

Study and population [7] Direct measurement evidence, abstract-level. 34 THA recipients; exact postoperative stage unverified

Protocol and timing Two 6MWTs on the same day separated by 1 hour of seated rest

Principal findings ICC 0.96; SEM 3.67 m; SRD95 10.17 m; second trial 3.71 m farther

Interpretive limits Full methods needed for corridor and selection; not interchangeable with Kennedy's mixed-joint, long-interval estimate

Source examined Original abstract only; full article not retrieved

De Leo D 2026

Study and population [8] Important-change study, abstract-level. 100 inpatients tested on postoperative days 1 and 3

Protocol and timing TUG, 10MWT and 2MWT; 15-point global change rating; anchor-based and distribution-based methods

Principal findings Estimate ranges: TUG 3.7–9.0 s, 10MWT 0.06–0.25 m/s, 2MWT 3.3–33.5 m; ROC AUC 0.73, 0.64 and 0.55 respectively

Interpretive limits Full anchor definitions, correlations and exact thresholds unavailable; walking discrimination weak; no universal MCID can be selected from the range

Source examined Original abstract only; full article not retrieved

Alalem N 2025

Study and population [22] Concurrent model-based measurement validation. 18 patients before THA, 12 at 3 months, eight at 1 year and seven controls

Protocol and timing Thigh IMU versus optical hip ROM; internally grouped cross-validation; best-agreement multiple linear regression included optically measured walking speed

Principal findings Best-agreement multiple linear regression model: MAE about 4.2 degrees and limits-of-agreement width 26.8 degrees. Random forest had lower MAE (3.2 degrees), so model superiority depends on the error metric

Interpretive limits Not a fully independent single-sensor workflow; attrition; distribution-derived classes; participant-side grouping requires scrutiny for generalization

Source examined Original article body retrieved

Instrumented assessment and recovery

Zügner R 2018

Study and population [20] Direct instrumented reproducibility and group comparison. 20 THA patients at 1–2 years, 20 hip-OA patients and 20 controls; mixed surgical approaches

Protocol and timing Two experienced observers replaced markers within 2 hours; six trials recorded and one randomly selected; modified Helen Hayes model

Principal findings Observer scatter varied by output and population; peak angles and ROM differed in reproducibility

Interpretive limits Same-session observer study, not between-day MDC; healthy control reliability alone is insufficient

Source examined Original article body retrieved

Bravi M 2020

Study and population [21] Concurrent device comparison, abstract-level. 20 recent THA/TKA patients using crutches and 10 healthy controls

Protocol and timing Single trunk IMU compared with optical motion capture during five consecutive walks

Principal findings Patient correlations were 0.704–0.986 for general spatiotemporal measures and 0.077–0.464 for gait phases

Interpretive limits Correlation is not agreement; mixed surgery; crutches affect phase detection; detailed full methods unavailable

Source examined Original abstract only; full article not retrieved

Lippi L 2024

Study and population [23] Concurrent agreement/rehabilitation context. 25 of 42 assessed participants; 20 hip and five knee recipients; four fracture indications; baseline about 12 days postoperative

Protocol and timing Simultaneous clinician/app scoring; thigh sensor for 30CST; 43.2-cm chair raised with cushions when needed; four-week rehabilitation interval

Principal findings High same-trial supervised walking correlations; perfect observed chair-rise count agreement in Figure 8. Walking Figures 5–7 show systematic offsets and nonzero limits of agreement; plotted units and clinical tolerances require clarification before claiming interchangeability

Interpretive limits Mixed procedures and indications; no unsupervised home or between-day validation; modified chair height; agreement figures now inspected; walking-unit interpretation and clinical tolerances remain unresolved

Source examined Original article body retrieved; agreement figures now visually inspected; walking-unit interpretation remains unresolved; mixed procedures/indications

Fary C 2023

Study and population [24] Longitudinal recovery description. 612 unilateral THA recipients with at least 24-week follow-up; output-specific denominators 582–605

Protocol and timing Passive smartphone gait metrics summarized by week against an averaged preoperative baseline

Principal findings Gait metrics returned to their preoperative group levels at different times

Interpretive limits Recovery defined by nonsignificance, not equivalence or individual importance; variable valid days; algorithm not validated specifically in THA; aids, carriage and clothing can affect capture

Source examined Original article body retrieved

Jeldi AJ 2017

Study and population [25] Longitudinal capacity/performance context, abstract-level. 30 THA recipients, mean age 67 years; preoperative, 3-month and 12-month observations

