Remote, Reliable, Repeatable: Real-World Test–Retest Validation of Hand Grip Strength Assessments

Remote, Reliable, Repeatable: Real-World Test–Retest Validation of Hand Grip Strength Assessments | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Remote, Reliable, Repeatable: Real-World Test–Retest Validation of Hand Grip Strength Assessments Sharah Abdul Mutalib, Daniel Jenkinson, Sonia Pike, Etienne Burdet, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6801648/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Handgrip strength (HGS) is a key indicator of health and functional status. As remote health assessments become more common, it is critical to understand how procedural supervision influences the reliability of remote HGS assessment. Objective To evaluate the test-retest reliability and measurement precision of remote HGS assessment under varying levels of procedural supervision as found in real world use. Methods Seventy-two adults were randomised into six groups reflecting different supervision levels over two sessions. HGS was measured using the GripAble Sensor and Able Assess platform. Test–retest reliability was evaluated using intraclass correlation coefficients (ICC), while measurement precision was quantified using the standard error of measurement (SEM) and minimal detectable change (MDC%). Agreement between sessions was further assessed using Bland–Altman analysis, reporting mean difference and 95% limits of agreement (LoA). Protocol compliance was rated from video recordings. Participants also completed a post-session questionnaire on remote assessment experience. Results All groups demonstrated a good-to-excellent test-retest reliability (ICC ≥ 0.93, ICC lower bounds ≥ 0.73), but measurement precision varied. Fully supervised groups achieved the lowest MDC% (as low as 8.5%), while unsupervised groups often exceeded 20% in the single trial reporting approach, indicating reduced sensitivity to change. Higher supervision corresponded with better protocol compliance. Participant feedback demonstrated high usability during real-world use: 97% rated the test as easy or very easy, > 75% felt comfortable performing it remotely, and > 95% were satisfied with the experience. Conclusion Remote HGS assessment shows high reliability, but measurement precision is shaped by supervision and procedural compliance. Based on these findings, it is recommended that to maximise measurement precision during remote sessions, in-person supervision should be provided during onboarding, possibly followed up with periodic supervision when conducting repeated longitudinal measurements of an individual. Integrating structured features, including standardised instructions, user specified configuration and compliance monitoring will further improve remote measurement performance without undermining usability during real-world use. Handgrip strength dynamometry real world remote test-retest measurement reliability precision Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Handgrip strength (HGS) has emerged as a robust biomarker of health status. It reflects overall muscular strength and is easily measured, serving as a convenient indicator of an individual’s functional capacity including the ability to perform activities of daily living (ADLs). HGS is particularly relevant in the context of aging, as it declines with advancing age and mirrors the process of sarcopenia – the progressive loss of muscle mass, strength and function in older adults. Low HGS also correlates with greater functional decline – including higher odds of disability and hospitalization – and significantly higher all-cause mortality rates (Vaishya et al. 2024 ). Numerous studies report that individuals with weaker HGS face elevated risks of aging-related diseases such as type 2 diabetes, cardiovascular disease, stroke, respiratory disease, chronic kidney disease, and certain cancers (Li et al. 2016 ; Chang et al. 2011 ; Celis-Morales et al. 2018 ). A large cohort analysis has found that each ~ 5 kg decrement in HGS corresponds to a substantial increase in all-cause and cardiovascular mortality risk, and HGS can outperform some traditional risk factors (e.g. blood pressure) in predicting long-term survival (Leong et al. 2015 ). A recent systematic review of 2.4 million adults across 69 countries confirmed age-related decline in HGS and established global norms, reinforcing its value as a benchmark for identifying low muscle strength and guiding early intervention (Tomkinson et al. 2024 ). Despite its clinical value, HGS remains underused in routine care. Barriers include limited access to appropriate equipment, the need for trained personnel, and uncertainty around how to interpret results in everyday workflows. HGS assessment has been shown to be reliable in clinical settings where it is performed under supervision. Expanding HGS assessment beyond the clinic and into remote settings could improve accessibility and enable scalable adoption of HGS across the care continuum. However, it also raises concerns about standardisation, patient compliance, and data quality in real-world settings. Health assessment outside clinical settings is not new, but has recently gained momentum across clinical domains such as geriatrics, neurology, cardiology, primary care, and mental health, driven in part by the recent COVID pandemic. Examples include video-based physical function tests (e.g. chair stand, gait speed), digital cognitive screening, remote vital sign monitoring, and telepsychiatric evaluations, including in cancer patients (Hoenemeyer et al. 2022 ) and stroke patients (Naef et al. 2025 ). Many approaches are supported by emerging evidence: remote chair stand tests have shown strong validity and test-retest reliability in healthy adults (Klein et al. 2025 ), though findings in older or frailer populations remain mixed (Heslop et al. 2023 ). Similarly, cognitive assessments delivered via telehealth, such as the Montreal Cognitive Assessment (MoCA) score, are comparable to in-person assessment, even among older adults (Loring, Lah, and Goldstein 2023 ). Nonetheless, several challenges persist. Many tools still lack large-scale validation, and alignment with in-clinic results is not always consistent. Discrepancies can arise from home-specific factors such as environmental distractions, inconsistent setup, and the absence of real-time guidance. Additional influences include the cognitive demands of self-directed assessment, unclear instructions, and whether the task is volitional or autonomic. The lack of the white coat effect – enhanced performance under observation – may also reduce engagement. Digital literacy barriers, particularly among older or underserved populations can compound these challenges. Clinicians also frequently cite concerns about patients’ ability to use remote tools effectively (Klein et al. 2025 ; Hailu et al. 2024 ). Even when validated, factors including, but not limited to, long-term compliance, user engagement, cognitive status and technology-related frustrations can limit the integration of remote assessment into routine care (Serrano et al. 2023 ). Against this backdrop, there is growing interest in evaluating whether HGS – a clinically meaningful and easily interpretable physical measure – can be effectively assessed in remote contexts using digital tools. To realise HGS’s utility as a widely adopted biomarker in decentralised care, it is critical to establish both the reliability of remote assessment and the accuracy with which participants follow protocols independently. Without such evidence, HGS will remain a validated yet underutilised metric in routine practice. To address this gap, we evaluated the test-retest reliability and protocol compliance of remote HGS assessment as a model for decentralised assessment. In this study, we used the Able Assess platform – a digital assessment platform combining a handheld device (GripAble) and app interface to guide and capture standardised HGS remotely. Our aim was to examine whether different remote assessment conditions affect (1) the consistency of HGS across repeated sessions and (2) participant compliance to the protocol. By simulating real-world use scenarios, this study explores the potential for remote HGS assessment to generate clinically valid and operationally scalable data, offering a blueprint for future decentralised measurement tools. Methods This parallel-group test-retest study assessed the reliability and protocol compliance of remote HGS assessments. Assessment was delivered using the Able Assess platform, which integrates the GripAble handheld sensor with an app-based interface, to standardise remote delivery of HGS assessments. To reflect real-world clinical settings, the study included varying assessment conditions: (1) fully independent self-assessment, (2) remote supervision via video, and (3) face-to-face guidance. The study followed the ethical standards of the 1964 Declaration of Helsinki and received approval from the Imperial College Research Ethics Committee and the Science, Engineering & Technology Research Ethics Committee (REF: 20IC5831). Able Assess Platform Able Assess is a digital assessment platform that integrates the GripAble sensor with an app-based interface to guide users through digitised assessments (e.g. HGS, timed-up-and-go, chair stand, etc.) and securely capture sensor data for analysis (Fig. 1 A). In this study, Able Assess (version 1.0.0) was used to support consistent protocol delivery, rather than to evaluate the platform itself – serving as a practical example to promote future generalisability. The GripAble sensor is a handheld digital dynamometer with dual load cells, offering ± 1.8 kg (± 2%) accuracy in isometric mode and ± 0.9 kg (± 1%) accuracy in isotonic mode across its 90 kg range, with sensitivity below 100 g (Mace et al. 2022 ). It has been fully validated for use in supervised clinical settings (Mutalib et al. 2022 ; Ergen, Kudin, and McGee 2024; Kudin, Ergen, and McGee 2025). It includes an IMU (accelerometer, gyroscope, magnetometer) and samples data at 50 Hz. For this study, the grip plate was used in isometric mode to align with standardised clinical protocols and existing normative datasets (Mutalib et al. 2024 ). The device connects via Bluetooth to the Able Assess app, which standardises assessment administration and records grip and motion data. Calibration of the GripAble handgrip was confirmed pre- and post-study using known weights, with accuracy within ± 2% across the measurement range. Participants Seventy-two participants were recruited into the study through convenience sampling across university campuses, community organisations, and personal networks. Participants were stratified by gender (36 males, 36 females) and age group (18–40, 41–60, and 61 + years), with 12 participants per gender within each age band. Eligible participants were aged 18 years or older, had no prior use of the GripAble device or similar hand dynamometers, and reported no history of upper limb musculoskeletal or neurological conditions, including pain in the hand, wrist, or forearm. Individuals not meeting these criteria were excluded. All participants provided written informed consent before enrolment. Experiment Protocol Hand Grip Strength Test-Retest Setup Figure 1 A and 1 B show the Able Assess platform and on-screen instructions to guide participants through the assessment process, respectively. The full protocol and group allocations are illustrated in Fig. 1 C. Participants were allocated to one of six assessment conditions using a 3×2 factorial design, capturing two sequential sessions: Session 1 (onboarding and initial assessment) and Session 2 (retest), conducted at least one week apart (Fig. 1 C). The design reflected (1) three onboarding assessment conditions during Session 1: face-to-face (in-person) supervision, video call supervision, or independent unsupervised onboarding via the app; and (2) two remote assessment conditions during Session 2: independent (unsupervised) or video-supervised administration of the HGS assessment. Supervised participants received live instructions by trained personnel (either face-to-face or via video call, depending on their group assignment), while independent users followed on-screen instructions. No coaching or technique correction was provided for the independent group, while those in the supervised groups received guidance as needed. During assessment, participants were instructed to be seated upright with feet flat on the floor, hips and knees at 90°, and the assessed arm adducted with the elbow at 90°, forearm in neutral position, and wrist slightly extended, following the American Society of Hand Therapists (ASHT) guidelines (Fig. 1 C). For clarity, each group is labelled using a two-part format that reflects the assessment conditions in Session 1 and Session 2, respectively: IND = independent /unsupervised, VID = video-supervised, F2F = face-to-face supervised. For example, the IND_VID group completed Session 1 independently and Session 2 supervised via video link. Video-Based Protocol Compliance Assessment Participants who consented were video recorded during the HGS assessment sessions. Assessments were conducted in line with ASHT guidelines for posture and positioning. A 12-item compliance scoring framework was independently developed to reflect key procedural elements: (1) participant seated, (2) seated posture, (3) number of trials, (4) device motion, (5) device positioning, (6) shoulder flexion, (7) shoulder adduction, (8) shoulder rotation, (9) elbow flexion, (10) forearm rotation, (11) wrist flexion, and (12) wrist deviation. Each criterion was scored on a three-point ordinal scale: 0 = not followed, 1 = partially followed, and 2 = fully followed, yielding a maximum total score of 24 per session. All ratings were completed by a single trained therapist to ensure consistency. Full scoring definitions are provided in Appendix A. Participant Experience Questionnaire At the end of Session 2, all participants (N = 72) completed a 10-item, five-point Likert questionnaire assessing: (1) clarity of app setup instructions, (2) need for additional assistance, (3) clarity of assessment instructions, (4) ease of using the GripAble device, (5) preparedness to follow the full protocol, (6) any setup- or assessment-related difficulties, (7) comfort performing the assessment without supervision, (8) perceived accuracy of strength capture, (9) home environmental factors, and (10) overall satisfaction. Appendix B shows the questionnaire items and available responses. Data Analysis Hand Grip Strength Data Participant demographics, including age, gender, self-reported handedness, height, weight, BMI, hand dimensions, and weekly exercise hours, were summarised using descriptive statistics stratified by gender and overall. To evaluate HGS performance, three commonly used reporting approaches were compared: the first trial (Trial 1), the average of three trials (Mean of 3), and the maximum of three trials (Max of 3). Normality of HGS distributions was assessed using the Shapiro–Wilk test for each session, hand, and reporting approach. HGS differences between assessment sessions were assessed using either a paired t-test or the non-parametric Wilcoxon signed-rank test, depending on the outcome of the normality test. Analyses were stratified by group, hand, and reporting approach. All statistical tests were two-tailed, with significance defined as p < 0.05. Implausible values were excluded from analysis to avoid skewing results. All analysis was performed in Python 3. To determine the consistency and precision of repeated measurements, test–retest reliability was assessed using the intraclass correlation coefficient (ICC) (Atkinson and Nevill 1998 ; Koo and Li 2016). ICC(2,1), a two-way random-effects model for absolute agreement, was selected in line with (Koo and Li 2016), with interpretation based on the lower bound of the 95% confidence interval to guard against overestimation. Measurement error was quantified via the standard error of measurement (SEM) and minimal detectable change (MDC), where SEM was calculated as: SEM = SD_pooled × √(1 – ICC), with SD_pooled being the square root of the average of the squared standard deviations from both sessions. MDC was computed as: MDC = 1.96 × √2 × SEM. SEM and MDC were computed using the point estimate of ICC to reflect true-score variability and detect the smallest meaningful change, respectively. Both metrics were also expressed as percentages of the group mean to aid comparability (SEM%, MDC%). To visualise individual-level agreement and detect systematic bias, Bland–Altman plots were generated for each group, hand, and reporting approach, alongside mean difference and 95% limits of agreement (LoA) calculations. Finally, to explore whether supervision influenced measurement fidelity, each group was assigned a supervision score (0–3), reflecting procedural oversight across onboarding and retesting phases: 0 = none (IND_IND), 1 = partial supervision during retesting (IND_VID), 2 = partial supervision during onboarding (VID_IND, F2F_IND), and 3 = fully supervised (VID_VID, F2F_VID). Relationships between supervision level and key precision metrics (MDC%, absolute mean difference, and LoA width) were examined using Spearman’s rank correlation (ρ), to test the hypothesis that greater supervision would reduce variability and improve assessment consistency. Video-Based Protocol Compliance Data To evaluate protocol compliance, scores were derived from video-based ratings across both sessions. For each participant, the mean compliance score was calculated when both sessions were available; otherwise, the single available session was used. Participants without any video data were excluded. Within-subject changes in compliance between Session 1 and Session 2 were evaluated using either a paired t-test or the non-parametric Wilcoxon signed-rank test, depending on the results of the Shapiro–Wilk test for normality. For this test, only participants with scores from both sessions were included To investigate whether higher compliance aligned with more structured supervision, Spearman’s rank correlation was used to assess the relationship between compliance scores and supervision scores. Additionally, participants were re-stratified into data-driven compliance groups – 0 = low ( 22) – rather than relying on predefined supervision-based categories (e.g., IND_IND, F2F_VID). These compliance-based groups were then used to re-evaluate test–retest reliability and measurement precision (ICC, mean difference, LoA, SEM, MDC), allowing us to examine whether actual protocol compliance, rather than nominal supervision level, better explained measurement quality. Participant Experience Questionnaire Responses were summarised by question type. Likert-style items were reported as the percentage of participants selecting each response. Multiple-selection questions were summarised by the percentage of participants selecting each unique combination of responses. Open-ended responses were categorised thematically, and category frequencies were reported. Results Test-Retest HGS Results Table 1 presents the demographic characteristics of participants, stratified by gender. Overall (N = 72) Male (N = 36) Female (N = 36) Age (years) 49.6 [18–82] 48.9 [18–82] 50.4 [18–82] Handedness (self-reported) R: 68, L: 4 R: 35, L: 1 R: 33, L: 3 Height (cm) 172.6 [155–190] 179.3 [170–190] 165.9 [155–175] Weight (kg) 72.3 [53–100] 78.7 [65–100] 65.8 [53–98] BMI (kg/m²) 24.2 [19–34] 24.5 [19–30] 23.9 [19–34] Hand Length (cm) 18.7 [17–22] 19.6 [18–22] 17.8 [17–19] Hand Circumference (cm) 20.6 [18–24] 21.7 [19–24] 19.5 [18–21] Exercise/week (hrs) 1.8 [0–4] 1.9 [0–4] 1.7 [0–4] Table 1 . Participant demographic characteristics stratified by gender. Values are presented as mean [range] for continuous variables and counts for categorical variables. BMI = body mass index; R = right; L = left. The Shapiro–Wilk test showed that HGS values were normally distributed in 9 out of 12 combinations of session, hand, and reporting approach ( p > 0.05). However, three combinations – Session 1 right hand using Trial 1 and Mean of 3 ( p = 0.023 and p = 0.026, respectively), and left hand using Mean of 3 ( p = 0.050) – showed minor deviations. Due to the relatively large sample sizes, parametric analyses were considered sufficiently robust for these tests. Test–retest reliability and measurement precision across all conditions are summarised in Table 2 , which includes ICC, SEM, and MDC values for each group, hand, and reporting approach. Figure 2 complements this table by visualising key reliability metrics across groups and reporting approaches. Overall, no significant differences in HGS were observed between Session 1 and Session 2 across hands and reporting approaches ( p > 0.05), except for the right-hand Trial 1 score, which increased from 21.83 kg (SD = 8.76) to 22.68 kg (SD = 8.84) ( t (70) = − 2.55, p = 0.013). Group-level results are presented in Table 2 . A significant increase was observed only in the F2F_IND group (right hand, Trial 1), from 21.95 kg (SD = 11.88) to 24.00 kg (SD = 10.94) ( t (11) = − 2.43, p = 0.033; indicated with *). Table 2 Test–retest reliability and measurement precision of HGS across all study groups. HGS values, SEM, and MDC are reported in kg. SEM% and MDC% are reported as percentages to aid interpretation relative to group means. * indicates a significant difference in HGS between Session 1 and 2. One trial from the IND_VID group was excluded due to an implausible reading of 0.3 kg. Group Hand N Reporting Statistics Session 1 Mean (SD) Session 2 Mean (SD) ICC (95% CI) SEM (SEM%) MDC (MDC%) IND_IND L 12 Trial 1 19.51 (6.3) 20.51 (7.0) 0.98 (0.91–0.99) 1.0 (5.01) 2.78 (13.89) 12 Mean of 3 20.07 (7.02) 20.36 (7.74) 0.98 (0.92–0.99) 1.12 (5.52) 3.1 (15.31) 12 Max of 3 21.27 (7.16) 22.28 (7.77) 0.98 (0.92–0.99) 1.12 (5.16) 3.12 (14.31) R 11 Trial 1 19.74 (5.27) 21.38 (6.94) 0.93 (0.73–0.98) 1.58 (7.71) 4.39 (21.37) 12 Mean of 3 20.88 (5.95) 21.83 (7.48) 0.97 (0.9–0.99) 1.15 (5.41) 3.2 (14.99) 12 Max of 3 22.33 (6.25) 23.23 (7.24) 0.98 (0.93–1.0) 0.84 (3.67) 2.32 (10.18) IND_VID L 12 Trial 1 20.93 (7.82) 22.58 (7.27) 0.94 (0.78–0.98) 1.87 (8.6) 5.18 (23.83) 12 Mean of 3 20.72 (7.86) 21.68 (6.46) 0.94 (0.81–0.98) 1.7 (8.04) 4.72 (22.28) 12 Max of 3 22.44 (8.67) 23.42 (7.33) 0.95 (0.85–0.99) 1.72 (7.5) 4.77 (20.79) R 12 Trial 1 23.57 (9.53) 24.26 (8.78) 0.97 (0.91–0.99) 1.47 (6.14) 4.07 (17.02) 12 Mean of 3 23.51 (9.95) 24.64 (9.56) 0.98 (0.94–0.99) 1.32 (5.47) 3.65 (15.17) 12 Max of 3 25.53 (10.19) 25.66 (9.74) 0.98 (0.94–1.0) 1.28 (5.01) 3.55 (13.88) F2F_IND L 12 Trial 1 20.73 (9.91) 21.5 (9.26) 0.97 (0.9–0.99) 1.62 (7.67) 4.49 (21.26) 12 Mean of 3 20.8 (9.78) 21.39 (8.85) 0.99 (0.96–1.0) 1.01 (4.8) 2.81 (13.31) 12 Max of 3 22.54 (10.08) 22.66 (9.14) 0.98 (0.94–1.0) 1.25 (5.51) 3.45 (15.28) R 12 Trial 1* 21.95 (11.88) 24.0 (10.94) 0.98 (0.88–0.99) 1.74 (7.56) 4.82 (20.97) 12 Mean of 3 22.64 (11.12) 23.86 (10.12) 0.99 (0.95–1.0) 1.19 (5.14) 3.31 (14.23) 12 Max of 3 24.53 (11.23) 25.43 (10.56) 0.99 (0.96–1.0) 1.14 (4.57) 3.17 (12.68) VID_IND L 12 Trial 1 19.72 (6.2) 19.11 (6.17) 0.94 (0.81–0.98) 1.46 (7.54) 4.06 (20.9) 12 Mean of 3 19.59 (6.37) 18.77 (5.9) 0.95 (0.84–0.99) 1.33 (6.96) 3.7 (19.3) 12 Max of 3 20.71 (6.38) 20.42 (6.16) 0.96 (0.86–0.99) 1.26 (6.15) 3.5 (17.04) R 12 Trial 1 19.69 (7.52) 19.92 (6.62) 0.96 (0.87–0.99) 1.39 (7.01) 3.85 (19.44) 12 Mean of 3 19.94 (6.87) 19.79 (6.41) 0.98 (0.92–0.99) 1.03 (5.2) 2.86 (14.42) 12 Max of 3 21.1 (7.25) 20.79 (6.33) 0.97 (0.91–0.99) 1.08 (5.17) 3.0 (14.33) F2F_VID L 12 Trial 1 23.22 (8.25) 23.07 (9.8) 0.98 (0.94–1.0) 1.15 (4.98) 3.2 (13.82) 12 Mean of 3 21.98 (7.77) 22.26 (8.97) 0.99 (0.96–1.0) 0.96 (4.35) 2.66 (12.05) 12 Max of 3 23.93 (8.42) 23.92 (9.71) 0.98 (0.93–0.99) 1.27 (5.32) 3.53 (14.74) R 12 Trial 1 23.91 (8.32) 24.43 (10.56) 0.98 (0.93–0.99) 1.37 (5.66) 3.8 (15.7) 12 Mean of 3 23.26 (8.91) 23.29 (9.02) 0.99 (0.96–1.0) 0.99 (4.25) 2.74 (11.79) 12 Max of 3 24.85 (9.06) 25.35 (9.7) 0.99 (0.95–1.0) 1.11 (4.42) 3.08 (12.26) VID_VID L 12 Trial 1 20.99 (7.51) 20.42 (7.35) 0.98 (0.94–0.99) 0.98 (4.74) 2.72 (13.14) 12 Mean of 3 20.28 (6.86) 19.94 (7.57) 0.99 (0.98–1.0) 0.62 (3.06) 1.71 (8.48) 12 Max of 3 21.99 (7.61) 20.93 (7.92) 0.98 (0.93–1.0) 1.0 (4.67) 2.78 (12.94) R 12 Trial 1 21.97 (9.39) 22.0 (9.08) 0.98 (0.93–0.99) 1.27 (5.76) 3.51 (15.96) 12 Mean of 3 21.9 (8.62) 21.58 (8.78) 0.99 (0.98–1.0) 0.68 (3.15) 1.9 (8.73) 12 Max of 3 23.18 (8.86) 22.89 (8.99) 0.99 (0.97–1.0) 0.85 (3.71) 2.37 (10.28) All groups demonstrated excellent test–retest reliability (ICC ≥ 0.93, ICC lower bounds > 0.73). The groups that received full procedural supervision during both onboarding and retesting – VID_VID and F2F_VID – demonstrated the most reliable and precise test-retest performance. These groups showed the lowest MDC% values (e.g., VID_VID: 11.52% [L], 11.66% [R]; F2F_VID: 13.54% [L], 13.25% [R]) and ICC lower bounds consistently in the “excellent” range (VID_VID: 0.95–0.96; F2F_VID: 0.94–0.95). In contrast, partially supervised groups – F2F_IND, VID_IND, and IND_VID – showed higher MDC% values, although all ICC lower bounds remained within the “good” to “excellent” range. For example, F2F_IND reached 16.62% (L) and 15.96% (R), with ICCs around 0.93; VID_IND and IND_VID showed slightly more variation in MDC% (e.g., IND_VID (L): 22.30%) and ICCs ranging from 0.81–0.93. The fully unsupervised group (IND_IND) did not exhibit the worst average performance in any single metric but showed the widest variability, particularly for the right hand. While MDC% values were moderate overall (14.50% [L], 15.51% [R]), their standard deviations differed notably – 0.73 for the left, but 5.61 for the right. Similarly, ICC lower bounds averaged within the “good” range (0.92 [L], 0.85 [R]), but again, the right hand showed substantial spread (SD = 0.11), driven by a single case (Trial 1, R) dipping just below the 0.75 threshold, into the “moderate” range of 0.73. Overall, these findings suggest that increased supervision leads to more precise and reliable HGS measurements, particularly under Mean of 3 and Max of 3 reporting, while variability remains most pronounced in single-trial assessments with limited procedural oversight. These trends were mirrored in the Bland–Altman analysis (Fig. 3 ). Mean differences across sessions were small in all groups (ranging from − 0.68 to 1.41 kg), but the smallest LoA were seen in fully supervised conditions – particularly VID_VID (e.g., -2.84 to 2.18 kg for the average of three trials). Wider LoAs were found in partially supervised and unsupervised groups such as IND_VID and F2F_IND, with IND_VID trial 1 scores showing the broadest range (–5.00 to 7.35 kg). To quantify the influence of procedural supervision on measurement consistency, we examined associations between supervision score and four key test-retest metrics – ICC lower bounds, absolute mean difference, MDC% and LoA width – for each HGS reporting approach (Fig. 4 ). Overall, higher levels of supervision were associated with better measurement precision and reliability, particularly for the Trial 1 and Mean of 3 reporting approaches. For Trial 1, the supervision score was positively associated with ICC lower bound (ρ = 0.71, p = 0.010) and was strongly associated with lower absolute mean difference (ρ = -0.74, p = 0.006), suggesting more consistent scores with increased supervision. The correlation with MDC% was moderate but marginally significant (ρ = − 0.55, p = 0.062), while LoA width showed no significant relationship (ρ = − 0.26, p = 0.411). The Mean of 3 method demonstrated the clearest pattern across all metrics, with supervision significantly associated with increased ICC lower bound (ρ = 0.78, p = 0.003), lower MDC% (ρ = − 0.82, p = 0.001), and reduced LoA width (ρ = − 0.78, p = 0.003). Although the correlation with mean difference was weaker (ρ = − 0.49, p = 0.102), the overall trend indicates enhanced test-retest reliability with procedural oversight. By contrast, the Max of 3 method exhibited weak and non-significant associations across all metrics, including ICC (ρ = 0.42, p = 0.172). This suggests that the Max approach may be inherently more stable across assessment conditions, or conversely, less sensitive to changes in protocol compliance. Protocol Compliance Results Of the 72 participants, 41 had video data for both Session 1 and Session 2. An additional 6 participants had video data for Session 1 only, and 3 had data for Session 2 only. The remaining 22 participants had missing recordings due to lack of consent. In total, 50 participants (69.4%) contributed at least one session to the compliance-based analysis. Normality of compliance scores was assessed using the Shapiro–Wilk test, which indicated significant deviation from normality in both sessions (Session 1: W = 0.877, p < 0.001; Session 2: W = 0.866, p < 0.001). Given these violations of the normality assumption, the Wilcoxon signed-rank test was used to compare scores across sessions. Among the 41 participants with video data for both sessions, compliance scores were significantly lower in Session 2 compared to Session 1 ( W = 87.5, p = 0.013). Mean scores declined from 22.12 (SD = 1.83) to 21.41 (SD = 2.17). At the individual level, 21 participants (51%) showed a decrease in compliance, 14 (34%) showed no change, and 6 (15%) showed an increase. As shown in Fig. 5 , compliance scores increased with greater levels of supervision during remote assessment, across both individual sessions and the mean of both sessions. In Session 1, the Spearman correlation between supervision score and compliance score was moderate with ρ = 0.51 ( p = 0.000), while Session 2 was slightly lower at ρ = 0.36 ( p = 0.017). When compliance scores were averaged across both sessions, the correlation remained moderate (ρ = 0.49, p = 0.000), indicating a consistent relationship between procedural supervision and compliance. There was, however, considerable variation in compliance scores within each supervision level, as illustrated by the spread of individual data points. Further analyses explored whether protocol compliance influenced test-retest performance. Participants were grouped into Low (N = 7), Medium (N = 21), and High (N = 22) compliance levels based on their mean compliance scores across sessions (Fig. 6 ) . Higher compliance was associated with improved reliability for the Mean of 3 and Max of 3 scoring methods. Specifically, significant positive correlations were observed between compliance level and ICC lower bounds for both Mean of 3 (ρ = 0.96, p = 0.003) and Max of 3 (ρ = 0.84, p = 0.038). For Max of 3, compliance was also significantly associated with reduced absolute mean difference in HGS between sessions (ρ = − 0.84, p = 0.038), suggesting better within-subject agreement. No significant associations were observed for the Trial 1 metric across any outcome. Participant Questionnaire Results All participants (N = 72) provided responses. Overall, participants found the remote setup and assessment highly accessible and user-friendly. Nearly 90% rated the setup and HGS assessment instructions as “clear” or “very clear,” and over 97% reported the device was “easy” or “very easy” to use. Most (76%) felt comfortable or very comfortable performing the assessment without supervision, and 58% believed the app accurately captured their strength. While 73.6% reported no difficulties, about a quarter experienced issues – 16.7% reported posture challenges alone, 4.2% reported posture plus difficulty understanding instructions, 1.4% reported posture plus physical discomfort, 1.4% experienced all three, and 2.8% reported physical discomfort only. No participants reported technical difficulties. Finally, satisfaction was uniformly high: over 95% of respondents were “satisfied” or “very satisfied” with their HGS assessment experience. Discussion This study demonstrates that remote HGS assessment yields excellent or good-to-excellent test-retest reliability, with ICCs consistently above 0.93 (95% CI 0.73-1.0) across all reporting approaches, hands, and supervision groups. These findings support the feasibility and