Home-based rehabilitation using wearable motion tracker with smart phone application feedback is as effective as conventional self-directed rehabilitation after total knee arthroplasty: randomized controlled trial Authors: Seok Jin Jung, Ji Hoon Kim, Seung Joon Rhee Published: 2025 ABSTRACT AbstractObjectiveTo investigate the effectiveness and pain improvement of self-directed rehabilitation (SDR) using a wearable motion tracker system (WMT) with smartphone application feedback after total knee arthroplasty.DesignA randomized controlled trial in institutional practice enrolled female patients who had undergone total knee arthroplasty. The experimental group was instructed to perform self-directed rehabilitation using the WMT device (WMT), and the control group performed self-directed rehabilitation protocol under physiotherapist supervision (SDR) according to discharge education at home with regular outpatient visit. 24 and 25 patients in the experimental and control group were analyzed. Knee ROM, WOMAC, EQ-5D-5 L, KOOS, and motion tracker gained gait speed, and gait ROM was assessed at preoperatively, and 3,6, 12 weeks postoperatively.ResultsIn the mixed-effect model analysis, significant differences between the groups were observed in WOMAC pain, stiffness, function subsets, and EQ-5D-5 L, KOOS stiffness, pain, and activities of daily living subsets. WOMAC stiffness and EQ-5D-5 L scores were lower at 3week, and KOOS stiffness scores were higher at 6week in the experimental group. Lower WOMAC and higher KOOS stiffness scores were observed at 12week in the experimental group. (95% CI for all).ConclusionsWMT-supported rehabilitation provided significant benefits in reducing postoperative pain and stiffness, and improving knee function and quality of life in the early postoperative period. Many of the results were superior to the level in conventional SDR with intermittent physiotherapist guided rehabilitation.Trial registration numberKCT0010323 Clinical Research Information Service (CRIS), Republic of Korea, 2025.03.20 : Retrospectively registered https://cris.nih.go.kr/cris/search/detailSearch.do?seq=29333%26;search_page=M%26;search_lang=%26;class_yn=.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13102-025-01374-1.Keywords: Arthroplasty, Feedback, Knee, Rehabilitation, Self-management, Wearable electronic devices Introduction Total knee arthroplasty (TKA) is a highly successful and worldwidely accepted surgical technique for the treatment of end-stage osteoarthritis (OA) [1]. TKA is known to induce significant pain relief, functional improvements, and higher quality of life [2, 3]. Immediate and sufficient rehabilitation after surgical procedures plays a key role in restoring knee function and range of motion (ROM) [4–6]. Physiotherapist-supervised programs are commonly used for the functional rehabilitation of patients who have undergone TKA. Such standard programs require certified professional therapists and specialized equipment that can be used in hospitals or outpatient facilities. Therefore, postoperative rehabilitation in TKA is potentially associated with an increased frequency of outpatient facility visits and related costs [4, 7, 8]. Therefore, there have been pursuits seeking patient-oriented alternative methods with relatively low cost and self-directed rehabilitation (SDR) [9–11]. A systematic review reported that supervised rehabilitation programs did not result in overall significant differences in outcomes compared to SDR [8]. Disadvantages of SDR include a lack of standardized protocols and the unwarranted accomplishment of protocols by patients. The compliance issue with SDR programs has always been questioned. Patients with poor SDR may exhibit poor knee ROM, which influences knee function, quality of life, and satisfaction after surgery [5, 12, 13]. To overcome these limitations of conventional SDR, we investigated the effectiveness of SDR using a wearable motion tracker system (WMT) after TKA. We hypothesized that patients who underwent self-directed home-based rehabilitation using a wearable motion tracker with a smartphone application feedback system would show better clinical outcomes than those who underwent conventional self-directed rehabilitation under physiotherapist