Protocol and timing Objectively measured daily activity with 6MWT, Harris Hip Score and Oxford Hip Score

Principal findings Pain and function improved without corresponding increases in activity-volume measures

Interpretive limits Capacity and everyday behavior differ; original full article unavailable; no chair-rise prognostic rule

Source examined Original abstract only; full article not retrieved

Stetter BJ 2025

Study and population [28] Exploratory classification and longitudinal subgroup description. 109 preoperative hip-OA participants with a postoperative subset and healthy controls; recruitment over 10 years

Protocol and timing 18 waveforms, PCA/clustering and internally cross-validated SVM classification; postoperative projection into preoperative models

Principal findings Distinct gait subpopulations with differing postoperative patterns

Interpretive limits Exploratory current-pattern classification and subgroup trajectories, not externally validated prediction of future individual clinical outcomes

Source examined Original article body retrieved

Gasparutto X 2021

Study and population [26] Instrumented recovery context. 71 primary hip-OA THA patients and 52 controls; preoperative and 6-month testing; important bilateral/other disease excluded

Protocol and timing 47-cm armchair; support allowed; self-selected TUG; fastest of first three trials; optical phase partition

Principal findings Different component tasks showed different deficits and recovery, with substantial individual variability

Interpretive limits Composite TUG is not pure STS; no phase-specific MIC, MDC or prospective prediction validation

Source examined Original article body retrieved

Prognostic questions and temporal ordering

Shibuya M 2020

Study and population [9] Development prognostic association. 317 of 336 primary unilateral THA recipients; 11 preoperative nonwalkers and eight postoperative complications excluded

Protocol and timing Preoperative usual/maximum speed over middle 10 m of 16-m course; aids permitted; later 50 m without human assistance

Principal findings Maximum-speed AUC 0.70 (0.64–0.76) for day-5 recovery; both speed measures retained adjusted time-to-event associations

Interpretive limits Retrospective single center; future-complication exclusions; local mean stay 13.5 days; no external validation or universal tertile thresholds

Source examined Original article body retrieved

Oosting E 2016

Study and population [10] Development prognostic cohort, abstract-level. 315 THA recipients in a prospective inpatient recovery cohort

Protocol and timing Preoperative performance tests with conventional and RAPT variables; later walking-independence endpoint

Principal findings Performance measures added prognostic information for local inpatient recovery

Interpretive limits Full original source unavailable; protocol-specific cutoffs not implementation-ready; discharge and functional independence are distinct; abstract implications reverse the speed inequality relative to its results, so the threshold is not copied as an implementation rule

Source examined Original abstract only; full article not retrieved

Oosting E 2021

Study and population [11] Development prognostic model. 1,559 elective primary anterior THAs, retrospectively analyzed; 44 TUG values missing

Protocol and timing Preoperative TUG; endpoint discharge within 36 hours in one institutional pathway

Principal findings Base-model AUC 0.75, increasing to 0.77 with TUG; 9.7-s cutoff sensitivity 79%, specificity 48%

Interpretive limits No external validation; home support related but not included in the model; cutoff unsuitable as an absolute discharge exclusion

Source examined Original article body retrieved

Nankaku M 2013

Study and population [12] Development prognostic cohort, abstract-level. 204 unilateral THA recipients; 118 independent and 86 cane-assisted at 6 months, based solely on self-report

Protocol and timing Retrospective cohort; preoperative TUG, hip/knee strength and clinical covariates; stepwise logistic model and in-sample ROC cutoff

Principal findings Original abstract: TUG cutoff 10 s, sensitivity 76.7%, specificity 93.2%, AUC 0.93 for self-reported ambulatory category

Interpretive limits Abstract-only methods; selected outcome is reported cane use, not all community mobility; model/cutoff optimism and external validation remain unverified

Source examined Original abstract only; full article not retrieved

Kamimura A 2014

Study and population [13] Development prognostic cohort. 48 Japanese women with primary unilateral OA THA and asymptomatic opposite hip; posterolateral approach; full weight bearing from day 14

Protocol and timing Preoperative HHD, mean of three efforts; 45-cm armless-chair TUG; follow-up at 3 weeks, 4 months and 7 months

Principal findings Selected factors differed by horizon; development cutoffs included 0.56 Nm/kg knee extension and 0.24 Nm/kg hip abduction

Interpretive limits 13.5-s TUG boundary borrowed to define ambulation status, not observed falls; small stepwise model; unusual reported confidence intervals; no general risk calculator

