Abstract
Background
The prevalence of hip osteoarthritis is increasing and frequently requires total hip arthroplasty (THA). However, a standardized rehabilitation protocol to ensure sustainable functional recovery from THA has yet to be established. Therefore, this study compared the effects of real-time feedback-assisted home-exercise versus standard written instructions on the functional performance of patients following unilateral THA.
Methods
This study was designed as a cluster-randomized controlled pilot trial with two parallel groups. Data from three-dimensional movement analysis, clinical examination, body-composition, activity monitoring and questionnaires were collected at baseline four weeks post-surgery, and follow-ups at three- and six-months post-surgery. Baseline measurements were followed by an eight-week supervised group exercise therapy. For home-exercise the control-group received printed instructions, whereas the intervention-group was provided a real-time feedback prototype.
Results
A total of 22 patients, aged 52 to 73 years (mean 62.9), participated in the study, 21 completed the intervention and thereof, 19 could be assessed at six-months post-surgery. No statistically significant differences were found for the primary outcomes frontal knee, pelvis and trunk range-of-motion during level walking, nor the secondary outcomes Harris Hip Score (HHS), SF-12 and IPAQ short form. Both groups improved their mean walking speed, cadence, step length, isometric knee flexion force of operated limb, SF-12 physical domain and HHS.
Conclusion
Real-time feedback assisted home-exercise after THA did not differ when compared to commonly used written instructions in improving function and quality of life. However, patients in the real-time feedback group reported greater motivation and self-confidence regarding exercise performance, which was not reflected in the recorded numbers of home-exercise sessions.
Trial registration
The study was prospectively registered on 7th December 2023 at ClinicalTrials.gov NCT06161194.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12891-026-10321-1.
Keywords: Hip replacement, Exercise therapy, Real-time feedback, Kinematic, Kinetic, Gait
Background
The incidence of hip osteoarthritis (OA) is increasing with symptomatic onset between the age of 50 to 75 years [1]. The main factors contributing to the development and progression of hip OA include obesity, nutrition, sociodemographic factors, injuries, or joint shape, and have been extensively studied [2, 3]. Further biomechanical aspects related to structural cartilage overload [1] induced by a sum of loads [4] have been identified. Some studies report an impact of functional gait deviations, including pelvic drop, extensive lateral trunk lean, or varus-valgus thrust on the progression of OA [4, 5]. Depending on the severity of symptoms, such as pain and functional loss, conservative exercise treatment [6] or surgical total hip arthroplasty (THA) is recommended. However, the optimal timing for THA remains difficult to define due to limited evidence [7].
In Austria, a total of 18,052 primary THA and 1,290 re-implantation surgeries were performed in 2015 [8]. Despite these numbers of THA surgeries, no national recommendation for pre- and post-operative management concerning functional rehabilitation is implemented [9, 10]. Some evidence-based knowledge exists on the effectiveness of exercise therapy concerning the reduction of functional deficits after THA [11–13]. In particular, motor-re-education approaches [14–16] and strategies to enhance exercise sustainability through real-time feedback are proposed in several studies [17–19]. Regarding motor-re-education, quality of performance, compliance, and adherence to exercise are essential. However, there is no system available that focuses on relevant functional deficits, such as valgus thrust, pelvic obliquity, or lateral trunk lean, via real-time feedback assisted home-exercise. Furthermore, few longitudinal studies report kinematic data, and to our knowledge none report three-dimensional kinetic data addressing those functional deficits [11, 12, 20, 21].
Therefore, we aimed at comparing the effect of home-exercise assisted by digital real-time feedback in the intervention group (IG) with written home-exercise instructions in the control group (CG). The primary null-hypothesis was that improvements of frontal knee, hip, and pelvis range-of-motions (ROM) were equal between IG and CG. Secondary hypotheses were tested for patient-reported outcome measures (PROMS). Functional performance was assessed at three- and six-month post-surgery using three-dimensional gait analysis (3DGA) during level walking (LW) and quality of life (QoL) was assessed using physical activity questionnaires.
Methods
Study design
The study was designed as a cluster-randomized controlled pilot trial with two parallel groups and published in Widhalm et al. 2024 [22]. Patients were recruited prior to their surgery. The baseline assessment (t0) was conducted four to six weeks after surgery, and randomization followed thereafter. During the active phase of eight weeks, inpatient supervised exercise therapy took place in both groups. In addition, IG performed a home-exercise program with the support of the home-SETT system, a prototype real-time feedback system developed by the study group [23, 24]. CG performed an identical home-exercise program without the support of a real-time feedback device (i.e. a home-exercise program following written instructions). Follow-up assessments took place as close as possible to eight weeks after baseline (t1) and six-month post-surgery (t2).
Participants
Due to the lack of existing comparable research on the effects of real-time feedback for exercising following THA, it was not possible to estimate an appropriate sample size. Accordingly, this study was conducted and reported as a pilot. Patients were recruited at the Orthopedic Hospital Speising, Vienna, Austria, following ethics approval from the ethics committee of the City of Vienna (23–103–0623) and authorization for this clinical investigation from the Austrian Federal Office for Safety in Health (INS 102397778).
Inclusion and exclusion criteria
Patients aged from 50–75 years, with a body mass index of 18.5 to 32.49 kg/m2, the ability to walk without walking aids at baseline assessment, scheduled for elective unilateral THA, and willing to comply with all study-related procedures and to provide informed consent were included. All patients who were screened for eligibility underwent the same standardized surgical procedure, a direct anterior approach, which is a muscle-sparing approach that utilizes an internervous and intermuscular plane between the sartorius and tensor fasciae latae muscle, allowing access to the hip joint without detaching major muscles from the femur or pelvis [25, 26]. This minimizes soft tissue disruption and has shown better functional outcomes when compared to the posterior and lateral approach [27]. Patients were mobilized on the day of surgery. To facilitate rapid rehabilitation, preoperative patient education, multimodal pain management, and early physiotherapy were offered.
