Sensor-based measures of knee brace adherence have low agreement with self-report methods: A multi-measure study among knee osteoarthritis patients.

Michael J Callaghan; David T Felson; Melanie A Holden; Belinda Borrelli; Elaine Nicholls; Fraser Birrell; Nadine E Foster; Nicola Halliday; Carol Ingram; Clare Jinks; George M Peat; PROP OA team (see table S1)

doi:10.1016/j.joca.2026.06.005

PubMed · 42442456

Sensor-based measures of knee brace adherence have low agreement with self-report methods: A multi-measure study among knee osteoarthritis patients.

Michael J Callaghan·David T Felson·Melanie A Holden·Belinda Borrelli·Elaine Nicholls·Fraser Birrell·Nadine E Foster·Nicola Halliday·Carol Ingram·Clare Jinks·George M Peat·PROP OA team (see table S1)

View original source ↗Export metadata ↓

Abstract

OBJECTIVE: To explore agreement between self-report and objectively measured adherence to brace wearing by patients with knee osteoarthritis. METHOD: A single-arm observational analysis nested within the PROP OA randomised controlled trial (ISRCTN28555470). Of 237 adults with symptomatic knee osteoarthritis randomised to brace treatment, 60 were included in this sub-study investigating three different methods of assessing knee brace wear time over 26 weeks: 1. Self-report questionnaires (SRQ) at 12 weeks and 26 weeks; 2. Short message service (SMS) questions (days worn in past week, typical hours per day when worn) administered from week 1 to week 24; 3. A skin temperature sensor embedded in the brace, sampling every 10 min for 26 weeks. The presence and reason for the sensor were concealed from participants. The estimated proportion of participants meeting "minimum brace use", defined a priori as ≥1 h on ≥2 days in past week, was described for each measurement method, overall and by brace type (unloader, neutral). For temperature sensor measurements, time spent above 24°C and time spent above 25°C were used. Agreement between the measures was summarised by percentage agreement and kappa (ĸ). RESULTS: The estimated proportions of participants meeting "minimum brace use" at 12 weeks were 83% (SRQ), 83% (SMS), 60% and 58% (temperature sensor, 24°C and 25°C thresholds, respectively). At 26 weeks, the corresponding estimates reduced to 72%, 71% (SMS at 24 weeks), 43% and 37%. Sensor data suggested the sharpest decline in brace use occurred within the first 12 weeks. Agreement between self-report measures was higher than between self-report measures and sensor (SRQ vs SMS at 12 weeks: 92% agreement, ĸ=0.67 (95%CI: 0.34, 1.00); SRQ vs Sensor at 12 weeks: 74%, 0.35 (0.10, 0.60); SMS vs Sens at 12 weeks: 76%, 0.36 (0.05, 0.66). Agreement between all measurement methods reduced at 26 weeks. CONCLUSIONS: This novel use of a temperature sensor to monitor brace adherence in knee osteoarthritis indicates that self-report adherence substantially overestimates knee brace wearing time, with implications for clinical trials and practice.

Explore related subjects

Humans Osteoarthritis, Knee Braces Patient Compliance Female Self Report Middle Aged Male Aged Skin Temperature Surveys and Questionnaires Adherence Bracing Knee Osteoarthritis Sensors Journal Article Randomized Controlled Trial Observational Study

Keep this discovery

BibTeX RIS

Michael J Callaghan, David T Felson, Melanie A Holden, Belinda Borrelli, Elaine Nicholls, Fraser Birrell, Nadine E Foster, Nicola Halliday, Carol Ingram, Clare Jinks, George M Peat, PROP OA team (see table S1). 2026-07-13. Sensor-based measures of knee brace adherence have low agreement with self-report methods: A multi-measure study among knee osteoarthritis patients.. https://doi.org/10.1016/j.joca.2026.06.005

Cite the original work for its findings. Save a collection to share your selection of sources.

Discover connections

Adherence subject Aged subject Female subject Humans subject Journal Article subject Male subject Middle Aged subject Observational Study subject Patient Compliance subject Randomized controlled trials study

Connections use source metadata and explicit phrase matches, not verified experimental comparisons.

KEEP EXPLORING

Related citations

PubMed2026

Effect of an oral complementary medicine combination for symptomatic knee osteoarthritis: A double-blind, randomized, placebo-controlled trial.

OBJECTIVE: Many individuals use complementary medicine therapies for osteoarthritis, despite limited scientific supporting evidence. These therapies have demonstrated anti-inflammatory, antioxidative, and chondroprotective effects via different molecular pathways, suggesting potential therapeutic benefits. This trial aimed to evaluate the efficacy and safety of a novel oral complementary medicine combination in adults with symptomatic knee osteoarthritis at 12 weeks. DESIGN: The 12-week, double-blind, randomized, placebo-controlled ATLAS trial was conducted in Australia from July 2024 to April 2025 (ACTRN12623000380695). Eligible participants were aged ≥40 years with symptomatic knee osteoarthritis and radiographic Kellgren-Lawrence grade ≥2. Participants were randomized 1:1 to receive the complementary medicine combination, comprising a total daily dose of Boswellia serrata extract 250 mg, pine bark extract 100 mg, methylsulfonylmethane 1500 mg, curcumin 500 mg, and piperine 5 mg, or placebo. The primary outcome was the 12-week change in knee pain (0-100 mm visual analogue scale (VAS)). A 95% confidence interval excluding differences >18 mm between groups was interpreted as the absence of a clinically meaningful difference. Harms were monitored weekly. RESULTS: Eighty-four participants were recruited (mean [SD], age 63.7 [8.0] years, BMI 29.0 [5.5] kg/m2), and 57% were female. At 12-weeks, both groups (intention-to-treat population: complementary medicine, n=42; placebo, n=41) showed a decrease in VAS pain (least squares means: complementary medicine 20.1 mm vs placebo 20.3 mm), however with no difference between groups: 0.16 mm (95% CI -6.81 to 7.12). The safety profile of experiencing any AE was similar between groups (risk difference, -3.4% [95% CI -24.8-17.9%]); and most adverse events were mild or moderate. Four participants experienced serious adverse events, unrelated to the treatment. CONCLUSIONS: The complementary medicine combination was not superior to placebo for improving pain over 12 weeks, and the evidence does not support the use of this combination to treat people with symptomatic knee osteoarthritis.