Source examined Original article body retrieved

Ikutomo H 2018

Study and population [14] Prospective falls association, abstract-level. 162 women included from 286 THA patients with severe hip OA

Protocol and timing Gait abnormality at 3 weeks; subsequent falls during the postoperative year; Cox model

Principal findings 31.5% fell; gait abnormality HR 2.91 (1.55–5.48)

Interpretive limits Original abnormality definition, selection, ascertainment and handling of falls before the week-3 index assessment unavailable; HR is not absolute risk

Source examined Original abstract only; full article not retrieved

Ninomiya K 2020

Study and population [15] Prospective falls cohort. 157 primary unilateral OA THA recipients, including 142 women; selected for home discharge within 5 days

Protocol and timing Preoperative maximum 10-m walk with runways, best of two; capped one-leg stance; prospective falls questionnaires

Principal findings 32 fell; maximum speed 1.19 versus 1.21 m/s in fallers/nonfallers, p = 0.803; stance also nondiscriminating

Interpretive limits Selected short-stay sample; strength/prior falls more informative; OR coding/scaling unclear; strength cutoff specificity only 50%

Source examined Original article body retrieved

Hill AM 2021

Study and population [16] Prospective falls cohort, abstract-level. 167 adults aged at least 60 after elective primary THR in Perth; 54.4% women

Protocol and timing Predischarge baseline; 12-month falls calendars plus monthly calls

Principal findings 67 participants fell (reported as 42%, with the abstract percentage denominator unclear); 140 falls and nine fractures

Interpretive limits Strong ascertainment context, not validation of a particular gait or chair-rise threshold; original full article unavailable

Source examined Original abstract only; full article not retrieved

Lin X 2022

Study and population [17] Contemporaneous/retrospective association. 12 THA patients at mean 12.4 months; direct lateral approach; OA-faller and healthy comparison groups

Protocol and timing Current treadmill gait, strength and fall efficacy related to recalled preceding-year falls

Principal findings Eight THA participants reported falls; no gait parameter correlated; high in-sample strength/fall-efficacy explained variance

Interpretive limits Retrospective/contemporaneous, not future prognosis; tiny selected sample, variable selection, recall and reverse-causation risks

Source examined Original article body retrieved

Okazawa K 2025

Study and population [18] Longitudinal association/development discrimination. 44 women in final analysis from 313 surgical patients; 65 initially measured; posterolateral THA

Protocol and timing Discharge gait at 13.8 ± 3.6 days; mailed 2-year Forgotten Joint Score; pressure walkway, four trials

Principal findings Early gait related to later joint awareness in selected responders

Interpretive limits Only 14% of surgical cohort in final analysis; stepwise/ROC development; outcome is not falls or participation

Source examined Original article body retrieved

Kawano T 2022

Study and population [19] Development clinical prediction rule, abstract-level. 321 primary THA recipients, including 56 men; retrospective development cohort

Protocol and timing Preoperative and 3-week tests; 1-year UCLA activity category

Principal findings Final rule used age, prior activity, hip/knee strength and postoperative 10MWT; STS was not retained

Interpretive limits Self-reported activity; cutoffs and high in-sample probabilities not externally validated here; full methods unavailable

Source examined Original abstract only; full article not retrieved

Christensen JC 2026

Study and population [29] Association with later-recovery outcomes; temporal status unverified. 56 primary unilateral THA recipients; functional outcomes assessed at 26 weeks

Protocol and timing Cumulative loading measures multiplied by daily steps and related to 6MWT, 30STS, WOMAC and strength

Principal findings Cumulative loading rate associated with several functional outcomes

Interpretive limits Abstract insufficient to verify predictor timing; STS is an outcome, not a validated predictor; full article unavailable

Source examined Original abstract only; full article not retrieved

Search and source access appendix

Search coverage

Across the three overlapping streams, there were 2,989 unique official PubMed records and 4,173 unique Scopus records. These are database-specific retrieval counts, not unique cross-database publications or eligible studies. They must not be added as independent evidence.

The exact dated PubMed queries were reconciled against official ESearch identifiers and complete EFetch records, including nonstandard record types. Scopus retrieval reached terminal coverage, with unique identifiers matching each reported total.