Patients were excluded in accordance with the protocol [22] in case of reasons which would lead to obvious limitations concerning their participation in the assessments and exercise therapy interventions. Due to understanding of questionnaires and instructions an inadequacy in German language, and non-sufficient adherence to exercise also led to exclusion.
Recruitment, informed consent, and randomization
At the preoperative assessment potential participants were informed by the physician (SLB), checked for eligibility, and in case of willingness written informed consent was obtained prior to surgery. On the day of hospital discharge, patients were assessed concerning their status of recovery regarding exclusion criteria. All subsequent assessments and interventions took place at the University of Applied Sciences Campus Vienna.
After the baseline assessment, patients were assigned to groups of two to five participants, which was arbitrarily done based on surgery date. Clusters were randomized in blocks of two to achieve an equal allocation of clusters to both arms. Due to the small number of planned clusters, no stratified randomization approach was applied. Due to the nature of the active treatments studied, participant blinding was not applicable.
The total active participation time for completing the assessments performed at baseline, three-months and six-months follow-up was six hours. A maximum of 24 h of inpatient supervised exercise sessions were offered, and participants were instructed to practice their home-exercise program for 20 h in total.
Safety considerations and participant insurance
The risk of adverse events was expected to be low. The patients were clinically assessed by the principal investigator (KW) prior to the baseline measurement and the treatment. Exercise therapy was instructed and supervised by KW and a second physiotherapist (LBM). If necessary, on-site visits at the patients’ homes were done by the study team to ensure safe home-exercise and system setup. All study-related procedures were covered by a clinical trial insurance.
Treatment
The eight-week supervised exercise therapy and home-exercise – real-time feedback assisted for IG and paper hand-out instruction for CG – treatment, and assessments at t0, t1 and t2 were performed as outlined in the study protocol [22].
Clinical assessment
Following the clinical assessment, including an interview and measuring height, ROM of the hip, knee, and ankle joint with a goniometer according to the neutral-zero method, isometric strength of knee extension and flexion, and a functional task check was conducted by KW. Thereafter, all patients were equipped with retroreflective markers as defined per protocol [22]. After calibration measurement and familiarization trials, five eligible LW trials were captured by a 28 infrared camera system with the software Nexus version 2.9.3 (Vicon Motion Systems Limited, Oxford, UK). Following patients were equipped with a waist-worn accelerometer (wGT3X-BT, ActiGraph LLC, Pensacola, FL, USA) to record their physical activity over seven consecutive days.
Group-exercise program
Twice a week, group-exercise therapy sessions of 90 min each were instructed by KW and LBM in groups (as defined per clusters) of two to five patients, resulting in a maximum of 16 sessions. Both groups, IG and CG, were instructed and supervised in an equal program. The patients’ well-being was assessed and documented at the beginning of each session by the physiotherapist.
Group-exercise therapy and home-exercise program for both groups followed the FITT (frequency, intensity, time, type of training) principle and reported following the Consensus on Exercise Reporting Template (CERT) [28]. Intensity of each exercise was increased or decreased following patients’ self-reported perceived rate of exertion with the help of a Numerical Rating Scale (NRS) graded from 0 (very light exertion) to 10 (maximum exertion) [29]. For all patients, the absolute exercise therapy load was recorded. Patient-reported hip pain before and after exercise sessions were assessed using a NRS graded from 0 (no pain) to 10 (worst pain imaginable). Treatments and a detailed list of the exercise therapy program are provided in the appendix of the study protocol [22].
Home-exercise program
Patients performed a daily 30-min home-exercise program, excluding days with group exercise program. They documented the day, performed exercises, repetitions, pain on the NRS-10 scale, and a perceived exertion rating in a paper exercise diary. Both groups received printed handouts of the home-exercise program. The IG group performed the home-exercise program supported by the home-SETT digital real-time exercise feedback prototype for digitally guided training. The CG was therefore an active control receiving equal usual care treatment, including an equal group-exercise and written home-exercise program. A brief description of the prototype and a completed example of a home-exercise protocol can be found in the appendix of the study protocol [22].
Outcome parameters
Raw data were processed using Nexus version 2.14 (Vicon Motion Systems Limited, Oxford, UK) and trajectory data were filtered with the standard Nexus woltring (MSE, 15) filter routine. Primary outcome parameters were the frontal plane trunk (fTrunkROM), pelvis (fPelvisROM) and knee (fKROM) range-of-motion during the loading phase of the gait cycle, which was determined from initial contact to toe-off of the contralateral limb [30]. Secondary outcome parameters were the results from the HHS [31], the SF-12 [32, 33] and the IPAQ-SF [34]. In addition, kinematic, kinetic and spatio-temporal parameters were analyzed. Accelerometer-based physical activity data were collected to complement patient characterization, which was then categorized, and their daily step counts were extracted [35]. Body muscle mass, left and right leg muscle mass, and body fat rate were assessed based on bioelectrical impedance. Motivation, adherence and system usability were extracted from the exercise diary, interview and system usability scoring. Age and gender were recorded. Details of processing and parameters are described in the study protocol [22].
Statistical methods
Data were inspected for outliers using boxplots. The alignment of the data with a normal distribution was tested using Shapiro–Wilk tests and additional graphical inspections of q-q plots. Metric sample baseline characteristics are indicated as means with standard deviation (SD) and categorical baseline characteristics as frequencies (n, %). Unadjusted outcome scores are reported as means with the corresponding SD. To model changes between baseline and follow-up, we included baseline measurements as covariates. Following suggestions by Klar and Darlington [36], we adjusted for each patient’s baseline assessment and for the baseline cluster mean. To confirm the robustness of primary findings, exploratory sensitivity analyses were conducted by adjusting for random slopes and further covariates (leg length normalized walking speed, leg muscle mass, age, sex, and adherence). Treatment effects are reported independently for follow-up t1 and follow-up t2, including unstandardized and standardized regression coefficients as effect estimates, standard errors, t statistics, and appropriate P values via the Satterthwaite degrees-of-freedom method. All descriptive statistics and corresponding numbers of subjects (n, %) are reported, along with losses to follow-ups. Effect sizes are calculated and reported accordingly (per protocol).