supervision after TKA. Materials and methods This prospective randomized controlled trial was approved by the Clinical Research Ethics Committee of the authors’ hospital (IRB No.: 2011-015-097) and conducted in accordance with the approved guidelines. Informed written consent was obtained from all patients. Patients scheduled to undergo primary TKA between April 2021 and December 2023 in a single university hospital were investigated for eligibility for enrollment in this study. The inclusion criteria were female patients who were between 55 and 70 years old with a diagnosis of osteoarthritis (Kellgren-Lawrence grade 3–4). Patients with posttraumatic deformity around the knee joint, rheumatologic disease, systemic connective tissue disease, cognitive dysfunction, and history of inflammatory or infectious knee joint arthritis or previous surgery around knee were excluded. The age limitation of the maximum 70 years old was set considering the capability of smartphone application handling. For blinding and allocation concealment, enrolled patients were randomly allocated into two groups according to a computerized random allocation sequence using Microsoft Excel (USA), separately from the operator and rater. Assessment in the outpatient clinic was performed by orthopaedic resident doctor who were blinded to group allocation. Sample size calculation and power analysis The primary outcome for sample size determination was the change in the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) total score from baseline to 12 weeks postoperatively. Based on previous data and clinically meaningful thresholds, we assumed a minimal clinically important difference of 10 points (SD 12) in the WOMAC score between groups. Using a two-sided significance level (α = 0.05) and a power of 80% (1 − β = 0.80), the required sample size per group was calculated via the following formula: where Zα/2 = 1.96, Zβ = 0.84, Δ = 10, and σ = 12, resulting in a minimum of 23 participants per group. Allowing for a 10% dropout rate, a target of 25 patients for each group was set. An orthopedic surgeon with over 10 years of knee surgery and arthroplasty experience performed all the surgeries. The operative procedure was performed through medial parapatellar incision and followed a modified measured resection policy, using posterior cruciate ligament substituting type implant in both groups. Distal femoral resection was performed independently using an intramedullary guide, followed by proximal tibial resection using an extramedullary guide. Bone cement was used for implant fixation. Two closed suction drain was placed until the first postoperative day. The drains were clamped until four hours after the operation and the clamp was half released afterwards. Blood transfusion was not performed in any of the cases in both groups. One group of patients (experimental group, WMT) was directed to use a wearable motion tracker system and receive feedback from a smartphone application (app) during the outpatient rehabilitation period, which started on the seventh postoperative day. The other group (control group, SDR) was directed to perform the conventional postoperative rehabilitation protocol according to discharge education from the seventh postoperative day at home and subject to intermittent physiotherapist supervision during their follow-up outpatient visit. Before the discharge, patients in both groups performed identical acute postoperative recovery rehabilitation for seven days. Patients in the experimental group were educated to use the wearable motion tracker device eCEN Care (eCEN Inc., Seoul, Republic of Korea) for home-based rehabilitation (Fig. 1 Wearable motion tracker and gait analysis system, eCEN Care). This system consists of two internal motion trackers that utilize a gyrosensor and a multi-axial accelerometer with a Bluetooth communication module to measure and send information regarding knee range of motion and gait pattern to a mobile app in the smartphone (Fig. 1. Wearable motion tracker and gait analysis system, eCEN Care). The mobile app provides at-home monitoring for patients and a web dashboard for medical staff to review the assessment results. The tracker was attached to the lateral side of the thigh and