The searches were intentionally broad. Surgical mechanics, treatment comparisons, revision/fracture procedures, unrelated acronyms and off-domain papers were retained during retrieval and separated during targeted appraisal. Selection for intensive appraisal favored direct measurement properties, important-change anchors, true temporal outcomes, contradictory findings and implementation-relevant technology. Search retrieval is more extensive than full-text appraisal.

PubMed used an explicit publication-date range through 2026/10/02. Scopus used PUBYEAR BEF 2027; year-based constraints and indexing lags can miss articles first available online in 2026 but assigned to a later issue. Targeted source searching and citation discovery reduce, but do not eliminate, such gaps. No claim of complete coverage of all publications is made.

Supplemental discovery

Bounded reference and forward-citation searches used the Tolk hip-OA measurement study, the 2021 Wang sit-to-stand review, the 2021 Labanca balance review, the 2021 abductor-strength review, the 2020 Shibuya prognosis study and the 2016 THA prognosis review. Each request was capped at 100 records per provider (OpenAlex/Semantic Scholar). Capped citation networks are not exhaustive. Targeted official-source web searches identified or checked recent important-change, wearable, falls, chair-rise and biomechanical reports.

Exact native queries

Gait measurement PubMed

("Arthroplasty, Replacement, Hip"[MeSH Terms] OR "total hip arthroplast*"[Title/Abstract] OR "total hip replacement*"[Title/Abstract]) AND ("Gait"[MeSH Terms] OR "Walking"[MeSH Terms] OR (gait[Title/Abstract] OR walk*[Title/Abstract] OR ambulat*[Title/Abstract] OR locomot*[Title/Abstract] OR "six minute"[Title/Abstract] OR "6 minute"[Title/Abstract] OR 6MWT[Title/Abstract] OR 10MWT[Title/Abstract] OR "40 m"[Title/Abstract] OR "40 meter"[Title/Abstract] OR "40 metre"[Title/Abstract])) AND (reliab*[Title/Abstract] OR valid*[Title/Abstract] OR reproducib*[Title/Abstract] OR psychometr*[Title/Abstract] OR clinimetr*[Title/Abstract] OR agreement[Title/Abstract] OR "measurement error"[Title/Abstract] OR "standard error"[Title/Abstract] OR "minimal detectable"[Title/Abstract] OR "minimum detectable"[Title/Abstract] OR "smallest detectable"[Title/Abstract] OR "minimal important"[Title/Abstract] OR "minimally important"[Title/Abstract] OR "minimum important"[Title/Abstract] OR responsiv*[Title/Abstract] OR interpretabil*[Title/Abstract] OR "floor effect"[Title/Abstract] OR "ceiling effect"[Title/Abstract]) AND ("1800/01/01"[Date - Publication] : "2026/10/02"[Date - Publication])

Gait measurement Scopus

TITLE-ABS-KEY(("total hip arthroplast*" OR "total hip replacement*" OR THA OR THR) AND (gait OR walk* OR ambulat* OR locomot* OR "six minute" OR "6 minute" OR 6MWT OR 10MWT OR "40 m" OR "40 meter" OR "40 metre") AND (reliab* OR valid* OR reproducib* OR psychometr* OR clinimetr* OR agreement OR "measurement error" OR "standard error" OR "minimal detectable" OR "minimum detectable" OR "smallest detectable" OR "minimal important" OR "minimally important" OR "minimum important" OR responsiv* OR interpretabil* OR "floor effect" OR "ceiling effect")) AND PUBYEAR BEF 2027

Gait prognosis PubMed

("Arthroplasty, Replacement, Hip"[MeSH Terms] OR "total hip arthroplast*"[Title/Abstract] OR "total hip replacement*"[Title/Abstract]) AND ("Gait"[MeSH Terms] OR "Walking"[MeSH Terms] OR (gait[Title/Abstract] OR walk*[Title/Abstract] OR ambulat*[Title/Abstract] OR locomot*[Title/Abstract] OR "six minute"[Title/Abstract] OR "6 minute"[Title/Abstract] OR 6MWT[Title/Abstract] OR 10MWT[Title/Abstract] OR "40 m"[Title/Abstract] OR "40 meter"[Title/Abstract] OR "40 metre"[Title/Abstract])) AND (prognos*[Title/Abstract] OR predict*[Title/Abstract] OR longitudinal[Title/Abstract] OR prospective[Title/Abstract] OR cohort[Title/Abstract] OR "follow up"[Title/Abstract] OR "follow-up"[Title/Abstract] OR recovery[Title/Abstract] OR deteriorat*[Title/Abstract] OR fall*[Title/Abstract] OR "natural history"[Title/Abstract] OR "return to work"[Title/Abstract] OR discharge[Title/Abstract] OR "risk factor"[Title/Abstract]) AND ("1800/01/01"[Date - Publication] : "2026/10/02"[Date - Publication])