An additional mixed-effects model was run separately for CG and IG for typically reported gait characteristics and secondary outcomes to investigate improvement throughout the trial. Treatment was coded as a binary numeric variable (0/1) to allow full standardization of coefficients. The α value was set to 0.05, and exact P values are reported. A total of 30 hypotheses were tested on the sample, resulting in a Bonferroni corrected alphaBonferroni = 0.0017.
Results
Overall, 22 patients with a mean (SD) age of 62.9 (5.9) years were assessed at baseline. 21 of them completed the treatment phase and were assessed three-months post-surgery, 19 thereof also at six-months post-surgery. Reasons for drop-outs during the trial were acute non-surgery-related issues during the treatment phase (n = 1), low-back-pain prior to t1 at three-month (n = 1), and no-show at t1 (n = 1). No adverse events were observed during the intervention phase, nor till the six-month follow-up. Patient characteristics are presented in Table 1.
Table 1.
Patient characteristics for all, IG and CG at t0, t1, t2 as MEAN (SD)
| Group | n t0/t1/t2 | Baseline t0 | 3-months t1 | 6-months t2 | |
|---|---|---|---|---|---|
| BMI [kg/m2] | All | 22/21/19 | 25.96 (3.34) | 26.12 (3.10) | 26.10 (3.17) |
| CG | 12/12/10 | 26.49 (2.49) | 26.26 (2.47) | 26.23 (2.36) | |
| IG | 10/9/9 | 25.33 (4.21) | 25.93 (3.94) | 25.94 (4.04) | |
| Bodyfat [% BW] | All | 20/20/18 | 32.20 (6.22) | 32.00 (6.05) | 31.42 (6.27) |
| CG | 10/11/9 | 29.25 (6.43) | 29.34 (5.96) | 28.27 (6.02) | |
| IG | 10/9/9 | 35.14 (4.60) | 35.21 (4.60) | 34.58 (4.99) | |
| Musclemass [% BW] | All | 20/20/18 | 30.85 (4.05) | 31.43 (3.78) | 31.63 (4.09) |
| CG | 10/11/9 | 33.32 (3.58) | 33.41 (3.26) | 33.85 (3.57) | |
| IG | 10/9/9 | 28.38 (2.87) | 29.01 (2.95) | 29.41 (3.43) | |
| Musclemass op [% BW] | All | 20/20/18 | 6.46 (0.81) | 6.69 (0.65) | 6.70 (0.60) |
| CG | 10/11/9 | 6.89 (0.83) | 6.99 (0.63) | 6.98 (0.56) | |
| IG | 10/9/9 | 6.02 (0.53) | 6.31 (0.48) | 6.42 (0.53) | |
| Musclemass non-op [% BW] | All | 20/20/18 | 6.53 (0.68) | 6.65 (0.59) | 6.65 (0.60) |
| CG | 10/11/9 | 6.75 (0.71) | 6.83 (0.63) | 6.80 (0.58) | |
| IG | 10/9/9 | 6.30 (0.60) | 6.44 (0.49) | 6.49 (0.62) | |
| avg MVPA [min/day] | All | 17/15/16 | 37.3 (24.4) | 60.7 (28.9) | 58.4 (28.5) |
| CG | 9/8/10 | 32.5 (27.3) | 58.8 (31.4) | 49.5 (31.2) | |
| IG | 8/7/6 | 42.8 (21.1) | 62.8 (28.0) | 59.9 (27.1) |
BMI Body-Mass-Index, op operated leg, non-op non-operated leg, avg MVPA average Moderate-to-Vigorous-Physical-Activity
Days from surgery to the assessments as mean (SD) were for baseline = 34 (5) days, three-month = 95 (6) days, and six-month = 199 (11) days. Figure 1 illustrates the patient flow through the trial according to the CONSORT extension on cluster randomized trials [37].
Fig. 1.
Flow diagram of the progress through the phases of the trial [37]. IG RTF = Intervention group Real-time Feedback, CG PHO = Control group Paper handout. * eligibility based on clinical data
The results of the main statistics for the primary outcomes fKROM, fPelvisROM, and fTrunkROM, as well as the secondary outcomes HHS, SF-12 and IPAQ-SF did not reveal any significant differences between CG and IG. Additional sensitivity analyses adjusting for random slopes and further covariates (leg length normalized walking speed, leg muscle mass, age, sex, and adherence) confirmed the absence of significant differences between CG and IG. Details are presented in Table 2 and Supplementary File 1.
Table 2.