calf along the anatomical axis of the femur and tibia, approximately 10 cm from the joint line, using an elastic band and a tracker hook. Each sensor digitized the patient’s motion and provided real-time feedback on performance through a mobile app. The app also includes a web-based platform that allows the clinical team to prescribe and monitor from a remote location (Fig. 2 Mechanical protocol of gait analysis signal transmission in the wearable tracker module). This system helped patients perform independent rehabilitation at home without the need for other supervision and complete the rehabilitation program. The standard rehabilitation protocol included calf-pumping, quadriceps muscle strengthening, and straight leg raise exercises from the first postoperative day to the 12th postoperative week. Knee range of motion (ROM) recovery was started from the 2nd postoperative day using a continuous passive motion machine and hanging-and-dangling method to the discharge day and continued by heel-slide and towel-pulled passive bending during the outpatient period. Straight leg lateral raise and hamstring strengthening exercises were started from the first postoperative week. Performing three sets of 20 repetitions of each exercise in a day was regarded as the completion of a single-day rehabilitation protocol. Experimental group and control group patients were encouraged to perform the same rehabilitation protocol during the seven-day in-patient period. After discharge, experimental group patients were encouraged to watch and repeat after the mobile application-provided rehabilitation protocol while wearing the WMT device at home, and control group was directed to perform timely rehabilitation on their regular outpatient visit. Wearable motion tracker and gait analysis system, eCEN Care Mechanical protocol of gait analysis signal transmission in the wearable tracker module Functional outcomes were assessed using knee ROM measurement and gait pattern analysis, which were collected through individual outpatient visits at selected time points (preoperatively, 3,6 and 12 weeks after surgery). For patient-reported outcome measurement, patients from both groups were assessed using the validated Korean version of Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) [14], EuroQol five-dimension five-level (EQ-5D-5 L) [15], and Knee injury and Osteoarthritis Outcome Score (KOOS) [16]. Statistical analyses were performed using SPSS version 23.0 (IBM Corp., Armonk, NY, USA). All continuous data were first tested for normality using the Shapiro-Wilk test. Differences in demographic and preoperative baseline variables between groups were assessed using independent t-tests or Mann-Whitney U tests as appropriate, and categorical variables were compared with the chi-square test. The primary effectiveness analysis utilized mixed-effect model analysis to evaluate changes in WOMAC scores (and secondary outcomes such as EQ-5D-5 L and KOOS) over time and between groups. If significant main or interaction effects were present, Bonferroni-adjusted post-hoc tests were applied for all pairwise comparisons to control the family-wise Type I error rate, as recommended in recent statistical guidelines. Sensitivity analyses were performed using mixed-effects linear models to adjust for potential confounders, including age, BMI, and relevant comorbidities. All statistical tests were two-sided, and results were considered statistically significant at p < 0.05. Results Sixty eligible patients were identified and they agreed to participate in the study. Thirty patients were randomly assigned to the control group (group SDR) and the experimental group (group WMT). 