Gait prognosis Scopus

TITLE-ABS-KEY(("total hip arthroplast*" OR "total hip replacement*" OR THA OR THR) AND (gait OR walk* OR ambulat* OR locomot* OR "six minute" OR "6 minute" OR 6MWT OR 10MWT OR "40 m" OR "40 meter" OR "40 metre") AND (prognos* OR predict* OR longitudinal OR prospective OR cohort OR "follow up" OR "follow-up" OR recovery OR deteriorat* OR fall* OR "natural history" OR "return to work" OR discharge OR "risk factor")) AND PUBYEAR BEF 2027

Gait protocol mechanism technology PubMed

("Arthroplasty, Replacement, Hip"[MeSH Terms] OR "total hip arthroplast*"[Title/Abstract] OR "total hip replacement*"[Title/Abstract]) AND ("Gait"[MeSH Terms] OR "Walking"[MeSH Terms] OR (gait[Title/Abstract] OR walk*[Title/Abstract] OR ambulat*[Title/Abstract] OR locomot*[Title/Abstract] OR "six minute"[Title/Abstract] OR "6 minute"[Title/Abstract] OR 6MWT[Title/Abstract] OR 10MWT[Title/Abstract] OR "40 m"[Title/Abstract] OR "40 meter"[Title/Abstract] OR "40 metre"[Title/Abstract])) AND (protocol[Title/Abstract] OR biomechan*[Title/Abstract] OR kinematic*[Title/Abstract] OR kinetic*[Title/Abstract] OR symmetr*[Title/Abstract] OR asymmetr*[Title/Abstract] OR "weight bearing"[Title/Abstract] OR "weight-bearing"[Title/Abstract] OR "ground reaction"[Title/Abstract] OR "force plate"[Title/Abstract] OR "force platform"[Title/Abstract] OR sensor*[Title/Abstract] OR wearable*[Title/Abstract] OR inertial[Title/Abstract] OR acceleromet*[Title/Abstract] OR markerless[Title/Abstract] OR "motion capture"[Title/Abstract] OR camera[Title/Abstract] OR video[Title/Abstract] OR algorithm[Title/Abstract] OR electromyogra*[Title/Abstract] OR activation[Title/Abstract] OR "sampling frequency"[Title/Abstract] OR "filter cutoff"[Title/Abstract]) AND ("1800/01/01"[Date - Publication] : "2026/10/02"[Date - Publication])

Gait protocol mechanism technology Scopus

TITLE-ABS-KEY(("total hip arthroplast*" OR "total hip replacement*" OR THA OR THR) AND (gait OR walk* OR ambulat* OR locomot* OR "six minute" OR "6 minute" OR 6MWT OR 10MWT OR "40 m" OR "40 meter" OR "40 metre") AND (protocol OR biomechan* OR kinematic* OR kinetic* OR symmetr* OR asymmetr* OR "weight bearing" OR "weight-bearing" OR "ground reaction" OR "force plate" OR "force platform" OR sensor* OR wearable* OR inertial OR acceleromet* OR markerless OR "motion capture" OR camera OR video OR algorithm OR electromyogra* OR activation OR "sampling frequency" OR "filter cutoff")) AND PUBYEAR BEF 2027

Source access and appraisal

Exclusions and related evidence

Primary elective OA THA is the principal inference population. Fracture/hemiarthroplasty, revision, resurfacing, bilateral THA and mixed hip/knee samples are excluded from unqualified primary-THA conclusions or explicitly labeled contextual evidence. The 2025 multidisciplinary preoperative 30STS/TUG article, DOI 10.3390/jcm14041085, was retrieved and excluded from THA inference because its sample was TKA-only. Stand-to-sit is not silently relabeled sit-to-stand. Companion Delaware biomechanics/feedback reports are not counted as independent replication. Nonoperative hip/knee OA reliability is kept separate from postoperative evidence.

The primary study appendix records evidence class, population, protocol, principal findings, limitations and source access. Unavailable full text is not an exclusion based on an assumed result. This appendix is an audit of the work performed, not a PRISMA flow diagram or a registered systematic-review protocol.

References

References are numbered in first citation order. Source descriptions identify the material examined and do not constitute a study quality rating. Each link identifies the original publication or explicitly named primary source version.