Effects of primary and secondary trial outcomes for level walking by treatment: control vs. intervention
| Baseline t0: mean ± SD | Follow-up t1: mean ± SD | Follow-up t2: mean ± SD | Follow-up 1 | Follow-up 2 | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Control | Intervention | Control | Intervention | Control | Intervention | ß | p | ß | p | |
| Knee frontal RoM (°) | 1.21 ± 0.51 | 1.65 ± 1.21 | 1.27 ± 0.44 | 1.69 ± 1.02 | 1.40 ± 0.81 | 1.72 ± 1.22 | 0.34 | 0.449 | 0.09 | 0.894 |
| Pelvis frontal RoM (°) | 1.57 ± 1.33 | 1.77 ± 0.68 | 2.25 ± 1.32 | 2.07 ± 0.70 | 2.00 ± 1.14 | 2.20 ± 1.01 | −0.34 | 0.333 | 0.10 | 0.790 |
| Trunk frontal RoM (°) | 1.48 ± 0.89 | 0.87 ± 0.53 | 1.04 ± 0.76 | 0.99 ± 0.50 | 1.07 ± 0.71 | 0.91 ± 0.56 | 0.13 | 0.725 | 0.10 | 0.846 |
| Harris Hip Score (0–100) | 72.9 ± 19.8 | 81.0 ± 9.07 | 94.5 ± 8.23 | 95.6 ± 5.00 | 95.8 ± 9.30 | 97.2 ± 2.49 | −0.87 | 0.768 | −0.33 | 0.926 |
| SF-12 PH (0–100) | 33.5 ± 4.64 | 31.7 ± 4.21 | 46.5 ± 11.7 | 47.5 ± 9.12 | 50.6 ± 8.68 | 50.0 ± 7.83 | 1.95 | 0.679 | −0.25 | 0.958 |
| SF-12 MH (0–100) | 51.2 ± 11.4 | 53.2 ± 10.9 | 51.5 ± 11.0 | 54.3 ± 5.13 | 55.4 ± 2.87 | 55.4 ± 4.48 | −0.56 | 0.883 | −2.01 | 0.444 |
| IPAQ-SF Met-mins per week | 1334 ± 1141 | 1805 ± 2288 | 2926 ± 1900 | 5061 ± 5054 | 2362 ± 1253 | 4999 ± 4800 | 2558 | 0.214 | 2214 | 0.422 |
ß … estimated effect of the intervention (vs. control) in the unit of the outcome, e.g. (°)
p … p-values calculated using mixed linear methods (Satterthwaite's method) to compare outcomes by treatment (i.e. control vs intervention), adjusted for individual baseline values and baseline cluster means (random intercept)
Comment: additional exploratory models adjusting for random slopes and further covariates (leg length normalized walking speed, leg muscle mass, age, sex, and adherence) did not result in any statistically significant effects
During the eight weeks treatment phase patients performed a mean (SD) of 32.6 (10.3) home-exercise sessions—CG 31.2 (12.3); IG 34.6 (6.8)—and participated in 13.1 (2.7) of 16 offered supervised exercise therapy sessions—CG 12.3 (1.7); IG 14.8 (1.4). An additional analysis showed statistically significant improvements in both groups in walking speed, cadence, step length, isometric knee flexion force op-side, HHS, and SF-12 physical with large effect sizes. In addition, IG improved significantly in muscle mass of the op-side. Detailed results are presented in Table 3 and model details of this longitudinal analysis in Supplementary File 1.
Table 3.
Results of linear mixed-effects models for overall change across 3 time points as available case analysis with mean (SD), by CG and IG
| Group | n t0/1/2 | Baseline t0 | 3-months t1 | 6-months t2 | ß | stand. ß | CI95 | p | |
|---|---|---|---|---|---|---|---|---|---|
| Walking Speed norm | CG | 12/12/10 | 0.34 (0.08) | 0.42 (0.05) | 0.44 (0.03) | 0.05 | 0.59 | 0.38, 0.81 | < 0.001* |
| IG | 10/9/9 | 0.35 (0.04) | 0.44 (0.05) | 0.45 (0.06) | 0.05 | 0.64 | 0.40, 0.88 | < 0.001* | |
| Step Width [m] | CG | 12/12/10 | 0.08 (0.02) | 0.08 (0.02) | 0.07 (0.02) | −0.00 | −0.16 | −0.34, 0.02 | 0.082 |
| IG | 10/9/9 | 0.09 (0.02) | 0.09 (0.01) | 0.09 (0.01) | 0.00 | 0.06 | −0.16, 0.28 | 0.595 | |
| Step Length norm op | CG | 12/12/10 | 0.63 (0.11) | 0.72 (0.05) | 0.74 (0.03) | 0.06 | 0.56 | 0.33, 0.79 | < 0.001* |
| IG | 10/9/9 | 0.70 (0.03) | 0.79 (0.04) | 0.79 (0.07) | 0.04 | 0.55 | 0.32, 0.79 | < 0.001* | |
| Step Length norm non-op | CG | 12/12/10 | 0.65 (0.13) | 0.74 (0.07) | 0.77 (0.06) | 0.06 | 0.48 | 0.30, 0.67 | < 0.001* |
| IG | 10/9/9 | 0.65 (0.06) | 0.75 (0.07) | 0.77 (0.08) | 0.05 | 0.59 | 0.40, 0.77 | < 0.001* | |
| Isometric force extension op | CG | 12/12/10 | 189 (57) | 231 (62) | 290 (68) | 48.15 | 0.54 | 0.43, 0.64 | < 0.001* |
| IG | 10/9/9 | 157 (29) | 202 (52) | 213 (31) | 28.32 | 0.53 | 0.26, 0.81 | < 0.001* | |
| Isometric force extension non-op | CG | 12/12/10 | 217 (56) | 239 (59) | 286 (56) | 31.76 | 0.42 | 0.28, 0.55 | < 0.001* |
| IG | 10/9/9 | 192 (41) | 208 (43) | 219 (30) | 13.06 | 0.28 | 0.02, 0.54 | 0.039 | |
| Isometric force flexion op | CG | 12/10/10 | 102 (41) | 137 (59) | 183 (68) | 37.59 | 0.48 | 0.31, 0.65 | < 0.001* |
| IG | 10/9/9 | 101 (23) | 130 (31) | 134 (25) | 16.93 | 0.47 | 0.28, 0.67 | < 0.001* | |
| Isometric force flexion non-op | CG | 12/12/10 | 114 (44) | 131 (50) | 171 (68) | 26.39 | 0.38 | 0.19, 0.56 | < 0.001* |
| IG | 10/9/9 | 112 (37) | 124 (22) | 133 (27) | 10.28 | 0.29 | 0.07, 0.50 | 0.013 | |
| Musclemass op [% BW] | CG | 10/11/9 | 6.89 (0.83) | 7.00 (0.63) | 7.00 (0.56) | −0.00 | 0.00 | −0.16, 0.15 | 0.956 |
| IG | 10/9/9 | 6.02 (0.53) | 6.31 (0.48) | 6.42 (0.53) | 0.20 | 0.32 | 0.17, 0.48 | < 0.001* | |
| Musclemass non-op [% BW] | CG | 10/11/9 | 6.75 (0.71) | 6.83 (0.61) | 6.80 (0.58) | −0.01 | −0.02 | −0.14, 0.10 | 0.782 |
| IG | 10/9/9 | 6.30 (0.60) | 6.44 (0.49) | 6.49 (0.62) | 0.09 | 0.14 | 0.02, 0.26 | 0.028 | |