reliability of remote assessment of HGS using the GripAble sensor and Able Assess platform. While ICC reflects between-subject variance and is useful for assessing consistency across a sample, the MDC provides complementary insight into within-subject agreement – or measurement precision – which is especially relevant in heterogeneous populations (Atkinson and Nevill 1998 ). In our study, MDC% varied across groups, suggesting that measurement precision may be influenced by the degree of procedural supervision and structured guidance. A high MDC% in HGS between sessions indicates that small changes may go undetected, making it harder to distinguish real improvements or declines. While no universal threshold exists, MDC% in healthy adults typically ranges from 10–20% of measured grip force (Savas et al. 2023), with values below 19.5% sometimes cited as indicative of good precision in clinical settings (Bohannon 2019 ; Kim, Park, and Shin 2014). In our study, only the fully supervised groups (VID_VID and F2F_VID) consistently met this benchmark across all reporting approaches, with the lowest MDC% observed in VID_VID (8.48%, left-hand mean). In contrast, groups with partial or no supervision demonstrated greater variability, especially with the Trial 1 reporting approach, where MDC% often exceeded 20% (e.g., VID_IND left: 23.8%, F2F_IND right: 21.3%), while the mean of three trials or max scores frequently hovered near the threshold. These findings show that some level of supervision, either in-person or remote, can help reduce variability and improve measurement consistency especially if this supervision is established early, during onboarding. However, previous in-person test-retest studies using GripAble have, despite high ICCs, also reported MDC% values exceeding 20% but only during single trial usage (Kudin, Ergen, and McGee 2025). Therefore, utilising a three trial reporting approach (Mean of 3 or Max of 3) will improve precision regardless of supervision level. The persistence of elevated MDC% even in in-person settings may also indicate that factors beyond supervision alone, such as the device’s susceptibility to procedural inconsistencies between sessions (e.g. subtle changes in hand placement or grip execution) may contribute to reduced measurement precision. In contrast, interrater studies conducted within a single session have shown MDC% remaining below 15% (Ergen, Kudin, and McGee 2024) – possibly because participants are more likely to maintain consistent positioning and technique across closely timed trials. Additional factors such as, but not limited to, measurement time of day, fatigue levels, and previous night’s sleep will also affect measurement accuracy and were not controlled in this study to mimic real-world use. Correlation analyses confirmed the decline in measurement precision with reduced supervision, showing strong associations between supervision score and MDC%, LoA width, and mean difference for selected reporting approaches. These results support the observation that greater supervision enhances both consistency and agreement. Additionally, some of the partially supervised groups (IND_VID and VID_IND) showed reduced reliability and precision of the left hand when compared to the right hand (see Fig. 2 ). Given the majority of participants were right-handed, this indicates that hand dominance is also a factor with the non-dominant hand being more affected by the level of supervision. Both maximum and average scores generally yielded lower MDC% values, but neither consistently outperformed the other. Maximum scores showed weaker associations with supervision, suggesting greater resistance to protocol deviations – a potential advantage in unsupervised settings. In contrast, average scores more often reflected supervision effects. Therefore, using maximum values may offer greater stability across varied conditions. Lower compliance – particularly in fully unsupervised groups – was likely driven by misinterpretation of instructions or inconsistent assessment execution. These behavioural observations were consistent with reduced test–retest reliability, reflected in lower ICC values in the same groups. However, the relationship between supervision level and compliance scores was inconsistent, with notable variation observed within supervision groups. This variability may partly reflect individual differences in engagement or motivation, which was not formally assessed in this study (e.g., using the Intrinsic Motivation Inventory). Compliance assessment relied solely on observable behaviours captured on video, which may overlook subtler deviations. This was supported by participant feedback: although most rated the instructions and device usability as clear and easy, about a quarter still reported posture-related difficulties during the test. Future implementations could benefit from sensor-based tools that enable detection of non-observable non-compliance and support real-time guidance to improve protocol compliance in remote settings. Beyond measurement metrics, user feedback highlighted the feasibility of remote HGS assessment: the vast majority of participants found the setup intuitive, the assessment easy to complete, and the overall experience highly satisfactory. Most felt comfortable performing the assessment without supervision, and few reported environmental barriers, reinforcing the practicality of remote HGS assessment in home settings. This study has several limitations. Firstly, the absence of a gold-standard in-person baseline makes it difficult to determine whether consistency in groups like IND_IND reflects true performance or systematic error. We intentionally omitted an in-person session to preserve participant naivety; however, including it as a third session would have enabled protocol benchmarking. Secondly, compliance was rated by a single observer and focused on overt postural deviations, potentially overlooking more subtle errors. Incorporating multiple raters and sensor-derived metrics in future studies could improve scoring reliability. Lastly, while this study used the GripAble sensor and Able Assess platform as the assessment framework, the generalisability of these findings to other remote assessment tools will depend on the extent to which those platforms support similar procedural structure, user guidance, and compliance monitoring. Conclusion Despite the growing evidence of HGS as a new vital sign of general health, it is yet to be embedded into routine practice and used across the care continuum in both supervised and unsupervised use-cases. This study provides evidence that remote HGS assessment using digital tools such as the GripAble sensor and Able Assess platform is feasible and yields high test-retest reliability under a range of real world conditions. Nonetheless, measurement precision was influenced by level of supervision and when introduced, protocol compliance, and trial aggregation, to name a few, reflecting the multifactorial nature of reliability in decentralised assessment environments. While fully unsupervised assessment may be viable in select populations – particularly those with prior training or strong intrinsic compliance – small reductions in individual-level precision suggest real world HGS assessment may benefit from additional procedural or technological safeguards. A key recommendation based on these findings is that a supervised baseline measurement, prior to any remote testing, should improve measurement precision during repeated testing of an individual. Therefore, future implementations should utilise supervised onboarding alongside adaptive assessment protocols, embedded instructional support, real-time feedback, and post-hoc validation using, for example, sensor-derived metrics to reinforce protocol compliance. Importantly, such enhancements should be designed to support – not burden – usability, especially for end users in remote or unsupervised settings. As remote assessment tools are increasingly adopted in clinical and research contexts, their impact will depend not only on technical accuracy, but also on their ability to deliver reliable, actionable data within systems that prioritise both data quality and patient-centred care. Declarations Author contributions SAM and MM conceived the study and designed the protocol. SAM conducted the data analysis and drafted the manuscript. DJ collected the data. SP assessed protocol compliance from video recordings. EB supervised the study as principal investigator. All authors reviewed and approved the final manuscript. Acknowledgements We thank all participants for their time and involvement in the study. Ethics approval and consent to participate Ethical approval was gained from the Science Engineering Technology Research Ethics Committee of the Imperial College of Science, Technology and Medicine, London, UK (REF: 20IC5831). All participants gave their written consent to take part in the experi­ment. The experiment followed the ethical standards in the 1964 Declaration of Helsinki. Consent for publication All participants received a thorough explanation of the risks and benefits of inclusion and gave their oral and written informed con­sent to publish the data. Declaration of competing interest SAM, SP, and MM are employees of GRIPABLE Ltd. DJ worked on the study during an internship at GRIPABLE Ltd. MM and EB are the authors on the GripAble sensor patent. References Atkinson, G., and A. M. Nevill. 1998. “Statistical Methods for Assessing Measurement Error (reliability) in Variables Relevant to Sports Medicine.” Sports Medicine (Auckland, N.Z.) 26 (4): 217–38. Bohannon, Richard W. 2019. “Minimal Clinically Important Difference for Grip Strength: A Systematic Review.” Journal of Physical Therapy Science 31 (1): 75–78. Celis-Morales, Carlos A., Paul Welsh, Donald M. Lyall, Lewis Steell, Fanny Petermann, Jana Anderson, Stamatina Iliodromiti, et al. 2018. “Associations of Grip Strength with Cardiovascular, Respiratory, and Cancer Outcomes and All Cause Mortality: Prospective Cohort Study of Half a Million UK Biobank Participants.” BMJ (Clinical Research Ed.) 361 (May):k1651. Chang, Yu-Tzu, Hung-Lien Wu, How-Ran Guo, Ya-Yun Cheng, Chin-Chung Tseng, Ming-Cheng Wang, Ching-Yuang Lin, and Junne-Ming Sung. 2011. “Handgrip Strength Is an Independent Predictor of Renal Outcomes in Patients with Chronic Kidney Diseases.” Nephrology, Dialysis, Transplantation : Official Publication of the European Dialysis and Transplant Association - European Renal Association 26 (11): 3588–95. Ergen, Halil Ibrahim, Roman Kudin, and Corey W. McGee. 2024. “Interrater Reliability and Precision of a Novel Hand Strength Assessment and Treatment Device: The GripAble.” The American Journal of Occupational Therapy : Official Publication of the American Occupational Therapy Association 78 (5). https://doi.org/10.5014/ajot.2024.050689. Hailu, Ruth, Jessica Sousa, Mitchell Tang, Ateev Mehrotra, and Lori Uscher-Pines. 2024. “Challenges and Facilitators in Implementing Remote Patient Monitoring Programs in Primary Care.” Journal of General Internal Medicine 39 (13): 2471–77. Heslop, Philip A., Christopher Hurst, Avan A. Sayer, and Miles D. Witham. 2023. “Remote Collection of Physical Performance Measures for Older People: A Systematic Review.” Age and Ageing 52 (1). https://doi.org/10.1093/ageing/afac327. Hoenemeyer, Teri W., William W. Cole, Robert A. Oster, Dorothy W. Pekmezi, Andrea Pye, and Wendy Demark-Wahnefried. 2022. “Test/Retest Reliability and Validity of Remote vs. In-Person Anthropometric and Physical Performance Assessments in Cancer Survivors and Supportive Partners.” Cancers 14 (4): 1075. Kim, Jae Kwang, Min Gyue Park, and Sung Joon Shin. 2014. “What Is the Minimum Clinically Important Difference in Grip Strength?” Clinical Orthopaedics and Related Research 472 (8): 2536–41. Klein, Thorsten, Annette Worth, Claudia Niessner, and Anke Hanssen-Doose. 2025. “Remote Assessment of Physical Fitness via Videoconferencing: A Systematic Review.” BMC Sports Science, Medicine & Rehabilitation 17 (1): 11. Koo, Terry K., and Mae Y. Li. 2016. “A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research.” Journal of Chiropractic Medicine 15 (2): 155–63. Kudin, Roman, Halil Ibrahim Ergen, and Corey W. McGee. 2025. “Test-Retest Reliability and Precision of GripAble: A Multipurpose Exergaming Device.” Games for Health Journal , February. https://doi.org/10.1089/g4h.2024.0216. Leong, Darryl P., Koon K. Teo, Sumathy Rangarajan, Patricio Lopez-Jaramillo, Alvaro Avezum Jr, Andres Orlandini, Pamela Seron, et al. 2015. “Prognostic Value of Grip Strength: Findings from the Prospective Urban Rural Epidemiology (PURE) Study.” Lancet (London, England) 386 (9990): 266–73. Li, Joule J., Gary A. Wittert, Andrew Vincent, Evan Atlantis, Zumin Shi, Sarah L. Appleton, Catherine L. Hill, Alicia J. Jenkins, Andrzej S. Januszewski, and Robert J. Adams. 2016. “Muscle Grip Strength Predicts Incident Type 2 Diabetes: Population-Based Cohort Study.” Metabolism: Clinical and Experimental 65 (6): 883–92. Loring, David W., James J. Lah, and Felicia C. Goldstein. 2023. “Telehealth Equivalence of the Montreal Cognitive Assessment (MoCA): Results from the Emory Healthy Brain Study (EHBS).” Journal of the American Geriatrics Society 71 (6): 1931–36. Mace, Michael, Sharah Abdul Mutalib, Matjaz Ogrinc, Nicola Goldsmith, and Etienne Burdet. 2022. “GripAble: An Accurate, Sensitive and Robust Digital Device for Measuring Grip Strength.” Journal of Rehabilitation and Assistive Technologies Engineering 9 (March):20556683221078455. Mutalib, Sharah Abdul, Michael Mace, Chloe Seager, Etienne Burdet, Virgil Mathiowetz, and Nicola Goldsmith. 2022. “Modernising Grip Dynamometry: Inter-Instrument Reliability between GripAble and Jamar.” BMC Musculoskeletal Disorders 23 (1): 80. Mutalib, Sharah Abdul, Deepika Sharma, Sonia Pike, Liz Gwynne, Samantha Hyde, Jennifer Morehouse, Helen Davey, et al. 2024. “GripAble: Interrater Reliability and Normative Grip Strength of UK Population.” Journal of Hand Therapy : Official Journal of the American Society of Hand Therapists , March. https://doi.org/10.1016/j.jht.2023.12.013. Naef, Aileen C., Guichande Duarte, Saskia Neumann, Migjen Shala, Meret Branscheidt, and Chris Easthope Awai. 2025. “Toward Unsupervised Capacity Assessments for Gait in Neurorehabilitation: Validation Study.” Journal of Medical Internet Research 27 (March):e66123. Savas, Sumru, Asli Kilavuz, Fatma Özge Kayhan Koçak, and Sibel Cavdar. 2023. “Comparison of Grip Strength Measurements by Widely Used Three Dynamometers in Outpatients Aged 60 Years and Over.” Journal of Clinical Medicine 12 (13). https://doi.org/10.3390/jcm12134260. Serrano, Luiza Palmieri, Karla C. Maita, Francisco R. Avila, Ricardo A. Torres-Guzman, John P. Garcia, Abdullah S. Eldaly, Clifton R. Haider, et al. 2023. “Benefits and Challenges of Remote Patient Monitoring as Perceived by Health Care Practitioners: A Systematic Review.” The Permanente Journal 27 (4): 100–111. Tomkinson, Grant R., Justin J. Lang, Lukáš Rubín, Ryan McGrath, Bethany Gower, Terry Boyle, Marilyn G. Klug, et al. 2024. “International Norms for Adult Handgrip Strength: A Systematic Review of Data on 2.4 Million Adults Aged 20 to 100+ Years from 69 Countries and Regions.” Journal of Sport and Health Science 14 (December):101014. Vaishya, Raju, Anoop Misra, Abhishek Vaish, Nicola Ursino, and Riccardo D’Ambrosi. 2024. “Hand Grip Strength as a Proposed New Vital Sign of Health: A Narrative Review of Evidences.” Journal of Health, Population, and Nutrition 43 (1): 7. Additional Declarations Competing interest reported. SAM, SP, and MM are employees of GRIPABLE Ltd. DJ worked on the study during an internship at GRIPABLE Ltd. MM and EB are the authors on the GripAble sensor patent. Supplementary Files Appendix.