1. Dobson F, Hinman RS, Roos EM, Abbott JH, Stratford P, Davis AM, et al. OARSI recommended performance-based tests to assess physical function in people diagnosed with hip or knee osteoarthritis. Osteoarthritis and cartilage. 2013;21(8):1042-52. DOI 10.1016/j.joca.2013.05.002 Source examined: Original article body retrieved.

Source note: SRC-d7dd25666a3f Dobson F 2013

2. Kennedy DM, Stratford PW, Wessel J, Gollish JD, Penney D. Assessing stability and change of four performance measures: a longitudinal study evaluating outcome following total hip and knee arthroplasty. BMC musculoskeletal disorders. 2005;6:3. DOI 10.1186/1471-2474-6-3 Source examined: Original article body retrieved.

Source note: SRC-002a53da5f5c Kennedy DM 2005

3. Tolk JJ, Janssen RPA, Prinsen CSAC, van der Steen MMC, Bierma Zeinstra SMA, Reijman M. Measurement properties of the OARSI core set of performance-based measures for hip osteoarthritis: a prospective cohort study on reliability, construct validity and responsiveness in 90 hip osteo-arthritis patients. Acta orthopaedica. 2019;90(1):15-20. DOI 10.1080/17453674.2018.1539567 Source examined: Original article body and original supplementary PDF retrieved; supplementary Tables 2–4 visually checked.

Source note: SRC-daf9772081ef Tolk JJ 2019

4. Unver B, Kahraman T, Kalkan S, Yuksel E, Karatosun V, Gunal I. Test-retest reliability of the 50-foot timed walk and 30-second chair stand test in patients with total hip arthroplasty. Acta orthopaedica Belgica. 2015;81(3):435-41. PubMed. Source examined: Original publisher PDF retrieved; Table II and appendix visually checked.

Source note: SRC-cf3ec549527c Unver B 2015

5. Temporiti F, Casirati C, Adamo P, Leo D, Marino G, Grappiolo G, et al. Indoor and outdoor 10-Meter Walk Test and Timed Up and Go in patients after total hip arthroplasty: a reliability and comparative study. Archives of physiotherapy. 2024;14:90-95. DOI 10.33393/aop.2024.3267 Source examined: Original article body retrieved.

Source note: SRC-62b6db62b834 Temporiti F 2024

6. Yuksel E, Unver B, Kalkan S, Karatosun V. Reliability and minimal detectable change of the 2-minute walk test and Timed Up and Go test in patients with total hip arthroplasty. Hip international : the journal of clinical and experimental research on hip pathology and therapy. 2021;31(1):43-49. DOI 10.1177/1120700019888614 Source examined: Original abstract only; full article not retrieved.

Source note: SRC-251dc48f5ff7 Yuksel E 2021

7. Unver B, Kahraman T, Kalkan S, Yuksel E, Karatosun V. Reliability of the six-minute walk test after total hip arthroplasty. Hip international : the journal of clinical and experimental research on hip pathology and therapy. 2013;23(6):541-5. DOI 10.5301/hipint.5000073 Source examined: Original abstract only; full article not retrieved.

Source note: SRC-8134a6da765f Unver B 2013

8. De Leo D, Temporiti F, Guzzetti L, Crimeni E, Grappiolo G, Gatti R. Minimal Clinically Important Difference of the Timed Up and Go Test, 10-Meter Walk Test, and 2-Minute Walk Test in Patients in the Acute Phase After Total Hip Arthroplasty. The Journal of arthroplasty. 2026. DOI 10.1016/j.arth.2026.06.062 Source examined: Original abstract only; full article not retrieved.

Source note: SRC-488ad489687f De Leo D 2026

9. Shibuya M, Nanri Y, Kamiya K, Fukushima K, Uchiyama K, Takahira N, et al. The maximal gait speed is a simple and useful prognostic indicator for functional recovery after total hip arthroplasty. BMC musculoskeletal disorders. 2020;21(1):84. DOI 10.1186/s12891-020-3093-z Source examined: Original article body retrieved.

Source note: SRC-9851d96893a5 Shibuya M 2020

10. Oosting E, Hoogeboom TJ, Appelman-de Vries SA, Swets A, Dronkers JJ, van Meeteren NL. Preoperative prediction of inpatient recovery of function after total hip arthroplasty using performance-based tests: a prospective cohort study. Disability and rehabilitation. 2016;38(13):1243-9. DOI 10.3109/09638288.2015.1076074 Source examined: Original abstract only; full article not retrieved.