| Harris-Hip-Score [sum] | CG | 12/12/10 | 72.4 (19.0) | 93.9 (8.1) | 95.8 (9.3) | 12.40 | 0.60 | 0.37, 0.83 | < 0.001* |
| IG | 10/9/9 | 81.0 (9.1) | 95.6 (5.0) | 97.2 (2.5) | 8.23 | 0.71 | 0.46, 0.97 | < 0.001* | |
| IPAQ-SF Met-mins per week | CG | 12/12/10 | 1338 (1088) | 2918 (1812) | 2598 (1031) | 414 | 0.22 | −0.10, 0.54 | 0.172 |
| IG | 10/9/9 | 1805 (2288) | 5623 (5018) | 4999 (4800) | 1347 | 0.26 | −0.07, 0.58 | 0.119 | |
| SF-12 Physical [sum] | CG | 12/12/10 | 33.6 (5.3) | 45.5 (11.7) | 50.6 (8.7) | 9.08 | 0.64 | 0.45, 0.82 | < 0.001* |
| IG | 10/9/9 | 31.2 (4.2) | 47.5 (9.1) | 50.0 (7.8) | 9.26 | 0.71 | 0.44, 0.98 | < 0.001* | |
| SF-12 Mental [sum] | CG | 12/12/10 | 52.0 (11.2) | 52.2 (10.8) | 55.4 (2.9) | 1.39 | 0.12 | −0.12, 0.37 | 0.316 |
| IG | 10/9/9 | 53.2 (10.9) | 54.3 (5.1) | 55.4 (4.5) | −0.02 | 0.00 | −0.19, 0.18 | 0.980 | |
| avg MVPA [min/day] | CG | 9/8/10 | 32.5 (27.3) | 58.8 (31.4) | 57.5 (30.7) | 11.03 | 0.30 | 0.12, 0.48 | 0.003 |
| IG | 8/7/6 | 42.8 (21.1) | 62.8 (28.0) | 59.9 (27.1) | 8.94 | 0.29 | −0.17, 0.75 | 0.205 |
β … unstandardized regression coefficient, stand. β … standardized regression coefficient, op … operated leg, non-op … non-operated leg, SF-12 Physical … Short-Form-12 Health Survey Physical Component,
SF-12 Mental … Short-Form-12 Health Survey Mental Component, avg MVPA … average Moderate-to-Vigorous-Physical-Activity
*Significant after Bonferroni correction
Analysis of the clinical assessment interviews showed an improvement in pain from t0 to t2 in both groups and a decrease in knee complaints. Patients in the IG suffering from Low-Back-Pain decreased in numbers. Furthermore, IG reported a positive perception of the assistive and motivational capability of Home-SETT. (Supplementary File 2).
Visual inspection of concurrent kinematic data for IG, CG, and healthy reference at t1 showed notable differences for fPelvisROM and fHROM during stance phase (Fig. 2). Additionally kinematic data for t2 and kinetic data are provided in Supplementary File 3.
Fig. 2.
Kinematic gait report for the operated leg of IG, CG and healthy reference at three-month. Parameters presented over 100% gait cycle for CG (red), IG (blue) and healthy reference (grey) with shadow indicating standard deviation from the mean
Discussion
To our knowledge, this is the first study to investigate the effect of real-time feedback assisted home-exercise in patients following THA, combining supervised group exercise with an interventional and control home-exercise program, and using quantitative biomechanical and PROMs outcomes for evaluation. Our results suggest that the approach of physiotherapeutic supervised biweekly group exercise accompanied by digital assisted home-exercise is comparable to written exercise instructions in improving functional deficits.
Other studies implementing telerehabilitation home-exercise after three weeks of inpatient rehabilitation also reported equivalent improvements when comparing telerehabilitation with usual care home-exercise [38]. However, comparability is limited due to differences in time to intervention onset, minutes exercised per week, co-morbidities, and the lack of 3DGA data.
Although IG and CG did not differ in defined outcome variables, patients of the IG reported personal benefits in terms of higher motivation for regular exercise, which is consistent with findings from serious-gaming studies [39]. Moreover, increased self-confidence regarding exercise performance quality was highlighted. In particular, the availability of multiple instruction modalities by means of text, audio, and video was appreciated. The benefit of exercise motivation might potentially increase long-term adherence, which could also have an impact on cost-effectiveness as reported by Fatoye et al. [40]. Especially function and QoL improve after joint replacement when exercise is assisted by telerehabilitation [41].
In contrast to previous literature [42] clinically described limping after THA, characterized by increased frontal plane trunk excursions, was not consistently observed in our study. Results for pelvic drop were not directly comparable with other studies, as our analysis focused on enhanced neuromuscular stabilization during stance phase of level walking rather than surgery induced muscular deficits [43]. However, fKROM, fPelvisROM and fTrunkROM has to be interpreted cautiously, as minimal detectable change values are not published. Schwartz et al. reported measurement errors for fKROM for the stance phase of 0.5° [44]. Unpublished MDC results using ICC (2,1) based on data of Durstberger et al. [45] with 20 healthy participants revealed for fKROM 2.11°, fPelvisROM 1.17°, and fTrunkROM 0.63°. These commonly used parameters for describing functional deficits might not be sensitive enough and methods based on continuous data should be established for pattern characterization [46].