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6801648","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":465497531,"identity":"5ff32eb2-4629-4b79-888d-31a336804f98","order_by":0,"name":"Sharah Abdul Mutalib","email":"","orcid":"","institution":"GripAble Ltd","correspondingAuthor":false,"prefix":"","firstName":"Sharah","middleName":"Abdul","lastName":"Mutalib","suffix":""},{"id":465497534,"identity":"04dab024-90b9-4eb3-8b5d-566dfb43d5ee","order_by":1,"name":"Daniel Jenkinson","email":"","orcid":"","institution":"GripAble Ltd","correspondingAuthor":false,"prefix":"","firstName":"Daniel","middleName":"","lastName":"Jenkinson","suffix":""},{"id":465497538,"identity":"3574f032-d21a-4c5c-ad72-8df26afa89a6","order_by":2,"name":"Sonia Pike","email":"","orcid":"","institution":"GripAble Ltd","correspondingAuthor":false,"prefix":"","firstName":"Sonia","middleName":"","lastName":"Pike","suffix":""},{"id":465497540,"identity":"4062f7c9-b9d6-4ed5-86b7-75b3fedd9664","order_by":3,"name":"Etienne Burdet","email":"","orcid":"","institution":"Imperial College London","correspondingAuthor":false,"prefix":"","firstName":"Etienne","middleName":"","lastName":"Burdet","suffix":""},{"id":465497545,"identity":"10ab5c4e-42e4-4570-ae42-a2c176bf10c8","order_by":4,"name":"Corey McGee","email":"","orcid":"","institution":"University of Minnesota","correspondingAuthor":false,"prefix":"","firstName":"Corey","middleName":"","lastName":"McGee","suffix":""},{"id":465497547,"identity":"4890d1e4-518c-4fd7-80b4-c3d7152d9be2","order_by":5,"name":"Mike Mace","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAwklEQVRIiWNgGAWjYDACZjYGhg8MDDxAxHiAaC2MMyBaGIjUwsDGwMwDZhCrRbedLfGxza97MubsZw8c+MBgJ0dQi9lhtsPGuX3FPJY9eQkHZzAkGxOhhb1NOrcngcfgQI7BYaDbEhuI0NL+2xKk5fwbg8N/iNPCdoyZ4QdQyw2gLQxEakmW7G1I4LGc8S7hYI8BMX45f8zww48/Cfbm/LkHH/yoICLEwICxjYHBAMwyIE4DEPwhRfEoGAWjYBSMOAAA8Do9YNic3O4AAAAASUVORK5CYII=","orcid":"","institution":"GripAble Ltd","correspondingAuthor":true,"prefix":"","firstName":"Mike","middleName":"","lastName":"Mace","suffix":""}],"badges":[],"createdAt":"2025-06-02 11:23:14","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6801648/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6801648/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":83913408,"identity":"4122382e-f85d-491a-bf6d-6405018f1fb0","added_by":"auto","created_at":"2025-06-04 12:21:22","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":608317,"visible":true,"origin":"","legend":"\u003cp\u003eOverview of the Able Assess platform, assessment protocol, and study design. (A) The Able Assess platform (Able Assess app and GripAble sensor) interface were used to deliver standardised remote HGS assessment. (B) Participants were guided through a step-by-step on-screen instruction during assessment, with visual and text-based instructions. (C) Study protocol showing posture guidelines, assessment conditions, timeline, and group allocation.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6801648/v1/3330a530878352d7731b446f.png"},{"id":83913412,"identity":"fda0fbc9-567b-4acc-a4bf-ae29d8bd83fd","added_by":"auto","created_at":"2025-06-04 12:21:23","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":425546,"visible":true,"origin":"","legend":"\u003cp\u003eTest–retest reliability and measurement precision by participant group (top) and reporting approaches (bottom). Each row presents distributions of ICC (lower 95% CI), MDC (kg), and MDC% across either participant groups (top) or HGS reporting approaches (bottom). Box plots illustrate each subgroup’s spread and central tendency, while overlaid points (colour indicating hand, marker shape indicating reporting approach or group) shows the individual data associated with Table 2. Horizontal dashed lines mark ICC interpretation thresholds (0.5 = moderate, 0.75 = good, 0.9 = excellent). In the group-based plots, vertical dashed lines indicate supervision levels, reflecting procedural oversight during onboarding and retesting (see Methods for supervision score criteria).\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6801648/v1/e2a6265bb9ba603c6f7f663b.png"},{"id":83913413,"identity":"c4be6fed-eb14-4f0b-98d3-20b5f9dd79fc","added_by":"auto","created_at":"2025-06-04 12:21:23","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":900643,"visible":true,"origin":"","legend":"\u003cp\u003eBland–Altman plots showing the agreement between Session 1 and Session 2 HGS measurements for each group and reporting approach (Trial 1, Mean of 3, Max of 3). Each subplot represents one group (rows) and one reporting approach (columns), with left and right hands indicated by marker shape and gender by colour. Solid lines indicate the mean difference (bias), and dotted lines represent the 95% limits of agreement (LoA), both expressed in absolute units (kg).\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6801648/v1/357cf678941eb5448db4bffe.png"},{"id":83913411,"identity":"a5c35a17-54df-4684-8bd6-26b365602bca","added_by":"auto","created_at":"2025-06-04 12:21:23","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":667839,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between supervision score and measurement precision across HGS reporting approaches. Each panel shows the relationship between supervision score (x-axis) and one of four reliability metrics — ICC (95% CI lower bound), MDC%, absolute mean difference, limits of agreement (LoA) width — across three HGS reporting approaches (Trial 1, Mean of 3, Max of 3; columns). Supervision scores reflect the level of procedural oversight across onboarding and retesting. Points represent individual data points from Table 2, colored by group and shaped by hand (circles = left, squares = right). \u003cstrong\u003eBlack solid lines\u003c/strong\u003e show fitted linear trends across \u003cstrong\u003eall groups\u003c/strong\u003e, with Spearman’s ρ and p-values indicating monotonic associations. \u003cstrong\u003eRed dashed lines\u003c/strong\u003e show the same trends \u003cstrong\u003eexcluding IND_IND\u003c/strong\u003e, with corresponding Spearman’s ρ and p-values.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6801648/v1/460e0184ca2a910ae457371d.png"},{"id":83914457,"identity":"f6c0f59d-e435-47e6-81f6-1898b57c0b5e","added_by":"auto","created_at":"2025-06-04 12:29:23","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":240586,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between procedural supervision and compliance scores across assessment sessions. Each panel shows participant compliance scores (y-axis) plotted against supervision score (x-axis), where supervision score reflects the level of oversight across onboarding and retest sessions. Groups are colour-coded and fitted with a linear trend line. Left and middle panels show Session 1 and Session 2 separately; the right panel presents the mean compliance score across sessions. Spearman’s rank correlation coefficients (ρ) and p-values indicate the strength and significance of association between supervision and compliance.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-6801648/v1/42ae9e2db635a1d08a00991d.png"},{"id":83913409,"identity":"3afa6f9d-7c81-4340-8913-170baf0e2ae4","added_by":"auto","created_at":"2025-06-04 12:21:23","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":415389,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between protocol compliance and test–retest reliability metrics across reporting approaches. Each panel shows the association between compliance level (x-axis: Low = 0, Medium = 1, High = 2) and one of four reliability or precision metrics – ICC lower bound, MDC%, absolute mean difference, and LoA width (rows) – across three HGS reporting approaches (Trial 1, Mean of 3, Max of 3; columns). Points represent left (circles) and right (squares) hand measurements, coloured by the compliance group. Black trend lines indicate fitted linear associations, with corresponding Spearman’s ρ and p-values displayed in each subplot.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-6801648/v1/b04525c6314f75865ab7efed.png"},{"id":86956667,"identity":"d1eafa01-a363-4549-b4c4-eef7a1fbe235","added_by":"auto","created_at":"2025-07-17 15:17:09","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4366605,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6801648/v1/7a205af1-f4bd-445c-bf37-69dcd02fed3d.pdf"},{"id":83913406,"identity":"c2da5c95-f91d-4b95-a591-bbb0eaf44c5f","added_by":"auto","created_at":"2025-06-04 12:21:22","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":20556,"visible":true,"origin":"","legend":"","description":"","filename":"Appendix.docx","url":"https://assets-eu.researchsquare.com/files/rs-6801648/v1/d2080e97b43107c3ad7ed0f9.docx"}],"financialInterests":"Competing interest reported. SAM, SP, and MM are employees of GRIPABLE Ltd. DJ worked on the study during an internship at GRIPABLE Ltd. MM and EB are the authors on the GripAble sensor patent.","formattedTitle":"Remote, Reliable, Repeatable: Real-World Test–Retest Validation of Hand Grip Strength Assessments","fulltext":[{"header":"Introduction","content":"\u003cp\u003e\u003cb\u003eHandgrip strength (HGS)\u003c/b\u003e has emerged as a robust biomarker of health status. It reflects overall muscular strength and is easily measured, serving as a convenient indicator of an individual’s functional capacity including the ability to perform activities of daily living (ADLs). HGS is particularly relevant in the context of aging, as it declines with advancing age and mirrors the process of sarcopenia – the progressive loss of muscle mass, strength and function in older adults. Low HGS also correlates with greater functional decline – including higher odds of disability and hospitalization – and significantly higher all-cause mortality rates (Vaishya et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Numerous studies report that individuals with weaker HGS face elevated risks of aging-related diseases such as type 2 diabetes, cardiovascular disease, stroke, respiratory disease, chronic kidney disease, and certain cancers (Li et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Chang et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Celis-Morales et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). A large cohort analysis has found that each ~ 5 kg decrement in HGS corresponds to a substantial increase in all-cause and cardiovascular mortality risk, and HGS can outperform some traditional risk factors (e.g. blood pressure) in predicting long-term survival (Leong et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). A recent systematic review of 2.4\u0026nbsp;million adults across 69 countries confirmed age-related decline in HGS and established global norms, reinforcing its value as a benchmark for identifying low muscle strength and guiding early intervention (Tomkinson et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDespite its clinical value, HGS remains underused in routine care. Barriers include limited access to appropriate equipment, the need for trained personnel, and uncertainty around how to interpret results in everyday workflows. HGS assessment has been shown to be reliable in clinical settings where it is performed under supervision. Expanding HGS assessment beyond the clinic and into remote settings could improve accessibility and enable scalable adoption of HGS across the care continuum. However, it also raises concerns about standardisation, patient compliance, and data quality in real-world settings.\u003c/p\u003e \u003cp\u003eHealth assessment outside clinical settings is not new, but has recently gained momentum across clinical domains such as geriatrics, neurology, cardiology, primary care, and mental health, driven in part by the recent COVID pandemic. Examples include video-based physical function tests (e.g. chair stand, gait speed), digital cognitive screening, remote vital sign monitoring, and telepsychiatric evaluations, including in cancer patients (Hoenemeyer et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and stroke patients (Naef et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Many approaches are supported by emerging evidence: remote chair stand tests have shown strong validity and test-retest reliability in healthy adults (Klein et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), though findings in older or frailer populations remain mixed (Heslop et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Similarly, cognitive assessments delivered via telehealth, such as the Montreal Cognitive Assessment (MoCA) score, are comparable to in-person assessment, even among older adults (Loring, Lah, and Goldstein \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eNonetheless, several challenges persist. Many tools still lack large-scale validation, and alignment with in-clinic results is not always consistent. Discrepancies can arise from home-specific factors such as environmental distractions, inconsistent setup, and the absence of real-time guidance. Additional influences include the cognitive demands of self-directed assessment, unclear instructions, and whether the task is volitional or autonomic. The lack of the white coat effect – enhanced performance under observation – may also reduce engagement. Digital literacy barriers, particularly among older or underserved populations can compound these challenges. Clinicians also frequently cite concerns about patients’ ability to use remote tools effectively (Klein et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Hailu et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Even when validated, factors including, but not limited to, long-term compliance, user engagement, cognitive status and technology-related frustrations can limit the integration of remote assessment into routine care (Serrano et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAgainst this backdrop, there is growing interest in evaluating whether HGS – a clinically meaningful and easily interpretable physical measure – can be effectively assessed in remote contexts using digital tools. To realise HGS’s utility as a widely adopted biomarker in decentralised care, it is critical to establish both the reliability of remote assessment and the accuracy with which participants follow protocols independently. Without such evidence, HGS will remain a validated yet underutilised metric in routine practice.\u003c/p\u003e \u003cp\u003eTo address this gap, we evaluated the \u003cb\u003etest-retest reliability\u003c/b\u003e and \u003cb\u003eprotocol compliance\u003c/b\u003e of remote HGS assessment as a model for decentralised assessment. In this study, we used the Able Assess platform – a digital assessment platform combining a handheld device (GripAble) and app interface to guide and capture standardised HGS remotely. Our aim was to examine whether different remote assessment conditions affect (1) the consistency of HGS across repeated sessions and (2) participant compliance to the protocol. By simulating real-world use scenarios, this study explores the potential for remote HGS assessment to generate clinically valid and operationally scalable data, offering a blueprint for future decentralised measurement tools.\u003c/p\u003e "},{"header":"Methods","content":"\u003cp\u003eThis parallel-group test-retest study assessed the reliability and protocol compliance of remote HGS assessments. Assessment was delivered using the Able Assess platform, which integrates the GripAble handheld sensor with an app-based interface, to standardise remote delivery of HGS assessments. To reflect real-world clinical settings, the study included varying assessment conditions: (1) fully independent self-assessment, (2) remote supervision via video, and (3) face-to-face guidance.\u003c/p\u003e\u003cp\u003eThe study followed the ethical standards of the 1964 Declaration of Helsinki and received approval from the Imperial College Research Ethics Committee and the Science, Engineering \u0026amp; Technology Research Ethics Committee (REF: 20IC5831).\u003c/p\u003e\u003cp\u003eAble Assess Platform\u003c/p\u003e\u003cp\u003eAble Assess is a digital assessment platform that integrates the GripAble sensor with an app-based interface to guide users through digitised assessments (e.g. HGS, timed-up-and-go, chair stand, etc.) and securely capture sensor data for analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). In this study, Able Assess (version 1.0.0) was used to support consistent protocol delivery, rather than to evaluate the platform itself – serving as a practical example to promote future generalisability.