25 patients in the control group and 24 in the experimental group were included in the final analysis (Fig. 3 CONSORT diagram). The mean ages of the patients in the control group (25 female) and experimental group (24 female) were 67 and 66 years, respectively. The mean body mass index (BMI) was 26.05 and 25.12, and the ROM was 116.80 and 113.33 for each group, respectively. There was no statistically significant difference in patient characteristics, including body mass index and preoperative knee ROM. All preoperative patient-reported outcome measurement scores were similar between the two groups (Table 1). Over the course of the 12-week postoperative follow-up, all measured outcomes showed significant improvement. However, experimental group experienced consistently greater gains, both in the short term and at the end of the observation period. Patients in the experimental group reported lower stiffness scores on the WOMAC scale across all timepoints, indicating less joint stiffness compared to the control group. Notably, the reduction in stiffness was already evident by week 3 and remained superior at week 12 (interaction coefficient for group × time at 3 and 12 weeks both statistically significant, p < 0.001). Similarly, the WOMAC function scores improved more rapidly in the experimental group, with participants achieving better knee-related function (coefficient for group effect − 3.8, p = 0.001). Quality of life, as assessed by the EQ-5D-5 L, also showed a more favorable course of results in the experimental group with statistical analysis indicating a significant overall benefit (coefficient for group effect − 0.81, p = 0.017). Improvements were also seen in the KOOS subscales, where the experimental group consistently outperformed the SDR group. Particularly strong effects were observed in the pain, activities of daily living (ADL), and overall knee function domains, with patients in the experimental group reporting less pain (group effect coefficient 5.28, p < 0.001) (Table 2). In the comparison at each time point for these variables, WOMAC stiffness and EQ-5D-5 L scores were statistically lower in the experimental group at 3 weeks, and the KOOS stiffness score was higher in the experimental group at 6 weeks after TKA. The WOMAC stiffness score was significantly lower, and the KOOS stiffness score was higher in the experimental group at 12 weeks after TKA (Table 3; Fig. 4 − 1 and 4 − 2 Evolution of the outcomes over time in both groups, based on the mixed-effect model). Gait speed in the two groups was similar from the preoperative period to the 12th week. However, the range of knee joint motion during gait was significantly larger in the experimental group as 55.34 ± 10.79° compared with 47.89 ± 10.42° in control group at the 12th week(p = 0.037) (Table 4; Fig. 5 Gait speed change in the individual patient(A) Comparison of mean gait speed change between control and study group(B) Change in range of motion during the gait in the individual patient(C) Comparison of mean gait range of motion change between control and study group(D)). CONSORT diagram of this study Table 1Preoperative patient demographics and patient reported outcome measurementsParameterControl group (n = 25)Study group (n = 24)p-valueAge (years)66.16 ± 3.71667.00 ± 2.7980.552Sex Female2524 Male00Side0.458 Right129 Left1315BMI26.05 ± 3.0225.12 ± 2.830.308BMD-1.34 ± 0.928-1.63 ± 1.2740.356Range of motion (°)116.80 ± 3.340113.33 ± 21.550.531WOMAC Pain11.82 ± 0.7010.08 ± 0.860.147 Stiffness4.84 ± 0.0393.92 ± 0.500.152 Function44.20 ± 2.2737.59 ± 2.760.070EQ-5D-5 L15.48 ± 0.7714.67 ± 0.620.418KOOS Pain48.96 ± 3.3254.24 ± 3.700.293 Stiffness58.12 ± 3.5960.38 ± 4.060.677 Sport function16.60 ± 3.3019.38 ± 3.620.572 Activities of daily living47.37 ± 3.4756.27 ± 3.250.068 Quality of life37.98 ± 2.5242.43 ± 2.470.214BMI: body mass index; BMD: bone mineral density Preoperative patient demographics and patient reported outcome measurements BMI: body mass index; BMD: bone mineral density Table 2Inter-group comparison for ROM, WOMAC, EQ-5D-5 L, and KOOS scores using mixed-effect model analysisOutcome\EffectROMWOMACEQ-5D-5 LKOOSPainStiffnessFunctionPainStiffnessFunctionADLQOLIntercept (SDR, Preop)116.8011.154.8444.2015.4848.9658.1250.0047.3737.98Group [WMT]-3.47 (0.198)-1.49 (0.061)-0.92 (< 0.001)-3.80 (0.001)-0.81 (0.017)5.28 (< 0.001)2.26 (0.148)3.50 (0.004)8.90 (< 0.001)4.45 (0.001)Time [3wks]-21.80 (< 0.001)-4.44 (< 0.001)-1.50 (< 0.001)-15.10 (< 0.001)-3.15 (< 0.001)14.99 (< 0.001)18.12 (< 0.001)15.20 (< 0.001)20.77 (< 0.001)25.24 (< 0.001)Time [6wks]-14.50 (< 0.001)-7.92 (< 0.001)-2.92 (< 0.001)-24.45 (< 0.001)-5.30 (< 0.001)29.22 (< 0.001)23.84 (< 0.001)26.90 (< 0.001)37.34 (< 0.001)33.99 (< 0.001)Time [12wks]-11.60 (< 0.001)-9.92 (< 0.001)-3.28 (< 0.001)-28.80 (< 0.001)-6.30 (< 0.001)38.66 (< 0.001)29.26 (< 