Stopping OsteoARthritis After an ACL Tear

Overview

By 2040, 25% of Canadians will have osteoarthritis, a disabling joint condition. Most people think osteoarthritis only affects older adults, but 50% of the 700,000 Canadian youth who hurt their knee playing sports annually will develop osteoarthritis by 40 years of age. These young people with old knees face knee pain and disability for much of their adult lives, interfering with parenting, work, and recreation. Yet, most do not know about osteoarthritis or how to reduce their risk. In this clinical trial, people who have torn the Anterior Cruciate ligament in their knee and had reconstruction surgery 9-36 months previously will be randomized to receive either a 6-month virtual education and exercise therapy program called Stop OsteoARthritis (SOAR) or a minimal intervention control program. Researchers will test if those who received the SOAR program have larger gains in knee health, including pain, symptoms, function, and quality of life at 6, 12, and 24 months. Researchers will also use MRIs (baseline and 24 months) to assess how the SOAR program influences knee cartilage degeneration and its cost-effectiveness.

PURPOSE: Assess the effectiveness, efficiency, and implementation of a 6-month evidence-informed, digital (online), education and exercise-therapy program (SOAR - Stop OsteoARthritis) versus a minimal intervention control for people aged 16-35 years at risk of early-onset knee osteoarthritis (OA) due to a first time Anterior Cruciate Ligament Reconstruction (ACLR). OBJECTIVES: Primary Effectiveness Objective: Assess if self-reported knee pain, symptoms, function, and QoL (average of the Knee injury and OA Outcome Score pain, other symptoms, function in sport and recreation, and quality of life subscale scores; KOOS4) of people at risk of knee OA who receive SOAR is superior to those receiving a minimal intervention CONTROL at 6 (primary end-point), 12, and 24 months. Primary Efficiency Objective: Assess the incremental cost-utility ratio of SOAR compared CONTROL (6, 12, 24 months) Primary Implementation Objective: Assess provider adoption of the SOAR program. Secondary Effectiveness Objectives: 1. Assess if the perceived self-management of people at risk of knee OA who receive SOAR is superior to those receiving a minimal intervention CONTROL (6, 12, 24 months) 2. Assess if the percentage of SOAR group participants who achieve a Patient Acceptable Symptom State (PASS) for knee-related pain, other symptoms, function in daily living, function in sport and recreation, and quality of life is superior to CONTROL (6, 12, 24 months) 3. Assess if knee OA MRI features of SOAR group participants are superior to CONTROL (24 months) Secondary Efficiency Objectives: 1. Describe the health resource use of SOAR and CONTROL participants (24 months) 2. Describe SOAR program delivery costs Secondary Implementation Objectives: 1. Describe provider perceived barriers and facilitators of SOAR delivery 2. Describe participants' perceived barriers and facilitators of SOAR delivery 3. Describe features of physiotherapy (PT) counseling that promote self-management Exploratory Effectiveness Objectives: Explore the superiority of SOAR to CONTROL on the following outcomes (at 6, 12, and 24 months); 1. Patient-specific function 2. Knee-related self-efficacy 3. Knee-related fear of movement and re-injury 4. Knee extensor strength 5. Knee flexor strength 6. Physical activity, 7. Health-related quality-of-life 8. Adiposity 9. Early cartilage degeneration (at 24 months only) RESEARCH DESIGN: The proposed trial is a two-arm, assessor-blinded, superiority, hybrid effectiveness-implementation type 1 randomized controlled trial with embedded cost-utility analyses and 1:1 interviews. This multi-focus design (effectiveness, implementation, efficacy) can accelerate research translation to real-world settings (Type 1: main focus-intervention effectiveness; secondary focus-understand implementation context). This design is appropriate as we have shown feasibility and indirect evidence of SOAR effect, the intervention is minimal risk, and there are no fully powered superiority trials to inform non-inferiority or equivalence designs. The nature of the interventions do not allow for full blinding (physiotherapists cannot be blinded to treatment). We will employ proven methods from our proof-of-concept RCT to reduce allocation (online randomization module with schedule prepared by an arms-length statistician) and confirmation bias (outcome assessors and data analysts will be blinded to allocation). STATISTICAL ANALYSES: Primary analyses will be intent-to-treat (by randomization). Data missing more than 5% will be imputed using multiple imputations by chained equations to avoid bias. To ensure best practice, all outcome and demographic/prognostic variables will be included in imputation equations. Demographics: Descriptive statistics will be calculated for demographic and potentially prognostic variables (time since injury and ACLR, concomitant injury, ACLR rehabilitation, graft type, reinjury, co-intervention, SES) and observed differences considered or controlled for when interpreting findings. Effectiveness: SOAR superiority will be assessed with a generalized linear mixed regression model (GLMM) for the primary outcome (KOOS4 at 6 months) adjusted for baseline measure, time since ACLR, and sex. Adjusting for continuous time since ACLR versus stratifying improves power. Similar GLMMs will assess SOAR superiority for continuous secondary (PIH) and exploratory (knee extensor strength, adiposity, physical activity) outcomes at stated time points. Mixed effect logistic regression models (adjusted for the same variables as GLMMs) will assess SOAR superiority for binary secondary outcomes (MRI lesion worsening, achieving PASS) at stated time points. Analyses will yield valid results under the missing at random (MAR) assumption. The robustness of estimates to potential MAR assumption violations will be assessed with state-of-art methods. Sex/Gender: To explore the effect of gender identity (women, man, gender-diverse) and sex (female, male, intersex) on outcome variables and intervention effect, all outcomes will be described by treatment group stratified by gender (KOOS4, Partner in Health Scale, cost-utility, physical activity, quality of life) or sex (MRI, knee extensor strength, adiposity) at all time points. We will estimate GLMMs and report intervention effect estimates stratified by sex (female, male) or gender (woman, man) as exploratory analyses to inform future studies. As 6% of our preliminary proof-of-concept randomized controlled trial participants identified gender diverse, descriptive statistics will also explore differences by cis and diverse gender. Implementation: The % of physiotherapists achieving ≥85% on the fidelity checklist and checklist items with ≤70% fidelity will be reported. Provider and participants' responses to survey questions asking about barriers and facilitators of SOAR implementation will be summarized. Interview recordings related to identifying features of participant-provider interactions that facilitate self-management will be transcribed and de-identified. Data will be coded using a constant comparative approach, and categories will be developed by comparing and identifying meaningful patterns across codes. High-order themes will elucidate the relationship between categories. We will look for uniqueness by gender and, if found, reanalyze the data with a gender lens. Analysis trustworthiness and credibility will be fostered through data immersion, memoing, reflexive journaling, and team discussions. An audit of analytic decisions will be kept. Efficiency outcome: Incremental cost-utility ratio will be estimated as (Cost\_SOAR - Cost\_CONTROL) ⁄(∆QALY\_SOAR - ∆QALY\_CONTROL) for the intervention and 6-month post-intervention period using nested imputation and nonparametric bootstrapping to model uncertainty around cost and QALY estimates. The contribution of each cost item to total healthcare resource use will be described by group, gender, and compliance (full, partial).