Physical Recovery During Bracing After Ankle Fracture (Re-BAF)

Background: Although patients generally acknowledge the need for immobilisation following ankle fracture, a qualitative study including ten patients found substantial challenges in adhering to immobilisation recommendations. In a randomised controlled trial (RCT) by Smeeing et al. (2020), braces and elastic bandages permitting greater ankle movement than standard walkers were not associated with poorer functional outcomes and were associated with a faster return to work (p = 0.02) without an increased risk of complications. However, the trial was terminated after inclusion of only half of the planned sample, raising concerns regarding statistical power and robustness of the findings. Compared to the walker, ankle stirrups may better align with patient preferences while promoting physical activity and faster recovery.

Aim: The aim of this Re-BAF trial is to investigate if ankle stirrups, allowing active weight-bearing, are superior to rigid foot-ankle braces on patients' daily physical activity levels and physical recovery.

Study design: Superiority trial including 280 adult patients (140 per arm) with surgically or non-surgically treated ankle fractures, nested within a larger Scandinavian randomized multicentre non-inferiority trial (n=1400). Participants will be randomized to four weeks of weight-bearing as tolerated in either an ankle stirrup or a foot-ankle brace.

Eligibility: Inclusion: patients aged ≥18 years with surgically or non-surgically treated ankle fractures. Exclusion: pathological fractures, open fractures, prolonged need for ankle immobilisation (e.g., non-union or poor wound healing), inability to adhere to trial procedures (e.g. neuropathy or severe psychiatric disorder), non-weight bearing treatment, allergy to adhesive tape used for the SENS apparatus, uninterest in participating in the trial, or inability to read Danish.

Recruitment: Patients will be recruited at their one- to two-week follow-up from orthopaedic departments across Denmark. Patient recruitment begins after one to two weeks, as surgically treated patients are typically immobilised in casts or foot-ankle braces and instructed to avoid or minimise weight-bearing during the initial postoperative period to allow swelling to subside and wounds to heal. Likewise, non-surgically treated patients are typically immobilised in a foot-ankle brace. However, weight-bearing recommendations vary between hospitals and surgeons, ranging from weight-bearing as tolerated to limited or non-weight-bearing until fracture stability is confirmed radiographically at the one- to two-week follow-up. The type of immobilisation (e.g., cast or brace) and weight-bearing restrictions during the initial weeks therefore vary across hospitals and treating surgeons.

Randomisation: Patients will be randomised to four weeks of weight-bearing as tolerated in either a foot-ankle brace (walker) or an ankle stirrup, using REDCap Randomize, allocated 1:1. Randomisation will be performed in blocks to ensure balanced allocation of surgically and non-surgically treated patients between groups, as well as equal distribution of intervention and control patients across recruitment centres. Weight-bearing will be unrestricted in both groups; thus, the only intervention difference will be the type of brace (ankle stirrup versus walker). Other variation will reflect usual clinical practice.

Methods: From baseline (randomisation), patients will wear a thigh-worn SENS Motion® accelerometer, a GDPR-compliant medical sensor that continuously records physical activity, movement intensity, and nocturnal activity. The device will be worn until the battery is depleted, which is expected to occur after approximately 20 weeks. Weekly surveys will capture pain, analgesic use, and fear of movement. Additionally, two questionnaires will be distributed electronically via e-Boks alongside the patient-reported outcome measures (PROMs) of the BAF trial to assess fear of movement and patient-reported physical activity.

Outcomes: The primary outcome is the between-group difference in total physical activity (accelerometer counts) from baseline (randomisation) to 12 weeks post-fracture, while secondary outcomes include between-group differences in daily step counts, nocturnal activity (sleep efficiency proxy), time to resume vigorous activities, pain, analgesic use, patient-reported physical activity, and fear of movement.

A detailed and pre-specified Statistical Analysis Plan (SAP) will be made publicly available prior to database lock and before any unblinding of the data.

Timepoints: Assessments will be performed at baseline (randomisation), 6 weeks, 12 weeks (primary endpoint). Accelerometer-based measurements and weekly surveys will continue until the battery of the SENS accelerometer is depleted (approximately 22 weeks post-fracture). Questionnaires will be collected at 6, 12, and 26 weeks post-fracture alongside the patient-reported outcome measures of the BAF trial.

Sample size calculation: As no anchor-based minimal clinically important difference (MCID) exists for physical activity in ankle fracture patients, an MCID of 17-35% in daily step counts, derived from research on chronic obstructive pulmonary disease, will be used. Applying a 17% threshold to total accelerometer counts per day, on a comparable patient group, results in an expected between-group difference of 50,500 activity counts/day. To detect statistically significant differences at a two-tailed α level of 0.05 and β = 0.80, assuming a standard deviation of 101,000 counts per day before and after the intervention, a sample of 63 participants from each subgroup (surgical/non-surgical) per intervention group (ankle stirrup vs. foot-ankle brace) is required. To account for potential dropouts, seven additional patients will be included per subgroup, resulting in a total sample size of 280 patients.

Blinding/masking: Participants and care providers cannot be blinded due to the nature of the brace interventions. Data are labelled using allocation codes only. Outcome assessments and statistical analyses will be performed blinded to treatment allocation. The primary investigator is not involved in the randomisation process and will remain blinded to allocation during data collection, management and analysis.