0.001)34.50 (< 0.001)45.71 (< 0.001)42.25 (< 0.001)Group[WMT] × Time[3wks]4.00 (0.258)0.57 (0.595) -0.64 (< 0.001)* 1.15 (0.441)-0.40 (0.328)2.32 (0.154)2.59 (0.167)3.10 (0.053) 4.32 (0.013)* 3.00 (0.078)Group[WMT] × Time[6wks]6.50 (0.067)0.63 (0.555)-0.34 (0.059)1.85 (0.218)-0.25 (0.539) 4.77 (0.003)* 5.00 (0.007)* 4.45 (0.006)* 3.81 (0.029)* 4.69 (0.006)* Group[WMT] × Time[12wks]5.20 (0.136)1.34 (0.209) -0.64 (< 0.001)* 2.20 (0.143)-0.55 (0.181)2.17 (0.182) 4.70 (0.012)* 2.90 (0.075)1.55 (0.376)2.48 (0.145)Age (per year)-0.20 (0.323)-0.04 (0.562)-0.01 (0.617)-0.15 (0.213)-0.03 (0.455)-0.08 (0.504)0.11 (0.432)-0.12 (0.360)-0.16 (0.283)-0.13 (0.320)BMI (per unit)-0.70 (0.260)0.13 (0.113)-0.02 (0.505)-0.25 (0.121)-0.07 (0.210)-0.01 (0.942)0.06 (0.737)-0.09 (0.618)-0.08 (0.673)-0.17 (0.349)SDR: control group; WMT: experimental group; wks: weeks; ADL: activities of daily living; QOL: quality of life; (Coefficient [p-value]) Inter-group comparison for ROM, WOMAC, EQ-5D-5 L, and KOOS scores using mixed-effect model analysis SDR: control group; WMT: experimental group; wks: weeks; ADL: activities of daily living; QOL: quality of life; (Coefficient [p-value]) Table 3Postoperative outcomes of control group and experimental groupOutcomeControl group (n = 25)Study group (n = 24)p-valueCohen’s dInterpretation Range of motion 3wks91.80 ± 2.2393.13 ± 2.480.6930.57M(favors WMT)6wks100.00 ± 2.31102.92 ± 2.290.3751.27L(favors WMT)12wks105.20 ± 2.49107.50 ± 2.680.5320.89L(favors WMT) WOMAC Pain3wks7.28 ± 0.516.21 ± 0.650.198-1.84L(favors WMT)6wks3.80 ± 0.532.79 ± 0.280.102-2.37VL(favors WMT)12wks1.80 ± 0.271.50 ± 0.420.549-0.85L(favors WMT) Stiffness3wks3.12 ± 0.242.42 ± 0.22 0.035* -3.04 VL(favors WMT)6wks1.92 ± 0.221.58 ± 0.130.206-1.87L(favors WMT)12wks1.56 ± 0.210.92 ± 0.18 0.024* -3.27 VL(favors WMT) Function3wks24.52 ± 1.8621.08 ± 2.140.231-1.72L(favors WMT)6wks10.99 ± 1.857.587 ± 1.090.135-2.02L(favors WMT)12wks4.44 ± 0.653.54 ± 1.410.561-0.76M(favors WMT) EQ-5D-5 L 3wks12.72 ± 0.3511.55 ± 0.46 0.044* -2.92 VL(favors WMT)6wks9.84 ± 0.499.29 ± 0.370.380-1.25L(favors WMT)12wks8.08 ± 0.317.13 ± 0.380.056-2.98VL(favors WMT) KOOS Pain3wks63.95 ± 1.9067.20 ± 2.390.2910.58M(favors WMT)6wks78.176 ± 2.2582.95 ± 1.480.0862.18VL(favors WMT)12wks87.62 ± 1.2589.79 ± 2.040.3651.19L(favors WMT) Stiffness3wks76.24 ± 1.6778.83 ± 1.940.6771.05L(favors WMT)6wks81.96 ± 1.8886.96 ± 0.76 0.022* 2.95 VL(favors WMT)12wks87.38 ± 1.392.08 ± 1.31 0.016* 3.61 VL(favors WMT) Sport function3wks37.00 ± 2.5536.25 ± 2.580.8370.15S(favors WMT)6wks47.00 ± 2.2052.29 ± 2.570.1231.03L(favors WMT)12wks57.80 ± 3.1261.67 ± 2.570.3461.31L(favors WMT) Activities of daily living3wks68.14 ± 1.7281.46 ± 2.630.2922.93VL(favors WMT)6wks84.71 ± 1.9688.52 ± 1.820.1632.01VL(favors WMT)12wks93.08 ± 0.9894.63 ± 1.900.4681.06L(favors WMT) Quality of life3wks63.22 ± 2.2366.12 ± 2.250.3650.41S/M(favors WMT)6wks71.97 ± 2.1176.80 ± 1.820.0912.43VL(favors WMT)12wks80.23 ± 1.4782.00 ± 2.140.4950.95L(favors WMT)S: small, M: medium, L: large, VL: very large, S/M: small to medium Postoperative outcomes of control group and experimental group S: small, M: medium, L: large, VL: very large, S/M: small to medium Evolution of the outcomes over time in both groups, based on the mixed-effect model Table 4Gait related values of control group and experimental group Gait Speed(m/s) Control group( n = 25) Study group( n = 24) P -Value Cohen’s d Interpretation Preop1.10 ± 0.111.08 ± 0.090.559 3week1.06 ± 0.091.05 ± 0.110.6830.56M(favors WMT) 6week1.12 ± 0.091.13 ± 0.110.6340.50M(favors WMT) 12week1.12 ± 0.101.11 ± 0.110.7560.69M/L(favors WMT) Gait ROM(°) Control group( n = 25) Study group( n = 24) P -Value Preop38.53 ± 12.5043.44 ± 14.810.227 3week34.12 ± 12.8441.10 ± 11.530.1430.10Negligible 6week43.93 ± 12.0050.09 ± 13.260.098-0.11Negligible 12week47.89 ± 10.4255.34 ± 10.790.0370.10NegligibleROM: range of motion; M: medium, M/L: medium to large Gait related values of control group and experimental group ROM: range of motion; M: medium, M/L: medium to large Gait speed change in the individual patient (A) Comparison of mean gait speed change between control and study group (B) Change in range of motion during the gait in the individual patient (C) Comparison of mean gait range of