Source note: SRC-fe647c46204d Oosting E 2016

11. Oosting E, Kapitein PJC, de Vries SV, Breedveld E. Predicting short stay total hip arthroplasty by use of the timed up and go-test. BMC musculoskeletal disorders. 2021;22(1):361. DOI 10.1186/s12891-021-04240-6 Source examined: Original article body retrieved.

Source note: SRC-5937dca2cf74 Oosting E 2021

12. Nankaku M, Tsuboyama T, Akiyama H, Kakinoki R, Fujita Y, Nishimura J, et al. Preoperative prediction of ambulatory status at 6 months after total hip arthroplasty. Physical therapy. 2013;93(1):88-93. DOI 10.2522/ptj.20120016 Source examined: Original abstract only; full article not retrieved.

Source note: SRC-0279313426ba Nankaku M 2013

13. Kamimura A, Sakakima H, Tsutsumi F, Sunahara N. Preoperative predictors of ambulation ability at different time points after total hip arthroplasty in patients with osteoarthritis. Rehabilitation research and practice. 2014;2014:861268. DOI 10.1155/2014/861268 Source examined: Original article body retrieved.

Source note: SRC-c61268adeafb Kamimura A 2014

14. Ikutomo H, Nagai K, Tagomori K, Miura N, Nakagawa N, Masuhara K. Gait Abnormality Predicts Falls in Women After Total Hip Arthroplasty. The Journal of arthroplasty. 2018;33(10):3215-3219. DOI 10.1016/j.arth.2018.05.044 Source examined: Original abstract only; full article not retrieved.

Source note: SRC-8b3544eb6f13 Ikutomo H 2018

15. Ninomiya K, Takahira N, Ikeda T, Suzuki K, Sato R, Hirakawa K. Predictors of falls in patients during the first year after total hip arthroplasty: A prospective cohort study. Health science reports. 2020;3(3):e184. DOI 10.1002/hsr2.184 Source examined: Original article body retrieved.

Source note: SRC-a75d340b9e57 Ninomiya K 2020

16. Hill AM, Ross-Adjie G, McPhail SM, Jacques A, Bulsara M, Cranfield A, et al. Incidence and Associated Risk Factors for Falls in Older Adults Postdischarge Who Undergo Elective Total Hip Replacement Surgery-A Prospective Cohort Study. The journals of gerontology. Series A, Biological sciences and medical sciences. 2021;76(10):1814-1820. DOI 10.1093/gerona/glaa283 Source examined: Original abstract only; full article not retrieved.

Source note: SRC-eb3200138bfd Hill AM 2021

17. Lin X, Wu W, Weijer RHA, Prins MR, van Dieën JH, Bruijn SM, et al. Strong relationship of muscle force and fall efficacy, but not of gait kinematics, with number of falls in the year after Total Hip Arthroplasty for osteoarthritis: An exploratory study. Clinical biomechanics (Bristol, Avon). 2022;92:105551. DOI 10.1016/j.clinbiomech.2021.105551 Source examined: Original article body retrieved.

Source note: SRC-de68397ec95d Lin X 2022

18. Okazawa K, Hamai S, Fujita T, Kawahara S, Hara D, Nakashima Y, et al. Effect of Early Postoperative Gait Parameters After Total Hip Arthroplasty on Forgotten Joint Score-12 at 2-Year Follow-Up. Geriatrics (Basel, Switzerland). 2025;10(1):7. DOI 10.3390/geriatrics10010007 Source examined: Original article body retrieved.

Source note: SRC-a52095124e80 Okazawa K 2025

19. Kawano T, Nankaku M, Murao M, Goto K, Kuroda Y, Kawai T, et al. Development of a Clinical Prediction Rule to Identify Physical Activity After Total Hip Arthroplasty. Archives of physical medicine and rehabilitation. 2022;103(10):1975-1982. DOI 10.1016/j.apmr.2022.03.015 Source examined: Original abstract only; full article not retrieved.

Source note: SRC-26a83e470153 Kawano T 2022

20. Zügner R, Tranberg R, Lisovskaja V, Kärrholm J. Different reliability of instrumented gait analysis between patients with unilateral hip osteoarthritis, unilateral hip prosthesis and healthy controls. BMC musculoskeletal disorders. 2018;19(1):224. DOI 10.1186/s12891-018-2145-0 Source examined: Original article body retrieved.