\u003c/p\u003e\u003cp\u003eThe GripAble sensor is a handheld digital dynamometer with dual load cells, offering ± 1.8 kg (± 2%) accuracy in isometric mode and ± 0.9 kg (± 1%) accuracy in isotonic mode across its 90 kg range, with sensitivity below 100 g (Mace et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). It has been fully validated for use in supervised clinical settings (Mutalib et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Ergen, Kudin, and McGee 2024; Kudin, Ergen, and McGee 2025). It includes an IMU (accelerometer, gyroscope, magnetometer) and samples data at 50 Hz. For this study, the grip plate was used in isometric mode to align with standardised clinical protocols and existing normative datasets (Mutalib et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The device connects via Bluetooth to the Able Assess app, which standardises assessment administration and records grip and motion data. Calibration of the GripAble handgrip was confirmed pre- and post-study using known weights, with accuracy within ± 2% across the measurement range.\u003c/p\u003e\u003cp\u003eParticipants\u003c/p\u003e\u003cp\u003eSeventy-two participants were recruited into the study through convenience sampling across university campuses, community organisations, and personal networks. Participants were stratified by gender (36 males, 36 females) and age group (18–40, 41–60, and 61 + years), with 12 participants per gender within each age band.\u003c/p\u003e\u003cp\u003eEligible participants were aged 18 years or older, had no prior use of the GripAble device or similar hand dynamometers, and reported no history of upper limb musculoskeletal or neurological conditions, including pain in the hand, wrist, or forearm. Individuals not meeting these criteria were excluded. All participants provided written informed consent before enrolment.\u003c/p\u003e\u003cp\u003eExperiment Protocol\u003c/p\u003e\n\u003ch3\u003eHand Grip Strength Test-Retest Setup\u003c/h3\u003e\n\u003cp\u003eFigure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA and \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB show the Able Assess platform and on-screen instructions to guide participants through the assessment process, respectively. The full protocol and group allocations are illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC. Participants were allocated to one of six assessment conditions using a 3\u0026times;2 factorial design, capturing two sequential sessions: Session 1 (onboarding and initial assessment) and Session 2 (retest), conducted at least one week apart (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC). The design reflected (1) three onboarding assessment conditions during Session 1: face-to-face (in-person) supervision, video call supervision, or independent unsupervised onboarding via the app; and (2) two remote assessment conditions during Session 2: independent (unsupervised) or video-supervised administration of the HGS assessment.\u003c/p\u003e \u003cp\u003e Supervised participants received live instructions by trained personnel (either face-to-face or via video call, depending on their group assignment), while independent users followed on-screen instructions. No coaching or technique correction was provided for the independent group, while those in the supervised groups received guidance as needed.\u003c/p\u003e \u003cp\u003e During assessment, participants were instructed to be seated upright with feet flat on the floor, hips and knees at 90\u0026deg;, and the assessed arm adducted with the elbow at 90\u0026deg;, forearm in neutral position, and wrist slightly extended, following the American Society of Hand Therapists (ASHT) guidelines (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC). For clarity, each group is labelled using a two-part format that reflects the assessment conditions in Session 1 and Session 2, respectively: \u003cb\u003eIND\u003c/b\u003e\u0026thinsp;=\u0026thinsp;independent /unsupervised, \u003cb\u003eVID\u003c/b\u003e\u0026thinsp;=\u0026thinsp;video-supervised, \u003cb\u003eF2F\u003c/b\u003e\u0026thinsp;=\u0026thinsp;face-to-face supervised. For example, the IND_VID group completed Session 1 independently and Session 2 supervised via video link.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eVideo-Based Protocol Compliance Assessment\u003c/h2\u003e \u003cp\u003e Participants who consented were video recorded during the HGS assessment sessions. Assessments were conducted in line with ASHT guidelines for posture and positioning. A 12-item compliance scoring framework was independently developed to reflect key procedural elements: (1) participant seated, (2) seated posture, (3) number of trials, (4) device motion, (5) device positioning, (6) shoulder flexion, (7) shoulder adduction, (8) shoulder rotation, (9) elbow flexion, (10) forearm rotation, (11) wrist flexion, and (12) wrist deviation. Each criterion was scored on a three-point ordinal scale: 0\u0026thinsp;=\u0026thinsp;not followed, 1\u0026thinsp;=\u0026thinsp;partially followed, and 2\u0026thinsp;=\u0026thinsp;fully followed, yielding a maximum total score of 24 per session. All ratings were completed by a single trained therapist to ensure consistency. Full scoring definitions are provided in Appendix A.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eParticipant Experience Questionnaire\u003c/h3\u003e\n\u003cp\u003eAt the end of Session 2, all participants (N\u0026thinsp;=\u0026thinsp;72) completed a 10-item, five-point Likert questionnaire assessing: (1) clarity of app setup instructions, (2) need for additional assistance, (3) clarity of assessment instructions, (4) ease of using the GripAble device, (5) preparedness to follow the full protocol, (6) any setup- or assessment-related difficulties, (7) comfort performing the assessment without supervision, (8) perceived accuracy of strength capture, (9) home environmental factors, and (10) overall satisfaction. Appendix B shows the questionnaire items and available responses.\u003c/p\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eData Analysis\u003c/h2\u003e \u003cdiv id=\"Sec6\" class=\"Section3\"\u003e \u003ch2\u003eHand Grip Strength Data\u003c/h2\u003e \u003cp\u003eParticipant demographics, including age, gender, self-reported handedness, height, weight, BMI, hand dimensions, and weekly exercise hours, were summarised using descriptive statistics stratified by gender and overall.\u003c/p\u003e \u003cp\u003eTo evaluate HGS performance, three commonly used reporting approaches were compared: the first trial (Trial 1), the average of three trials (Mean of 3), and the maximum of three trials (Max of 3). Normality of HGS distributions was assessed using the Shapiro\u0026ndash;Wilk test for each session, hand, and reporting approach. HGS differences between assessment sessions were assessed using either a paired t-test or the non-parametric Wilcoxon signed-rank test, depending on the outcome of the normality test. Analyses were stratified by group, hand, and reporting approach. All statistical tests were two-tailed, with significance defined as \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05. Implausible values were excluded from analysis to avoid skewing results. All analysis was performed in Python 3.\u003c/p\u003e \u003cp\u003eTo determine the consistency and precision of repeated measurements, test\u0026ndash;retest reliability was assessed using the intraclass correlation coefficient (ICC) (Atkinson and Nevill \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1998\u003c/span\u003e; Koo and Li 2016). ICC(2,1), a two-way random-effects model for absolute agreement, was selected in line with (Koo and Li 2016), with interpretation based on the lower bound of the 95% confidence interval to guard against overestimation. Measurement error was quantified via the standard error of measurement (SEM) and minimal detectable change (MDC), where SEM was calculated as: SEM\u0026thinsp;=\u0026thinsp;SD_pooled \u0026times; \u0026radic;(1 \u0026ndash; ICC), with SD_pooled being the square root of the average of the squared standard deviations from both sessions. MDC was computed as: MDC\u0026thinsp;=\u0026thinsp;1.96 \u0026times; \u0026radic;2 \u0026times; SEM. SEM and MDC were computed using the point estimate of ICC to reflect true-score variability and detect the smallest meaningful change, respectively. Both metrics were also expressed as percentages of the group mean to aid comparability (SEM%, MDC%).\u003c/p\u003e \u003cp\u003eTo visualise individual-level agreement and detect systematic bias, Bland\u0026ndash;Altman plots were generated for each group, hand, and reporting approach, alongside mean difference and 95% limits of agreement (LoA) calculations.\u003c/p\u003e \u003cp\u003eFinally, to explore whether supervision influenced measurement fidelity, each group was assigned a supervision score (0\u0026ndash;3), reflecting procedural oversight across onboarding and retesting phases: 0\u0026thinsp;=\u0026thinsp;none (IND_IND), 1\u0026thinsp;=\u0026thinsp;partial supervision during retesting (IND_VID), 2\u0026thinsp;=\u0026thinsp;partial supervision during onboarding (VID_IND, F2F_IND), and 3\u0026thinsp;=\u0026thinsp;fully supervised (VID_VID, F2F_VID). Relationships between supervision level and key precision metrics (MDC%, absolute mean difference, and LoA width) were examined using Spearman\u0026rsquo;s rank correlation (ρ), to test the hypothesis that greater supervision would reduce variability and improve assessment consistency.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e\n\u003ch3\u003eVideo-Based Protocol Compliance Data\u003c/h3\u003e\n\u003cp\u003e To evaluate protocol compliance, scores were derived from video-based ratings across both sessions. For each participant, the mean compliance score was calculated when both sessions were available; otherwise, the single available session was used. Participants without any video data were excluded.\u003c/p\u003e \u003cp\u003eWithin-subject changes in compliance between Session 1 and Session 2 were evaluated using either a paired t-test or the non-parametric Wilcoxon signed-rank test, depending on the results of the Shapiro\u0026ndash;Wilk test for normality. For this test, only participants with scores from both sessions were included\u003c/p\u003e \u003cp\u003e To investigate whether higher compliance aligned with more structured supervision, Spearman\u0026rsquo;s rank correlation was used to assess the relationship between compliance scores and supervision scores. Additionally, participants were re-stratified into data-driven compliance groups \u0026ndash; 0\u0026thinsp;=\u0026thinsp;low (\u0026lt;\u0026thinsp;20), 1\u0026thinsp;=\u0026thinsp;moderate (20\u0026ndash;22), and 3\u0026thinsp;=\u0026thinsp;high (\u0026gt;\u0026thinsp;22) \u0026ndash; rather than relying on predefined supervision-based categories (e.g., IND_IND, F2F_VID). These compliance-based groups were then used to re-evaluate test\u0026ndash;retest reliability and measurement precision (ICC, mean difference, LoA, SEM, MDC), allowing us to examine whether actual protocol compliance, rather than nominal supervision level, better explained measurement quality.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eParticipant Experience Questionnaire\u003c/h2\u003e \u003cp\u003eResponses were summarised by question type. Likert-style items were reported as the percentage of participants selecting each response. Multiple-selection questions were summarised by the percentage of participants selecting each unique combination of responses. Open-ended responses were categorised thematically, and category frequencies were reported.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eTest-Retest HGS Results\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003epresents the demographic characteristics of participants, stratified by gender.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOverall (N\u0026thinsp;=\u0026thinsp;72)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMale (N\u0026thinsp;=\u0026thinsp;36)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eFemale (N\u0026thinsp;=\u0026thinsp;36)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAge (years)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e49.6 [18\u0026ndash;82]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e48.9 [18\u0026ndash;82]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e50.4 [18\u0026ndash;82]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHandedness (self-reported)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eR: 68, L: 4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eR: 35, L: 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eR: 33, L: 3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHeight (cm)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e172.6 [155\u0026ndash;190]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e179.3 [170\u0026ndash;190]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e165.9 [155\u0026ndash;175]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eWeight (kg)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e72.3 [53\u0026ndash;100]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e78.7 [65\u0026ndash;100]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e65.8 [53\u0026ndash;98]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eBMI (kg/m\u0026sup2;)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e24.2 [19\u0026ndash;34]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e24.5 [19\u0026ndash;30]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e23.9 [19\u0026ndash;34]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHand Length (cm)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e18.7 [17\u0026ndash;22]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e19.6 [18\u0026ndash;22]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e17.8 [17\u0026ndash;19]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHand Circumference (cm)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e20.6 [18\u0026ndash;24]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e21.7 [19\u0026ndash;24]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e19.5 [18\u0026ndash;21]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eExercise/week (hrs)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.8 [0\u0026ndash;4]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.9 [0\u0026ndash;4]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.7 [0\u0026ndash;4]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. \u003cem\u003eParticipant demographic characteristics stratified by gender. Values are presented as mean [range] for continuous variables and counts for categorical variables. BMI\u0026thinsp;=\u0026thinsp;body mass index; R\u0026thinsp;=\u0026thinsp;right; L\u0026thinsp;=\u0026thinsp;left.\u003c/em\u003e\u003c/p\u003e \u003cp\u003eThe Shapiro\u0026ndash;Wilk test showed that HGS values were normally distributed in 9 out of 12 combinations of session, hand, and reporting approach (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05). However, three combinations \u0026ndash; Session 1 right hand using Trial 1 and Mean of 3 (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.023 and \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.026, respectively), and left hand using Mean of 3 (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.050) \u0026ndash; showed minor deviations. Due to the relatively large sample sizes, parametric analyses were considered sufficiently robust for these tests.\u003c/p\u003e \u003cp\u003eTest\u0026ndash;retest reliability and measurement precision across all conditions are summarised in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, which includes ICC, SEM, and MDC values for each group, hand, and reporting approach. Figure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e complements this table by visualising key reliability metrics across groups and reporting approaches.