Regarding the HHS, patients in this study achieved a mean score of 96.5 (SD 6.8) at six-month follow-up, which is slightly higher than values reported in a walking skill based study (mean 93, range 91–96) at five months post-surgery [16]. As minimal clinically important improvement for HHS scores 15.9 to 18 points of 100 has been reported [47], the mean improvement of 20 points in our cohort suggests that patients started from a relatively high functional level four weeks post-surgery.
Limitations
Readers should be aware that this clinical investigation is a pilot, and not sufficiently powered. The estimate of 49 patients starting the intervention, which was planned by protocol [22], could not be reached because funding constraints required us to stop recruitment after nine months. Nevertheless, results of the study allow for conceptualization of a fully powered trial, as the real-time feedback approach appears safe for patients and yields outcomes comparable to usual care. Deviating from the protocol, no HHS data and activity monitoring data was collected prior to surgery. Furthermore, the HOOS [48] and KOOS [49] were not analyzed, as OA-oriented assessments turned out to be not suitable for patients after hip replacement. We decided so, as most of the items are highly related to pain, and patients participating in the study were no longer limited in function due to pain as the structural cause had been addressed through arthroplasty. Analysis of the agreement between the real-time feedback prototype and the optoelectronic reference system was not feasible, as an insufficient number of exercise repetitions could be derived due to technical limitations.
Regarding comorbidities such as low-back-pain or possible functional deficits due to prior injuries, no notable features were observed in 3DGA kinematics. Surprisingly, only one participant of those who started the treatment did not show up at the six-month follow-up. Dropouts occurred predominantly between recruitment and baseline assessment, accounting for 36.8%.
Conclusions
Real-time feedback and paper-based instructions for home-exercise, in addition to inpatient supervised exercise therapy, can be safely applied after unilateral THA and may improve function and QoL. Supervised exercise therapy combined with home-exercise – whether written or via Home-SETT instructed – starting four weeks post-surgery leads to statistically significant improvements. Whether additional physiotherapeutic guidance during the first four postoperative weeks and between six and twelve months post-surgery could further enhance long-term outcomes remains unclear. Accordingly, these protocol adaptations and a two-year follow-up should be considered in future studies.
Supplementary Information
Acknowledgements
The authors would like to thank rector Heimo Sandtner from the University of Applied Sciences Campus Vienna for approving the internal PhD-grant as additional staff resources to conduct this research.
Declaration of Generative AI and AI-assisted technologies in the writing process
During the preparation of this work, the authors did not use any AI or AI-assisted technologies.
Abbreviations
- 3DGA
Three-dimensional gait analysis
- CG
Control group
- fKROM
Frontal knee range-of-motion in degrees
- fPelvisROM
Frontal pelvis range-of-motion in degrees
- fTrunkROM
Frontal trunk range-of-motion in degrees
- HHS
Harris-Hip-Score
- HOOS
Hip osteoarthritis outcome score
- IG
Intervention group
- IPAQ-SF
International physical activity questionnaire - Short form
- KOOS
Knee osteoarthritis outcome score
- LW
Level walking
- MDC
Minimal detectable change
- n
Count
- NRS
Numeric rating scale (graded from 0 to 10)
- OA
Osteoarthritis
- PI
Principal investigator
- PROM
Patient-reported outcome measure
- QoL
Quality of life
- ROM
Range-of-motion
- SD
Standard deviation
- SF-12
SF-12 health survey
- t0
Baseline assessment 4–6 weeks post-surgery
- t1
Follow-up post 8-week intervention
- t2
Follow-up 6-months post-surgery
- THA
Total hip arthroplasty
Authors’ contributions
Conceptualization: KW, LBM, SLB; Study protocol: KW; Intervention: KW, LBM; Data acquisition: KW, HP; Development of feedback application: SD, LBM, HP, KW; Statistical workflow: PP, KW; Formal analysis: KW, HP, SD, PP; Supervision: HK, PA; Writing – Original draft: KW; Writing – Review & Editing: SD, PP, HP, LBM, SLB, HK, PA; All authors provided feedback on the drafts and approved the final version.
Funding
This clinical trial was performed within the research project SETT – Smart real-time training feedback (www.sett.at), which was funded within the 30th University of Applied Sciences Call for Proposals "Research at the Viennese Universities of Applied Sciences" by the Municipal Department 23 of the City of Vienna. Additionally means from a University of Applied Sciences Campus Vienna internal PhD-grant were provided.
Data availability
Processed study data are available from the corresponding author upon reasonable request.
Declarations
Ethical approval and consent to participate
Ethical approval was obtained via the ethics committee of the City of Vienna, Austria (23–103-0623) and the study was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants prior to their inclusion in the study. Conducting this clinical investigation was confirmed by the Austrian Federal Office for Safety in Health (INS 102397778).
Consent for publication
Not applicable.