Humans↗

PubMed2026

Weight-bearing locomotion mitigates the progression of post-traumatic knee osteoarthritis in male rodents: A systematic review and meta-analysis.

OBJECTIVE: Although knee post-traumatic osteoarthritis (PTOA) has been associated with altered loading, how weight-bearing (WB) activities should be modulated post-injury to preserve knee health remains unknown. This systematic review and meta-analysis examined the effects of WB locomotion on knee PTOA in rodents to provide insight into potential clinical implications. DESIGN: Studies published in PubMed, Cochrane Library, Embase, and CINAHL through 02/2025 and meeting the following criteria were included: 1) rodents with knee PTOA, 2) compared locomotor exercises (treadmill/wheel) to no exercise, 3) outcome measures of PTOA (OARSI/Mankin OA scores, bone quality, osteophyte condition, cartilage quality/morphology). If significant heterogeneity across studies was observed, locomotion speed, time/number of intervention sessions, frequency and duration of intervention, exercise initiation, follow-up time, animal species, PTOA model were analyzed as potential moderator variables that may influence findings. RESULTS: Twenty-two studies met the criteria, resulting in a total of 156 effect sizes (ESs) for meta-analysis (all males). A positive ES denotes less PTOA. Based on the CAMARADES checklist, 5 studies were ranked low risk of bias and 14 studies were ranked moderate risk of bias. Locomotion significantly reduced PTOA severity evaluated using OARSI/Mankin scores (ES=0.927, 95% CI: [0.590, 1.264]) and improved bone quality (ES=0.379, 95% CI: [-0.023, 0.780]) with substantial heterogeneity across studies. Follow-up analyses indicated that greater ESs for reducing OA scores were associated with a shorter duration of each training session, more training sessions, longer intervention duration, and longer follow-up time (all P≤0.004). Greater ESs for improving bone quality were associated with a longer follow-up time and later exercise initiation post-injury (both P≤0.014). Locomotion resulted in small to moderate but non-significant positive effects for isolated measures assessing osteophyte condition (ES=0.379, 95% CI: [-0.023, 0.780]), cartilage quality (ES=0.430, 95% CI: [-0.034, 0.893]), and cartilage morphology (ES=0.464, 95% CI: [-0.006, 0.934]). CONCLUSION: WB locomotion reduced the overall degree of knee PTOA in male rodents. Composite knee OA scores may be more responsive to exercise interventions than isolated cartilage/osteophyte measures. The benefits of locomotion may be more prominent with a longer follow-up time or with exercise protocols consisting of shorter but more individual sessions and longer overall intervention durations.

Animals↗

PubMed2026

CoLchicine for Treatment of OsteoArthritis of the Knee (CLOAK): Clinical and biochemical outcomes from a three-month double-blind, placebo-controlled study.

OBJECTIVE: Knee osteoarthritis (KOA) causes pain and progressive disability, but pharmacologic treatments are limited. Colchicine inhibits inflammation that might modulate KOA, but efficacy trials have yielded mixed results. We tested whether colchicine, without concurrent NSAIDs, improved KOA pain, function, synovial effusion size, and OA-associated inflammatory serum biomarkers. METHODS: Participants with symptomatic KOA and radiographic Kellgren-Lawrence grades 2/3 were randomized to receive three months of daily colchicine or placebo in a double-blind manner, with no concurrent NSAID use. The primary outcome was between-group change in visual analog score (VAS) for index knee pain. Secondary outcomes included changes in Knee Osteoarthritis Outcome Scores (KOOS), size (depth in millimeters) of sonographically-identified effusions, acetaminophen use, and changes in OA-related serum biomarkers. RESULTS: From baseline to end of study of 120 enrolled participants, no significant differences were observed in improvement of VAS pain, KOOS scores or effusion size. Subsets of participants with more severe VAS pain, worse radiographic disease, or higher hsCRP or serum urate levels at baseline also showed no significant clinical benefit from colchicine compared to placebo. In contrast to the clinical outcomes, colchicine treatment was associated with significant or trending improvement in multiple OA-related serum biomarkers including hsCRP and β-NGF (p < 0.05) and PGE2, IL-1ra, IL-8, and VEGF (p < 0.16). CONCLUSION: This double-blind placebo-controlled trial of colchicine for KOA failed to demonstrate improvement in pain, function, or synovial effusion size in comparison to placebo at three months. Early improvement in OA-associated inflammatory biomarkers suggests a possible longer-term clinical benefit. Clinical Trials Registration No NCT03913442.

Humans↗