\u003c/p\u003e \u003cp\u003eOverall, no significant differences in HGS were observed between Session 1 and Session 2 across hands and reporting approaches (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05), except for the right-hand Trial 1 score, which increased from 21.83 kg (SD\u0026thinsp;=\u0026thinsp;8.76) to 22.68 kg (SD\u0026thinsp;=\u0026thinsp;8.84) (\u003cem\u003et\u003c/em\u003e(70) = \u0026minus;\u0026thinsp;2.55, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.013). Group-level results are presented in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. A significant increase was observed only in the F2F_IND group (right hand, Trial 1), from 21.95 kg (SD\u0026thinsp;=\u0026thinsp;11.88) to 24.00 kg (SD\u0026thinsp;=\u0026thinsp;10.94) (\u003cem\u003et\u003c/em\u003e(11) = \u0026minus;\u0026thinsp;2.43, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.033; indicated with *).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eTest\u0026ndash;retest reliability and measurement precision of HGS across all study groups. HGS values, SEM, and MDC are reported in kg. SEM% and MDC% are reported as percentages to aid interpretation relative to group means. * indicates a significant difference in HGS between Session 1 and 2. One trial from the IND_VID group was excluded due to an implausible reading of 0.3 kg.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"9\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGroup\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHand\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eN\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eReporting\u003c/p\u003e \u003cp\u003eStatistics\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSession 1 Mean (SD)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eSession 2\u003c/p\u003e \u003cp\u003eMean (SD)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eICC\u003c/p\u003e \u003cp\u003e(95% CI)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eSEM\u003c/p\u003e \u003cp\u003e(SEM%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eMDC\u003c/p\u003e \u003cp\u003e(MDC%)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e \u003cp\u003e\u003cb\u003eIND_IND\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTrial 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e19.51 (6.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e20.51 (7.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98 (0.91\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.0 (5.01)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e2.78 (13.89)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e20.07 (7.02)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e20.36 (7.74)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98 (0.92\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.12 (5.52)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.1 (15.31)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMax of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e21.27 (7.16)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e22.28 (7.77)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98 (0.92\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.12 (5.16)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.12 (14.31)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTrial 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e19.74 (5.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e21.38 (6.94)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.93 (0.73\u0026ndash;0.98)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.58 (7.71)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e4.39 (21.37)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e20.88 (5.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e21.83 (7.48)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.97 (0.9\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.15 (5.41)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.2 (14.99)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMax of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e22.33 (6.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e23.23 (7.24)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98 (0.93\u0026ndash;1.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.84 (3.67)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e2.32 (10.18)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e \u003cp\u003e\u003cb\u003eIND_VID\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTrial 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e20.93 (7.82)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e22.58 (7.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.94 (0.78\u0026ndash;0.98)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.87 (8.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e5.18 (23.83)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e20.72 (7.86)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e21.68 (6.46)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.94 (0.81\u0026ndash;0.98)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.7 (8.04)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e4.72 (22.28)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMax of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e22.44 (8.67)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e23.42 (7.33)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.95 (0.85\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.72 (7.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e4.77 (20.79)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTrial 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e23.57 (9.53)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e24.26 (8.78)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.97 (0.91\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.47 (6.14)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e4.07 (17.02)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e23.51 (9.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e24.64 (9.56)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98 (0.94\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.32 (5.47)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.65 (15.17)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMax of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e25.53 (10.19)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e25.66 (9.74)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98 (0.94\u0026ndash;1.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.28 (5.01)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.55 (13.88)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e \u003cp\u003e\u003cb\u003eF2F_IND\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTrial 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e20.73 (9.91)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e21.5 (9.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.97 (0.9\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.62 (7.67)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e4.49 (21.26)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e20.8 (9.78)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e21.39 (8.85)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.99 (0.96\u0026ndash;1.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.01 (4.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e2.81 (13.31)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMax of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e22.54 (10.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e22.66 (9.14)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98 (0.94\u0026ndash;1.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.25 (5.51)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.45 (15.28)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eTrial 1*\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e21.95 (11.88)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e24.0 (10.94)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98 (0.88\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.74 (7.56)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e4.82 (20.97)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e22.64 (11.12)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e23.86 (10.12)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.99 (0.95\u0026ndash;1.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.19 (5.14)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.31 (14.23)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMax of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e24.53 (11.23)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e25.43 (10.56)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.99 (0.96\u0026ndash;1.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.14 (4.57)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.17 (12.68)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e \u003cp\u003e\u003cb\u003eVID_IND\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTrial 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e19.72 (6.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e19.11 (6.17)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.94 (0.81\u0026ndash;0.98)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.46 (7.54)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e4.06 (20.9)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e19.59 (6.37)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e18.77 (5.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.95 (0.84\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.33 (6.96)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.7 (19.3)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMax of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e20.71 (6.38)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e20.42 (6.16)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.96 (0.86\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.26 (6.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.5 (17.04)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTrial 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e19.69 (7.52)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e19.92 (6.62)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.96 (0.87\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.39 (7.01)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.85 (19.44)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e19.94 (6.87)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e19.79 (6.41)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98 (0.92\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.03 (5.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e2.86 (14.42)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMax of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e21.1 (7.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e20.79 (6.33)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.97 (0.91\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.08 (5.17)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.0 (14.33)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e \u003cp\u003e\u003cb\u003eF2F_VID\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTrial 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e23.22 (8.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e23.07 (9.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98 (0.94\u0026ndash;1.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.15 (4.98)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.2 (13.82)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e21.98 (7.77)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e22.26 (8.97)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.99 (0.96\u0026ndash;1.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.96 (4.35)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e2.66 (12.05)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMax of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e23.93 (8.42)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e23.92 (9.71)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98 (0.93\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.27 (5.32)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.53 (14.74)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTrial 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e23.91 (8.32)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e24.43 (10.56)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98 (0.93\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.37 (5.66)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.8 (15.7)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e23.26 (8.91)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e23.29 (9.02)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.99 (0.96\u0026ndash;1.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.99 (4.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e2.74 (11.79)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMax of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e24.85 (9.06)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e25.35 (9.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.99 (0.95\u0026ndash;1.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.11 (4.42)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.08 (12.26)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e \u003cp\u003e\u003cb\u003eVID_VID\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTrial 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e20.99 (7.51)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e20.42 (7.35)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98 (0.94\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.98 (4.74)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e2.72 (13.14)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e20.28 (6.86)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e19.94 (7.57)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.99 (0.98\u0026ndash;1.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.62 (3.06)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e1.71 (8.48)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMax of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e21.99 (7.61)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e20.93 (7.92)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98 (0.93\u0026ndash;1.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.0 (4.67)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e2.78 (12.94)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTrial 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e21.97 (9.39)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e22.0 (9.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98 (0.93\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.27 (5.76)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3.51 (15.96)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e21.9 (8.62)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e21.58 (8.78)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.99 (0.98\u0026ndash;1.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.68 (3.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e1.9 (8.73)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMax of 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e23.18 (8.86)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e22.89 (8.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.99 (0.97\u0026ndash;1.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.85 (3.71)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e2.37 (10.28)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAll groups demonstrated excellent test\u0026ndash;retest reliability (ICC\u0026thinsp;\u0026ge;\u0026thinsp;0.93, ICC lower bounds\u0026thinsp;\u0026gt;\u0026thinsp;0.73). The groups that received full procedural supervision during both onboarding and retesting \u0026ndash; VID_VID and F2F_VID \u0026ndash; demonstrated the most reliable and precise test-retest performance. These groups showed the lowest MDC% values (e.g., VID_VID: 11.52% [L], 11.66% [R]; F2F_VID: 13.54% [L], 13.25% [R]) and ICC lower bounds consistently in the \u0026ldquo;excellent\u0026rdquo; range (VID_VID: 0.95\u0026ndash;0.96; F2F_VID: 0.94\u0026ndash;0.95).