Competing interests
The authors declare not to have any financial or academic conflict of interest.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Hunter DJ, Bierma-Zeinstra S. Osteoarthritis. Lancet. 2019;393(10182):1745–59. [DOI] [PubMed] [Google Scholar]
- 2.Vina ER, Kwoh CK. Epidemiology of osteoarthritis: literature update. Curr Opin Rheumatol. 2018;30(2):160–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ackerman IN, Kemp JL, Crossley KM, Culvenor AG, Hinman RS. Hip and knee osteoarthritis affects younger people. Too J Orthop Sports Phys Ther. 2017;47(2):67–79. [DOI] [PubMed] [Google Scholar]
- 4.Tateuchi H. Gait- and postural-alignment-related prognostic factors for hip and knee osteoarthritis: Toward the prevention of osteoarthritis progression. Phys Ther Res. 2019;22(1):31–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Diamond LE, Devaprakash D, Cornish B, Plinsinga ML, Hams A, Hall M, et al. Feasibility of personalised hip load modification using real-time biofeedback in hip osteoarthritis: a pilot study. Osteoarthritis Cartilage Open. 2022;4(1):100230. [DOI] [PMC free article] [PubMed]
- 6.Fransen M, McConnell S, Hernandez-Molina G, Reichenbach S. Exercise for osteoarthritis of the hip. Cochrane Database Syst Rev. 2014;4:CD007912. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Frydendal T, Christensen R, Mechlenburg I, Mikkelsen LR, Overgaard S, Ingwersen KG. Total hip arthroplasty versus progressive resistance training in patients with severe hip osteoarthritis: protocol for a multicentre, parallel-group, randomised controlled superiority trial. BMJ Open. 2021;11(10):e051392. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Leitner L, Türk S, Heidinger M, Stöckl B, Posch F, Maurer-Ertl W, et al. Trends and economic impact of hip and knee arthroplasty in Central Europe: findings from the Austrian National Database. Sci Rep. 2018;8(1):4707. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Krysa JA, Ho C, O’Connell P, Pohar MK. Clinical practice recommendations for prehabilitation and post-operative rehabilitation for arthroplasty: A scoping review. Musculoskeletal Care. 2022;20(3):503–15. [DOI] [PubMed] [Google Scholar]
- 10.Peter WF, Nelissen RG, Vlieland TP. Guideline recommendations for post-acute postoperative physiotherapy in total hip and knee arthroplasty: are they used in daily clinical practice? Musculoskeletal Care. 2014;12(3):125–31. [DOI] [PubMed] [Google Scholar]
- 11.Bahl JS, Nelson MJ, Taylor M, Solomon LB, Arnold JB, Thewlis D. Biomechanical changes and recovery of gait function after total hip arthroplasty for osteoarthritis: a systematic review and meta-analysis. Osteoarthritis Cartilage. 2018;26(7):847–63. [DOI] [PubMed] [Google Scholar]
- 12.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] [PMC free article] [PubMed] [Google Scholar]
- 13.Mikkelsen LR, Petersen AK, Søballe K, Mikkelsen SS, Mechlenburg I. Asymmetry and pelvic movements 6 months after total hip replacement Secondary analyses from a randomized controlled trial. Acta Orthop Belg. 2019;85(3):338–45. [PubMed] [Google Scholar]
- 14.Judd DL, Winters JD, Stevens-Lapsley JE, Christiansen CL. Effects of neuromuscular reeducation on hip mechanics and functional performance in patients after total hip arthroplasty: A case series. Clin Biomech (Bristol, Avon). 2016;32:49–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Saueressig T, Owen PJ, Zebisch J, Herbst M, Belavy DL. Evaluation of Exercise Interventions and Outcomes After Hip Arthroplasty: A Systematic Review and Meta-analysis. JAMA Netw Open. 2021;4(2):e210254-e. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Heiberg KE, Bruun-Olsen V, Ekeland A, Mengshoel AM. Effect of a walking skill training program in patients who have undergone total hip arthroplasty: followup one year after surgery. Arthritis Care Res (Hoboken). 2012;64(3):415–23. [DOI] [PubMed] [Google Scholar]
- 17.Raaben M, Vogely HC, Blokhuis TJ. Real-time visual biofeedback to improve therapy compliance after total hip arthroplasty: A pilot randomized controlled trial. Gait Posture. 2018;61:306–10. [DOI] [PubMed] [Google Scholar]
- 18.Uccheddu F, Furferi R, Governi L, Carfagni M. RGB-D-Based Method for Measuring the Angular Range of Hip and Knee Joints during Home Care Rehabilitation. Sensors. 2022;22(1):184. [DOI] [PMC free article] [PubMed]
- 19.van Dijk-Huisman HC, Weemaes ATR, Boymans T, Lenssen AF, de Bie RA. Smartphone App with an Accelerometer Enhances Patients' Physical Activity Following Elective Orthopedic Surgery: A Pilot Study. Sensors (Basel, Switzerland). 2020;20(15):4317. [DOI] [PMC free article] [PubMed]
- 20.Bertaux A, Gueugnon M, Moissenet F, Orliac B, Martz P, Maillefert JF, et al. Gait analysis dataset of healthy volunteers and patients before and 6 months after total hip arthroplasty. Sci Data. 2022;9(1):399. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Loppini M, Temporiti F, Furone R, Galli M, Grappiolo G, Gatti R. Static and dynamic pelvic kinematics after one-stage bilateral or unilateral total hip arthroplasty. Hip Int. 2021;31(6):729–34. [DOI] [PubMed] [Google Scholar]
- 22.Widhalm K, Maul L, Durstberger S, Putz P, Leder-Berg S, Kainz H, et al. Efficacy of Real-Time Feedback Exercise Therapy in Patients Following Total Hip Arthroplasty: Protocol for a Pilot Cluster-Randomized Controlled Trial. JMIR Res Protoc. 2024;13:e59755. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Widhalm K, Maul L, Durstberger S, Klupper C, Putz P. Real-time digital feedback for exercise therapy of lower extremity functional deficits: a mixed methods study of user requirements. Stud Health Technol Inform. 2023;301:108–14. [DOI] [PubMed] [Google Scholar]
- 24.Widhalm K, Maul L, Durstberger S, Putz P, Klupper C, Werner F. Real-Time Digitized Visual Feedback in Exercise Therapy for Lower Extremity Functional Deficits: Qualitative Study of Usability Factors During Prototype Testing. JMIR Serious Games. 2024;12:e51771. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Post ZD, Orozco F, Diaz-Ledezma C, Hozack WJ, Ong A. Direct anterior approach for total hip arthroplasty: indications, technique, and results. J Am Acad Orthop Surg. 2014;22(9):595–603. [DOI] [PubMed] [Google Scholar]