motion change between control and study group (D) Discussion The findings of this study suggest that self-directed home-based rehabilitation with wearable motion tracking technology leads to enhanced outcomes in post-TKA recovery compared with traditional SDR. Our hypothesis was proven by the key results that the experimental group (WMT) showed superior improvements in pain reduction, knee stiffness, and overall function and quality of life, particularly in the first 12-week period. The results of this study suggest that the wearable motion tracker combined with app feedback provided patients with more precise and immediate feedback, which might contribute to better adherence and optimized exercise performance. The experimental group achieved significantly lower WOMAC stiffness scores (Table 3), indicating reduced stiffness compared to the control group. Similarly, the EQ-5D-5 L score showed notable improvement at 3 weeks in the experimental group (p = 0.044). These findings underscore the efficacy of WMT-based rehabilitation in the early postoperative phase, where stiffness and discomfort are the most impactful. The ROM during gait was notably greater in the experimental group by the 12th week (55.34° ± 10.79°) than in the control group (47.89° ± 10.42°, p = 0.037) (Table 4). Importantly, the data suggest that these benefits are not only statistically significant, but also likely to have meaningful clinical impact in daily life—particularly in the critical early and intermediate phases of postoperative rehabilitation. The greater and more rapid improvements observed in the WMT group support the integration of wearable motion tracker-based feedback systems into standard postoperative care to enhance both symptom control and functional restoration after knee replacement surgery. These findings are in accordance with those of previous studies suggesting that SDR alone can achieve results comparable to those of supervised rehabilitation programs. López-Liria et al. [17] found no significant difference between home- and hospital-based rehabilitation programs for improving knee function after TKA. Xu et al. [18] reported that a self-developed home-based exercise program was not inferior to standard supervised PT in improving mobility and quality of life, with added economic benefits and time savings for patients. It has been reported that even unsupervised home exercises are non-inferior to outpatient PT in terms of knee flexion recovery after TKA [19]. The lack of significant differences between the supervised and unsupervised groups underscores the potential for home-based wearable systems to replace supervised PT, particularly for patients with adequate self-motivation and support. The integration of wearable technology and mobile applications for post-TKA rehabilitation has shown promising outcomes across multiple studies, supporting our finding that these tools enhance patient engagement and clinical results in home-based programs. Ramkumar et al. [20] validated the feasibility of remote patient monitoring systems using wearable technology, showing that patients maintained high compliance and mobility levels postoperatively with significant improvements in PROMs and reduced opioid use. This resonates with our study results, in which patients in the wearable-based group demonstrated better adherence to the rehabilitation protocol. Correia et al. [21] reported that digital biofeedback systems could achieve superior outcomes compared to traditional in-person rehabilitation, with advantages in knee ROM and PROMs such as the timed up-and-go test. Our findings reinforce the effectiveness of real-time feedback, as the wearable motion tracker system in our study also improved knee ROM and functional outcomes, likely because of enhanced patient engagement through continuous feedback. Osterloh et al. [22] explored digital-assisted group rehabilitation and found improvements in the quality of life, pain, and physical function after TKA, suggesting that digital tools can complement or even replace traditional rehabilitation programs for some patients. The real-time individualized feedback provided by wearable trackers in our study parallels these findings, promoting faster functional gains and higher patient