\u003c/p\u003e \u003cp\u003eIn contrast, partially supervised groups \u0026ndash; F2F_IND, VID_IND, and IND_VID \u0026ndash; showed higher MDC% values, although all ICC lower bounds remained within the \u0026ldquo;good\u0026rdquo; to \u0026ldquo;excellent\u0026rdquo; range. For example, F2F_IND reached 16.62% (L) and 15.96% (R), with ICCs around 0.93; VID_IND and IND_VID showed slightly more variation in MDC% (e.g., IND_VID (L): 22.30%) and ICCs ranging from 0.81\u0026ndash;0.93.\u003c/p\u003e \u003cp\u003eThe fully unsupervised group (IND_IND) did not exhibit the worst average performance in any single metric but showed the widest variability, particularly for the right hand. While MDC% values were moderate overall (14.50% [L], 15.51% [R]), their standard deviations differed notably \u0026ndash; 0.73 for the left, but 5.61 for the right. Similarly, ICC lower bounds averaged within the \u0026ldquo;good\u0026rdquo; range (0.92 [L], 0.85 [R]), but again, the right hand showed substantial spread (SD\u0026thinsp;=\u0026thinsp;0.11), driven by a single case (Trial 1, R) dipping just below the 0.75 threshold, into the \u0026ldquo;moderate\u0026rdquo; range of 0.73.\u003c/p\u003e \u003cp\u003eOverall, these findings suggest that increased supervision leads to more precise and reliable HGS measurements, particularly under \u003cem\u003eMean of 3\u003c/em\u003e and \u003cem\u003eMax of 3\u003c/em\u003e reporting, while variability remains most pronounced in single-trial assessments with limited procedural oversight. These trends were mirrored in the Bland\u0026ndash;Altman analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Mean differences across sessions were small in all groups (ranging from \u0026minus;\u0026thinsp;0.68 to 1.41 kg), but the smallest LoA were seen in fully supervised conditions \u0026ndash; particularly VID_VID (e.g., -2.84 to 2.18 kg for the average of three trials). Wider LoAs were found in partially supervised and unsupervised groups such as IND_VID and F2F_IND, with IND_VID trial 1 scores showing the broadest range (\u0026ndash;5.00 to 7.35 kg).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo quantify the influence of procedural supervision on measurement consistency, we examined associations between supervision score and four key test-retest metrics \u0026ndash; ICC lower bounds, absolute mean difference, MDC% and LoA width \u0026ndash; for each HGS reporting approach (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Overall, higher levels of supervision were associated with better measurement precision and reliability, particularly for the Trial 1 and Mean of 3 reporting approaches.\u003c/p\u003e \u003cp\u003eFor Trial 1, the supervision score was positively associated with ICC lower bound (ρ\u0026thinsp;=\u0026thinsp;0.71, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.010) and was strongly associated with lower absolute mean difference (ρ = -0.74, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.006), suggesting more consistent scores with increased supervision. The correlation with MDC% was moderate but marginally significant (ρ = \u0026minus;\u0026thinsp;0.55, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.062), while LoA width showed no significant relationship (ρ = \u0026minus;\u0026thinsp;0.26, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.411). The Mean of 3 method demonstrated the clearest pattern across all metrics, with supervision significantly associated with increased ICC lower bound (ρ\u0026thinsp;=\u0026thinsp;0.78, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.003), lower MDC% (ρ = \u0026minus;\u0026thinsp;0.82, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.001), and reduced LoA width (ρ = \u0026minus;\u0026thinsp;0.78, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.003). Although the correlation with mean difference was weaker (ρ = \u0026minus;\u0026thinsp;0.49, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.102), the overall trend indicates enhanced test-retest reliability with procedural oversight. By contrast, the Max of 3 method exhibited weak and non-significant associations across all metrics, including ICC (ρ\u0026thinsp;=\u0026thinsp;0.42, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.172). This suggests that the Max approach may be inherently more stable across assessment conditions, or conversely, less sensitive to changes in protocol compliance.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eProtocol Compliance Results\u003c/h2\u003e \u003cp\u003eOf the 72 participants, 41 had video data for both Session 1 and Session 2. An additional 6 participants had video data for Session 1 only, and 3 had data for Session 2 only. The remaining 22 participants had missing recordings due to lack of consent. In total, 50 participants (69.4%) contributed at least one session to the compliance-based analysis.\u003c/p\u003e \u003cp\u003eNormality of compliance scores was assessed using the Shapiro\u0026ndash;Wilk test, which indicated significant deviation from normality in both sessions (Session 1: \u003cem\u003eW\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.877, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Session 2: \u003cem\u003eW\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.866, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Given these violations of the normality assumption, the Wilcoxon signed-rank test was used to compare scores across sessions.\u003c/p\u003e \u003cp\u003eAmong the 41 participants with video data for both sessions, compliance scores were significantly lower in Session 2 compared to Session 1 (\u003cem\u003eW\u003c/em\u003e\u0026thinsp;=\u0026thinsp;87.5, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.013). Mean scores declined from 22.12 (SD\u0026thinsp;=\u0026thinsp;1.83) to 21.41 (SD\u0026thinsp;=\u0026thinsp;2.17). At the individual level, 21 participants (51%) showed a decrease in compliance, 14 (34%) showed no change, and 6 (15%) showed an increase.\u003c/p\u003e \u003cp\u003eAs shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, compliance scores increased with greater levels of supervision during remote assessment, across both individual sessions and the mean of both sessions. In Session 1, the Spearman correlation between supervision score and compliance score was moderate with ρ\u0026thinsp;=\u0026thinsp;0.51 (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.000), while Session 2 was slightly lower at ρ\u0026thinsp;=\u0026thinsp;0.36 (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.017). When compliance scores were averaged across both sessions, the correlation remained moderate (ρ\u0026thinsp;=\u0026thinsp;0.49, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.000), indicating a consistent relationship between procedural supervision and compliance. There was, however, considerable variation in compliance scores within each supervision level, as illustrated by the spread of individual data points.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFurther analyses explored whether protocol compliance influenced test-retest performance. Participants were grouped into Low (N\u0026thinsp;=\u0026thinsp;7), Medium (N\u0026thinsp;=\u0026thinsp;21), and High (N\u0026thinsp;=\u0026thinsp;22) compliance levels based on their mean compliance scores across sessions (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e\u003cb\u003e)\u003c/b\u003e. Higher compliance was associated with improved reliability for the Mean of 3 and Max of 3 scoring methods. Specifically, significant positive correlations were observed between compliance level and ICC lower bounds for both Mean of 3 (ρ\u0026thinsp;=\u0026thinsp;0.96, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.003) and Max of 3 (ρ\u0026thinsp;=\u0026thinsp;0.84, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.038). For Max of 3, compliance was also significantly associated with reduced absolute mean difference in HGS between sessions (ρ = \u0026minus;\u0026thinsp;0.84, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.038), suggesting better within-subject agreement. No significant associations were observed for the Trial 1 metric across any outcome.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eParticipant Questionnaire Results\u003c/h2\u003e \u003cp\u003eAll participants (N\u0026thinsp;=\u0026thinsp;72) provided responses. Overall, participants found the remote setup and assessment highly accessible and user-friendly. Nearly 90% rated the setup and HGS assessment instructions as \u0026ldquo;clear\u0026rdquo; or \u0026ldquo;very clear,\u0026rdquo; and over 97% reported the device was \u0026ldquo;easy\u0026rdquo; or \u0026ldquo;very easy\u0026rdquo; to use. Most (76%) felt comfortable or very comfortable performing the assessment without supervision, and 58% believed the app accurately captured their strength. While 73.6% reported no difficulties, about a quarter experienced issues \u0026ndash; 16.7% reported posture challenges alone, 4.2% reported posture plus difficulty understanding instructions, 1.4% reported posture plus physical discomfort, 1.4% experienced all three, and 2.8% reported physical discomfort only. No participants reported technical difficulties. Finally, satisfaction was uniformly high: over 95% of respondents were \u0026ldquo;satisfied\u0026rdquo; or \u0026ldquo;very satisfied\u0026rdquo; with their HGS assessment experience.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study demonstrates that remote HGS assessment yields excellent or good-to-excellent test-retest reliability, with ICCs consistently above 0.93 (95% CI 0.73-1.0) across all reporting approaches, hands, and supervision groups. These findings support the feasibility and reliability of remote assessment of HGS using the GripAble sensor and Able Assess platform. While ICC reflects between-subject variance and is useful for assessing consistency across a sample, the MDC provides complementary insight into within-subject agreement \u0026ndash; or measurement precision \u0026ndash; which is especially relevant in heterogeneous populations (Atkinson and Nevill \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1998\u003c/span\u003e). In our study, MDC% varied across groups, suggesting that measurement precision may be influenced by the degree of procedural supervision and structured guidance.\u003c/p\u003e \u003cp\u003eA high MDC% in HGS between sessions indicates that small changes may go undetected, making it harder to distinguish real improvements or declines. While no universal threshold exists, MDC% in healthy adults typically ranges from 10\u0026ndash;20% of measured grip force (Savas et al. 2023), with values below 19.5% sometimes cited as indicative of good precision in clinical settings (Bohannon \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Kim, Park, and Shin 2014). In our study, only the fully supervised groups (VID_VID and F2F_VID) consistently met this benchmark across all reporting approaches, with the lowest MDC% observed in VID_VID (8.48%, left-hand mean). In contrast, groups with partial or no supervision demonstrated greater variability, especially with the Trial 1 reporting approach, where MDC% often exceeded 20% (e.g., VID_IND left: 23.8%, F2F_IND right: 21.3%), while the mean of three trials or max scores frequently hovered near the threshold.\u003c/p\u003e \u003cp\u003eThese findings show that some level of supervision, either in-person or remote, can help reduce variability and improve measurement consistency especially if this supervision is established early, during onboarding. However, previous in-person test-retest studies using GripAble have, despite high ICCs, also reported MDC% values exceeding 20% but only during single trial usage (Kudin, Ergen, and McGee 2025). Therefore, utilising a three trial reporting approach (Mean of 3 or Max of 3) will improve precision regardless of supervision level. The persistence of elevated MDC% even in in-person settings may also indicate that factors beyond supervision alone, such as the device\u0026rsquo;s susceptibility to procedural inconsistencies between sessions (e.g. subtle changes in hand placement or grip execution) may contribute to reduced measurement precision. In contrast, interrater studies conducted within a single session have shown MDC% remaining below 15% (Ergen, Kudin, and McGee 2024) \u0026ndash; possibly because participants are more likely to maintain consistent positioning and technique across closely timed trials. Additional factors such as, but not limited to, measurement time of day, fatigue levels, and previous night\u0026rsquo;s sleep will also affect measurement accuracy and were not controlled in this study to mimic real-world use.\u003c/p\u003e \u003cp\u003eCorrelation analyses confirmed the decline in measurement precision with reduced supervision, showing strong associations between supervision score and MDC%, LoA width, and mean difference for selected reporting approaches. These results support the observation that greater supervision enhances both consistency and agreement. Additionally, some of the partially supervised groups (IND_VID and VID_IND) showed reduced reliability and precision of the left hand when compared to the right hand (see Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Given the majority of participants were right-handed, this indicates that hand dominance is also a factor with the non-dominant hand being more affected by the level of supervision.\u003c/p\u003e \u003cp\u003eBoth maximum and average scores generally yielded lower MDC% values, but neither consistently outperformed the other. Maximum scores showed weaker associations with supervision, suggesting greater resistance to protocol deviations \u0026ndash; a potential advantage in unsupervised settings. In contrast, average scores more often reflected supervision effects. Therefore, using maximum values may offer greater stability across varied conditions.\u003c/p\u003e \u003cp\u003eLower compliance \u0026ndash; particularly in fully unsupervised groups \u0026ndash; was likely driven by misinterpretation of instructions or inconsistent assessment execution. These behavioural observations were consistent with reduced test\u0026ndash;retest reliability, reflected in lower ICC values in the same groups. However, the relationship between supervision level and compliance scores was inconsistent, with notable variation observed within supervision groups. This variability may partly reflect individual differences in engagement or motivation, which was not formally assessed in this study (e.g., using the Intrinsic Motivation Inventory). Compliance assessment relied solely on observable behaviours captured on video, which may overlook subtler deviations. This was supported by participant feedback: although most rated the instructions and device usability as clear and easy, about a quarter still reported posture-related difficulties during the test. Future implementations could benefit from sensor-based tools that enable detection of non-observable non-compliance and support real-time guidance to improve protocol compliance in remote settings.\u003c/p\u003e \u003cp\u003eBeyond measurement metrics, user feedback highlighted the feasibility of remote HGS assessment: the vast majority of participants found the setup intuitive, the assessment easy to complete, and the overall experience highly satisfactory. Most felt comfortable performing the assessment without supervision, and few reported environmental barriers, reinforcing the practicality of remote HGS assessment in home settings.