Source note: SRC-68525777726d Zugner R 2018

21. Bravi M, Gallotta E, Morrone M, Maselli M, Santacaterina F, Toglia R, et al. Concurrent validity and inter trial reliability of a single inertial measurement unit for spatial-temporal gait parameter analysis in patients with recent total hip or total knee arthroplasty. Gait & posture. 2020;76:175-181. DOI 10.1016/j.gaitpost.2019.12.014 Source examined: Original abstract only; full article not retrieved.

Source note: SRC-7793ba556d88 Bravi M 2020

22. Alalem N, Gasparutto X, Rose-Dulcina K, DiGiovanni P, Hannouche D, Armand S. Validity of a Single Inertial Measurement Unit to Measure Hip Range of Motion During Gait in Patients Undergoing Total Hip Arthroplasty. Sensors (Basel, Switzerland). 2025;25(11):3363. DOI 10.3390/s25113363 Source examined: Original article body retrieved.

Source note: SRC-17bd12852330 Alalem N 2025

23. Lippi L, Desimoni F, Canonico M, Massocco G, Turco A, Polverelli M, et al. System for Tracking and Evaluating Performance (Step-App®): validation and clinical application of a mobile telemonitoring system in patients with knee and hip total arthroplasty. A prospective cohort study. European journal of physical and rehabilitation medicine. 2024;60(2):349-360. DOI 10.23736/s1973-9087.24.08128-0 Source examined: Original article body retrieved; agreement figures not visually verified; mixed procedures/indications.

Source note: SRC-19e0ae8bbb16 Lippi L 2024

24. Fary C, Cholewa J, Abshagen S, Van Andel D, Ren A, Anderson MB, et al. Stepping Beyond Counts in Recovery of Total Hip Arthroplasty: A Prospective Study on Passively Collected Gait Metrics. Sensors (Basel, Switzerland). 2023;23(14):6538. DOI 10.3390/s23146538 Source examined: Original article body retrieved.

Source note: SRC-e9a65ab0e76f Fary C 2023

25. Jeldi AJ, Deakin AH, Allen DJ, Granat MH, Grant M, Stansfield BW. Total Hip Arthroplasty Improves Pain and Function but Not Physical Activity. The Journal of arthroplasty. 2017;32(7):2191-2198. DOI 10.1016/j.arth.2017.02.002 Source examined: Original abstract only; full article not retrieved.

Source note: SRC-1e73689ae4a0 Jeldi AJ 2017

26. Gasparutto X, Gueugnon M, Laroche D, Martz P, Hannouche D, Armand S. Which functional tasks present the largest deficits for patients with total hip arthroplasty before and six months after surgery? A study of the timed up-and-go test phases. PloS one. 2021;16(9):e0255037. DOI 10.1371/journal.pone.0255037 Source examined: Original article body retrieved.

Source note: SRC-76ecfb3605f8 Gasparutto X 2021

27. Dobson F, Hinman R S, Hall M, Marshall C J, Sayer T, Anderson C, et al. Reliability and measurement error of the Osteoarthritis Research Society International (OARSI) recommended performance-based tests of physical function in people with hip and knee osteoarthritis. Osteoarthritis and cartilage. 2017;25(11):1792-1796. DOI 10.1016/j.joca.2017.06.006 Source examined: Original article body retrieved; numerical original tables incomplete in extraction; nonoperative mixed OA context.

Source note: SRC-55a59aa57129 Dobson F 2017

28. Stetter BJ, Dully J, Stief F, Holder J, Steingrebe H, Zaucke F, et al. Explainable machine learning for orthopedic decision-making: predicting functional outcomes of total hip replacement from gait biomechanics. Arthritis research & therapy. 2025;27(1):229. DOI 10.1186/s13075-025-03709-2 Source examined: Original article body retrieved.

Source note: SRC-cf6aff60f42d Stetter BJ 2025

29. Christensen JC, Judd DL, Thomsen-Freitas PB, Davis-Wilson H, Christiansen CL, Stevens-Lapsley JE. Kinetic Biomarkers of Cumulative Loading and Daily Step Count as Predictors of Functional Recovery Following Primary Unilateral Total Hip Arthroplasty. Journal of orthopaedic research : official publication of the Orthopaedic Research Society. 2026;44(5):e70203. DOI 10.1002/jor.70203 Source examined: Original abstract only; full article not retrieved.

Source note: SRC-b9e40de78da0 Christensen JC 2026