- 26.Cadossi M, Sambri A, Tedesco G, Mazzotti A, Terrando S, Faldini C. Anterior approach in total hip replacement. Orthopedics. 2017;40(3):e553–6. [DOI] [PubMed] [Google Scholar]
- 27.Ang JJM, Onggo JR, Stokes CM, Ambikaipalan A. Comparing direct anterior approach versus posterior approach or lateral approach in total hip arthroplasty: a systematic review and meta-analysis. Eur J Orthopaedic Surg Traumatol. 2023;33(7):2773–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Slade SC, Dionne CE, Underwood M, Buchbinder R. Consensus on Exercise Reporting Template (CERT): Explanation and Elaboration Statement. Br J Sports Med. 2016;50(23):1428–37. [DOI] [PubMed] [Google Scholar]
- 29.Williams N. The Borg Rating of Perceived Exertion (RPE) scale. Occup Med. 2017;67(5):404–5. [Google Scholar]
- 30.Perry J, Burnfield JM. Gait analysis: normal and pathological function. 2nd ed. Thorofare, N.J.: SLACK; 2010. [Google Scholar]
- 31.Nilsdotter A, Bremander A. Measures of hip function and symptoms: Harris Hip Score (HHS), Hip Disability and Osteoarthritis Outcome Score (HOOS), Oxford Hip Score (OHS), Lequesne Index of Severity for Osteoarthritis of the Hip (LISOH), and American Academy of Orthopedic Surgeons (AAOS) Hip and Knee Questionnaire. Arthritis Care Res. 2011;63(S11):S200–7. [DOI] [PubMed] [Google Scholar]
- 32.Gandek B, Ware J, Aaronson N, Apolone G, Bjorner J, Brazier J, et al. Cross-Validation of Item Selection and Scoring for the SF-12 Health Survey in Nine Countries: Results from the IQOLA Project. J Clin Epidemiol. 1998;51:1171–8. [DOI] [PubMed] [Google Scholar]
- 33.Drixler K, Morfeld M, Glaesmer H, Brähler E, Wirtz MA. Validation of the Short-Form-Health-Survey-12 (SF-12 Version 2.0) assessing health-related quality of life in a normative German sample. Zeitschrift fur Psychosomatische Medizin und Psychotherapie. 2020;66(3):272–86. [DOI] [PubMed] [Google Scholar]
- 34.Craig CL, Marshall AL, Sjostrom M, Bauman AE, Booth ML, Ainsworth BE, et al. International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc. 2003;35(8):1381–95. [DOI] [PubMed] [Google Scholar]
- 35.Putz P, Klinger M, Male C, Pabinger I. Lower physical activity and altered body composition in patients with haemophilia compared with healthy controls. Haemophilia. 2021;27(2):e260–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Klar N, Darlington G. Methods for modelling change in cluster randomization trials. Stat Med. 2004;23(15):2341–57. [DOI] [PubMed] [Google Scholar]
- 37.Campbell MK, Piaggio G, Elbourne DR, Altman DG. Consort 2010 statement: extension to cluster randomised trials. BMJ. 2012;345:e5661. [DOI] [PubMed] [Google Scholar]
- 38.Eichler S, Salzwedel A, Rabe S, Mueller S, Mayer F, Wochatz M, et al. The Effectiveness of Telerehabilitation as a Supplement to Rehabilitation in Patients After Total Knee or Hip Replacement: Randomized Controlled Trial. JMIR Rehabil Assist Technol. 2019;6(2):e14236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Ling Y, Ter Meer LP, Yumak Z, Veltkamp RC. Usability Test of Exercise Games Designed for Rehabilitation of Elderly Patients After Hip Replacement Surgery: Pilot Study. JMIR Serious Games. 2017;5(4):e19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Fatoye F, Gebrye T, Mbada C, Useh U. Economic evaluations of digital health interventions for the management of musculoskeletal disorders: systematic review and meta-analysis. J Med Internet Res. 2023;25:e41113. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Molina-Garcia P, Mora-Traverso M, Prieto-Moreno R, Díaz-Vásquez A, Antony B, Ariza-Vega P. Effectiveness and cost-effectiveness of telerehabilitation for musculoskeletal disorders: a systematic review and meta-analysis. Ann Phys Rehabil Med. 2024;67(1):101791. [DOI] [PubMed] [Google Scholar]
- 42.Brettmann K, Vogt L, Galm R, Hartge S, Banzer W. Visual feedback gait training following total hip arthroplasty. Deutsche Zeitschrift für Sportmedizin. 2006;57(7/8):189–94. [Google Scholar]
- 43.Qi X, Jie K, Chen J, Cao H, Koch JA, Li J, et al. Post-THA gait training to improve pelvic obliquity and decrease leg length discrepancy in DDH patients: a retrospective study. J Int Med Res. 2020;48(3):300060519898034. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Schwartz MH, Trost JP, Wervey RA. Measurement and management of errors in quantitative gait data. Gait Posture. 2004;20(2):196–203. [DOI] [PubMed] [Google Scholar]
- 45.Durstberger S, Kaufmann C, Widhalm K, Klinger M, Putz P. O 108—Preliminary results of test-retest-reliability of outcomes derived from the Plug-in Gait and Cleveland Models when attached by an advanced operator. Gait Posture. 2018;65:226–7. [Google Scholar]
- 46.Penasso H, Durstberger S, Maul L, Putz P, Widhalm K. Kinematic signatures of knee joint reaction load patterns during self-paced level walking in healthy adults. Gait Posture. 2025;121:178–9. [Google Scholar]
- 47.Singh JA, Schleck C, Harmsen S, Lewallen D. Clinically important improvement thresholds for Harris Hip Score and its ability to predict revision risk after primary total hip arthroplasty. BMC Musculoskelet Disord. 2016;17:256. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Blasimann A, Dauphinee SW, Staal JB. Translation, cross-cultural adaptation, and psychometric properties of the German version of the hip disability and osteoarthritis outcome score. J Orthop Sports Phys Ther. 2014;44(12):989–97. [DOI] [PubMed] [Google Scholar]
- 49.Kessler S, Lang S, Puhl W, Stöve J. The Knee Injury and Osteoarthritis Outcome Score–a multifunctional questionnaire to measure outcome in knee arthroplasty. Z Orthop Ihre Grenzgeb. 2003;141(3):277–82. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Processed study data are available from the corresponding author upon reasonable request.