\u003c/p\u003e \u003cp\u003eThis study has several limitations. Firstly, the absence of a gold-standard in-person baseline makes it difficult to determine whether consistency in groups like IND_IND reflects true performance or systematic error. We intentionally omitted an in-person session to preserve participant naivety; however, including it as a third session would have enabled protocol benchmarking. Secondly, compliance was rated by a single observer and focused on overt postural deviations, potentially overlooking more subtle errors. Incorporating multiple raters and sensor-derived metrics in future studies could improve scoring reliability. Lastly, while this study used the GripAble sensor and Able Assess platform as the assessment framework, the generalisability of these findings to other remote assessment tools will depend on the extent to which those platforms support similar procedural structure, user guidance, and compliance monitoring.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eDespite the growing evidence of HGS as a new vital sign of general health, it is yet to be embedded into routine practice and used across the care continuum in both supervised and unsupervised use-cases. This study provides evidence that remote HGS assessment using digital tools such as the GripAble sensor and Able Assess platform is feasible and yields high test-retest reliability under a range of real world conditions. Nonetheless, measurement precision was influenced by level of supervision and when introduced, protocol compliance, and trial aggregation, to name a few, reflecting the multifactorial nature of reliability in decentralised assessment environments. While fully unsupervised assessment may be viable in select populations \u0026ndash; particularly those with prior training or strong intrinsic compliance \u0026ndash; small reductions in individual-level precision suggest real world HGS assessment may benefit from additional procedural or technological safeguards. A key recommendation based on these findings is that a supervised baseline measurement, prior to any remote testing, should improve measurement precision during repeated testing of an individual.\u003c/p\u003e \u003cp\u003eTherefore, future implementations should utilise supervised onboarding alongside adaptive assessment protocols, embedded instructional support, real-time feedback, and post-hoc validation using, for example, sensor-derived metrics to reinforce protocol compliance. Importantly, such enhancements should be designed to support \u0026ndash; not burden \u0026ndash; usability, especially for end users in remote or unsupervised settings. As remote assessment tools are increasingly adopted in clinical and research contexts, their impact will depend not only on technical accuracy, but also on their ability to deliver reliable, actionable data within systems that prioritise both data quality and patient-centred care.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor contributions\u003c/h2\u003e\n\u003cp\u003eSAM and MM conceived the study and designed the protocol. SAM conducted the data analysis and drafted the manuscript. DJ collected the data. SP assessed protocol compliance from video recordings. EB supervised the study as principal investigator. All authors reviewed and approved the final manuscript.\u003c/p\u003e\n\u003ch2\u003eAcknowledgements\u003c/h2\u003e\n\u003cp\u003eWe thank all participants for their time and involvement in the study.\u003c/p\u003e\n\u003ch2\u003eEthics approval and consent to participate\u003c/h2\u003e\n\u003cp\u003eEthical approval was gained from the Science Engineering Technology Research Ethics Committee of the Imperial College of Science, Technology and Medicine, London, UK (REF: 20IC5831). All participants gave their written consent to take part in the experi\u0026shy;ment. The experiment followed the ethical standards in the 1964 Declaration of Helsinki.\u003c/p\u003e\n\u003ch2\u003eConsent for publication\u003c/h2\u003e\n\u003cp\u003eAll participants received a thorough explanation of the risks and benefits of inclusion and gave their oral and written informed con\u0026shy;sent to publish the data.\u003c/p\u003e\n\u003ch2\u003eDeclaration of competing interest\u003c/h2\u003e\n\u003cp\u003eSAM, SP, and MM are employees of GRIPABLE Ltd. DJ worked on the study during an internship at GRIPABLE Ltd. MM and EB are the authors on the GripAble sensor patent.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAtkinson, G., and A. M. Nevill. 1998. \u0026ldquo;Statistical Methods for Assessing Measurement Error (reliability) in Variables Relevant to Sports Medicine.\u0026rdquo; \u003cem\u003eSports Medicine (Auckland, N.Z.)\u003c/em\u003e 26 (4): 217\u0026ndash;38.\u003c/li\u003e\n\u003cli\u003eBohannon, Richard W. 2019. \u0026ldquo;Minimal Clinically Important Difference for Grip Strength: A Systematic Review.\u0026rdquo; \u003cem\u003eJournal of Physical Therapy Science\u003c/em\u003e 31 (1): 75\u0026ndash;78.\u003c/li\u003e\n\u003cli\u003eCelis-Morales, Carlos A., Paul Welsh, Donald M. Lyall, Lewis Steell, Fanny Petermann, Jana Anderson, Stamatina Iliodromiti, et al. 2018. \u0026ldquo;Associations of Grip Strength with Cardiovascular, Respiratory, and Cancer Outcomes and All Cause Mortality: Prospective Cohort Study of Half a Million UK Biobank Participants.\u0026rdquo; \u003cem\u003eBMJ (Clinical Research Ed.)\u003c/em\u003e 361 (May):k1651.\u003c/li\u003e\n\u003cli\u003eChang, Yu-Tzu, Hung-Lien Wu, How-Ran Guo, Ya-Yun Cheng, Chin-Chung Tseng, Ming-Cheng Wang, Ching-Yuang Lin, and Junne-Ming Sung. 2011. \u0026ldquo;Handgrip Strength Is an Independent Predictor of Renal Outcomes in Patients with Chronic Kidney Diseases.\u0026rdquo; \u003cem\u003eNephrology, Dialysis, Transplantation : Official Publication of the European Dialysis and Transplant Association - European Renal Association\u003c/em\u003e 26 (11): 3588\u0026ndash;95.\u003c/li\u003e\n\u003cli\u003eErgen, Halil Ibrahim, Roman Kudin, and Corey W. McGee. 2024. \u0026ldquo;Interrater Reliability and Precision of a Novel Hand Strength Assessment and Treatment Device: The GripAble.\u0026rdquo; \u003cem\u003eThe American Journal of Occupational Therapy : Official Publication of the American Occupational Therapy Association\u003c/em\u003e 78 (5). https://doi.org/10.5014/ajot.2024.050689.\u003c/li\u003e\n\u003cli\u003eHailu, Ruth, Jessica Sousa, Mitchell Tang, Ateev Mehrotra, and Lori Uscher-Pines. 2024. \u0026ldquo;Challenges and Facilitators in Implementing Remote Patient Monitoring Programs in Primary Care.\u0026rdquo; \u003cem\u003eJournal of General Internal Medicine\u003c/em\u003e 39 (13): 2471\u0026ndash;77.\u003c/li\u003e\n\u003cli\u003eHeslop, Philip A., Christopher Hurst, Avan A. Sayer, and Miles D. Witham. 2023. \u0026ldquo;Remote Collection of Physical Performance Measures for Older People: A Systematic Review.\u0026rdquo; \u003cem\u003eAge and Ageing\u003c/em\u003e 52 (1). https://doi.org/10.1093/ageing/afac327.\u003c/li\u003e\n\u003cli\u003eHoenemeyer, Teri W., William W. Cole, Robert A. Oster, Dorothy W. Pekmezi, Andrea Pye, and Wendy Demark-Wahnefried. 2022. \u0026ldquo;Test/Retest Reliability and Validity of Remote vs. In-Person Anthropometric and Physical Performance Assessments in Cancer Survivors and Supportive Partners.\u0026rdquo; \u003cem\u003eCancers\u003c/em\u003e 14 (4): 1075.\u003c/li\u003e\n\u003cli\u003eKim, Jae Kwang, Min Gyue Park, and Sung Joon Shin. 2014. \u0026ldquo;What Is the Minimum Clinically Important Difference in Grip Strength?\u0026rdquo; \u003cem\u003eClinical Orthopaedics and Related Research\u003c/em\u003e 472 (8): 2536\u0026ndash;41.\u003c/li\u003e\n\u003cli\u003eKlein, Thorsten, Annette Worth, Claudia Niessner, and Anke Hanssen-Doose. 2025. \u0026ldquo;Remote Assessment of Physical Fitness via Videoconferencing: A Systematic Review.\u0026rdquo; \u003cem\u003eBMC Sports Science, Medicine \u0026amp; Rehabilitation\u003c/em\u003e 17 (1): 11.\u003c/li\u003e\n\u003cli\u003eKoo, Terry K., and Mae Y. Li. 2016. \u0026ldquo;A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research.\u0026rdquo; \u003cem\u003eJournal of Chiropractic Medicine\u003c/em\u003e 15 (2): 155\u0026ndash;63.\u003c/li\u003e\n\u003cli\u003eKudin, Roman, Halil Ibrahim Ergen, and Corey W. McGee. 2025. \u0026ldquo;Test-Retest Reliability and Precision of GripAble: A Multipurpose Exergaming Device.\u0026rdquo; \u003cem\u003eGames for Health Journal\u003c/em\u003e, February. https://doi.org/10.1089/g4h.2024.0216.\u003c/li\u003e\n\u003cli\u003eLeong, Darryl P., Koon K. Teo, Sumathy Rangarajan, Patricio Lopez-Jaramillo, Alvaro Avezum Jr, Andres Orlandini, Pamela Seron, et al. 2015. \u0026ldquo;Prognostic Value of Grip Strength: Findings from the Prospective Urban Rural Epidemiology (PURE) Study.\u0026rdquo; \u003cem\u003eLancet (London, England)\u003c/em\u003e 386 (9990): 266\u0026ndash;73.\u003c/li\u003e\n\u003cli\u003eLi, Joule J., Gary A. Wittert, Andrew Vincent, Evan Atlantis, Zumin Shi, Sarah L. Appleton, Catherine L. Hill, Alicia J. Jenkins, Andrzej S. Januszewski, and Robert J. Adams. 2016. \u0026ldquo;Muscle Grip Strength Predicts Incident Type 2 Diabetes: Population-Based Cohort Study.\u0026rdquo; \u003cem\u003eMetabolism: Clinical and Experimental\u003c/em\u003e 65 (6): 883\u0026ndash;92.\u003c/li\u003e\n\u003cli\u003eLoring, David W., James J. Lah, and Felicia C. Goldstein. 2023. \u0026ldquo;Telehealth Equivalence of the Montreal Cognitive Assessment (MoCA): Results from the Emory Healthy Brain Study (EHBS).\u0026rdquo; \u003cem\u003eJournal of the American Geriatrics Society\u003c/em\u003e 71 (6): 1931\u0026ndash;36.\u003c/li\u003e\n\u003cli\u003eMace, Michael, Sharah Abdul Mutalib, Matjaz Ogrinc, Nicola Goldsmith, and Etienne Burdet. 2022. \u0026ldquo;GripAble: An Accurate, Sensitive and Robust Digital Device for Measuring Grip Strength.\u0026rdquo; \u003cem\u003eJournal of Rehabilitation and Assistive Technologies Engineering\u003c/em\u003e 9 (March):20556683221078455.\u003c/li\u003e\n\u003cli\u003eMutalib, Sharah Abdul, Michael Mace, Chloe Seager, Etienne Burdet, Virgil Mathiowetz, and Nicola Goldsmith. 2022. \u0026ldquo;Modernising Grip Dynamometry: Inter-Instrument Reliability between GripAble and Jamar.\u0026rdquo; \u003cem\u003eBMC Musculoskeletal Disorders\u003c/em\u003e 23 (1): 80.\u003c/li\u003e\n\u003cli\u003eMutalib, Sharah Abdul, Deepika Sharma, Sonia Pike, Liz Gwynne, Samantha Hyde, Jennifer Morehouse, Helen Davey, et al. 2024. \u0026ldquo;GripAble: Interrater Reliability and Normative Grip Strength of UK Population.\u0026rdquo; \u003cem\u003eJournal of Hand Therapy : Official Journal of the American Society of Hand Therapists\u003c/em\u003e, March. https://doi.org/10.1016/j.jht.2023.12.013.\u003c/li\u003e\n\u003cli\u003eNaef, Aileen C., Guichande Duarte, Saskia Neumann, Migjen Shala, Meret Branscheidt, and Chris Easthope Awai. 2025. \u0026ldquo;Toward Unsupervised Capacity Assessments for Gait in Neurorehabilitation: Validation Study.\u0026rdquo; \u003cem\u003eJournal of Medical Internet Research\u003c/em\u003e 27 (March):e66123.\u003c/li\u003e\n\u003cli\u003eSavas, Sumru, Asli Kilavuz, Fatma \u0026Ouml;zge Kayhan Ko\u0026ccedil;ak, and Sibel Cavdar. 2023. \u0026ldquo;Comparison of Grip Strength Measurements by Widely Used Three Dynamometers in Outpatients Aged 60 Years and Over.\u0026rdquo; \u003cem\u003eJournal of Clinical Medicine\u003c/em\u003e 12 (13). https://doi.org/10.3390/jcm12134260.\u003c/li\u003e\n\u003cli\u003eSerrano, Luiza Palmieri, Karla C. Maita, Francisco R. Avila, Ricardo A. Torres-Guzman, John P. Garcia, Abdullah S. Eldaly, Clifton R. Haider, et al. 2023. \u0026ldquo;Benefits and Challenges of Remote Patient Monitoring as Perceived by Health Care Practitioners: A Systematic Review.\u0026rdquo; \u003cem\u003eThe Permanente Journal\u003c/em\u003e 27 (4): 100\u0026ndash;111.\u003c/li\u003e\n\u003cli\u003eTomkinson, Grant R., Justin J. Lang, Luk\u0026aacute;\u0026scaron; Rub\u0026iacute;n, Ryan McGrath, Bethany Gower, Terry Boyle, Marilyn G. Klug, et al. 2024. \u0026ldquo;International Norms for Adult Handgrip Strength: A Systematic Review of Data on 2.4 Million Adults Aged 20 to 100+ Years from 69 Countries and Regions.\u0026rdquo; \u003cem\u003eJournal of Sport and Health Science\u003c/em\u003e 14 (December):101014.\u003c/li\u003e\n\u003cli\u003eVaishya, Raju, Anoop Misra, Abhishek Vaish, Nicola Ursino, and Riccardo D\u0026rsquo;Ambrosi. 2024. \u0026ldquo;Hand Grip Strength as a Proposed New Vital Sign of Health: A Narrative Review of Evidences.\u0026rdquo; \u003cem\u003eJournal of Health, Population, and Nutrition\u003c/em\u003e 43 (1): 7.\u003cbr\u003e\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Handgrip strength, dynamometry, real world, remote, test-retest, measurement, reliability, precision","lastPublishedDoi":"10.21203/rs.3.rs-6801648/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6801648/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eHandgrip strength (HGS) is a key indicator of health and functional status. As remote health assessments become more common, it is critical to understand how procedural supervision influences the reliability of remote HGS assessment.\u003c/p\u003e\u003ch2\u003eObjective\u003c/h2\u003e \u003cp\u003eTo evaluate the test-retest reliability and measurement precision of remote HGS assessment under varying levels of procedural supervision as found in real world use.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eSeventy-two adults were randomised into six groups reflecting different supervision levels over two sessions. HGS was measured using the GripAble Sensor and Able Assess platform. Test\u0026ndash;retest reliability was evaluated using intraclass correlation coefficients (ICC), while measurement precision was quantified using the standard error of measurement (SEM) and minimal detectable change (MDC%). Agreement between sessions was further assessed using Bland\u0026ndash;Altman analysis, reporting mean difference and 95% limits of agreement (LoA). Protocol compliance was rated from video recordings. Participants also completed a post-session questionnaire on remote assessment experience.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eAll groups demonstrated a good-to-excellent test-retest reliability (ICC\u0026thinsp;\u0026ge;\u0026thinsp;0.93, ICC lower bounds\u0026thinsp;\u0026ge;\u0026thinsp;0.73), but measurement precision varied. Fully supervised groups achieved the lowest MDC% (as low as 8.5%), while unsupervised groups often exceeded 20% in the single trial reporting approach, indicating reduced sensitivity to change. Higher supervision corresponded with better protocol compliance. Participant feedback demonstrated high usability during real-world use: 97% rated the test as easy or very easy, \u0026gt;\u0026thinsp;75% felt comfortable performing it remotely, and \u0026gt;\u0026thinsp;95% were satisfied with the experience.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eRemote HGS assessment shows high reliability, but measurement precision is shaped by supervision and procedural compliance. Based on these findings, it is recommended that to maximise measurement precision during remote sessions, in-person supervision should be provided during onboarding, possibly followed up with periodic supervision when conducting repeated longitudinal measurements of an individual. Integrating structured features, including standardised instructions, user specified configuration and compliance monitoring will further improve remote measurement performance without undermining usability during real-world use.\u003c/p\u003e","manuscriptTitle":"Remote, Reliable, Repeatable: Real-World Test–Retest Validation of Hand Grip Strength Assessments","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-04 12:21:18","doi":"10.21203/rs.3.rs-6801648/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"7b0275cf-97f0-41ba-8b65-567be71e260c","owner":[],"postedDate":"June 4th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-07-17T15:08:49+00:00","versionOfRecord":[],"versionCreatedAt":"2025-06-04 12:21:18","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6801648","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6801648","identity":"rs-6801648","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}