satisfaction. Recent studies have demonstrated that a smartwatch paired with a mobile application could serve as an effective alternative to formal physical therapy (PT) after TKA, showing comparable outcomes in range of motion (ROM) and patient-reported outcome measures (PROMs), such as KOOS and EQ-5D-5 L [23]. This aligns with our study, in which wearable technology effectively supported self-directed rehabilitation, potentially reducing the need for supervised PT and the associated costs. The use of wearable technology and digital feedback systems for rehabilitation after TKA has been increasingly validated as a viable, cost-effective, and efficient alternative to supervised PT. These systems not only offer substantial improvements in pain and stiffness, but also address logistical challenges associated with traditional rehabilitation methods, making high-quality care accessible for a broader patient population. Our study extends this observation, demonstrating that wearable motion trackers provide an effective alternative for conventional intermittent physiotherapist-supervised self-directed rehabilitation, especially in the early recovery stages, enhancing quality of life, and reducing postoperative stiffness. There are limitations in this study, although it demonstrated some improvements in early postoperative outcomes. First, the study was performed in a small patient group. To overcome this limitation, we applied strict and narrow inclusion criteria and randomization. Second, this study reported relatively short-term outcomes in the 12th postoperative week. Despite three months is a good period to observe acute postoperative recovery after TKA, functional recovery may persist until one year after surgery. We believe that the long-term effects and integrating power analysis in the study design could further substantiate these findings. Third, the sample population was relatively cooperative and limited to the age range of 55–70 years. The restriction of age group and limited female sex was a policy that considered smartphone friendliness and incidence of TKA. However, further improvements in the user interface and voice recognition technology will help overcome this limitation. Future studies should include a broader demographic group to determine if similar outcomes can be achieved in a less compliant or more diverse patient population. Conclusion The results of this study suggest that WMT-supported SDR can provide significant benefits in reducing postoperative pain and stiffness, and improving knee function and quality of life, particularly in the early postoperative period. As such, wearable motion trackers may serve as viable alternatives to self-directed rehabilitation with intermittent physiotherapist supervision, potentially reducing healthcare costs and improving accessibility for patients undergoing TKA. Electronic supplementary material Supplementary file Acknowledgements This study was supported by the 2024 overseas training grant from Pusan National University Hospital. Author contributions Author SeokJin, Jung and author Seung Joon, Rhee made substantial contributions to the conception or design of the manuscript, author JiHoon, Kim, for acquisition, analysis, and interpretation of the data. All authors participated in drafting the manuscript. All authors have contributed equally to the manuscript and have read and approved the final version of the manuscript. Funding No funding was received. Data availability The data associated with the paper is available from the corresponding author upon reasonable request. Declarations Reporting This study was conducted and reported in accordance with the Consolidated Standards of Reporting Trials (CONSORT) guidelines. A CONSORT checklist is provided as supplementary material. Ethics approval and consent to participate This study was approved by the Institutional Review Board of Pusan National University hospital (2011-015-097). Informed consent was obtained from all patients. This study was carried out with strict adherence to the Declaration of Helsinki. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary file Data Availability Statement The data associated with the paper is available from the corresponding author upon reasonable request.