COMplex Fracture Orthopedic Rehabilitation (COMFORT) - Real-time visual biofeedback on weight bearing versus standard training methods in the treatment of proximal femur fractures in the elderly: study protocol for a multicenter randomized controlled trial

This protocol is described according to the SPIRIT (Standard Protocol Items: Recommendations for Interventional Trials) Statement (see Additional file 1). The study is an international, multicenter trial conducted between March 2017 and August 2018 in the Netherlands and Sweden. Rehabilitation centers focused on geriatric rehabilitation are asked to participate in this study. In order to ensure the quality and expertise of the participating centers, a minimum inclusion rate of at least 20 patients per year is expected. Therefore, the number of eligible patients is screened in each center before their participation in the study. Measurements will be performed by trained physical therapists at the site and all physical therapists are experienced in geriatric trauma care. Before the start of the study, all physical therapists agreed on the study treatment regime during multiple consensus meetings and were trained to uniformly implement this study treatment. Participating geriatric rehabilitation clinics are listed in Table 1.

This trial is a two-arm, parallel-design, prospective RCT. Based on a previous pilot study, the average length of rehabilitation is 44 days, with a standard deviation of 17. Using a power of 0.80, an alpha of 5%, and an expected reduction of seven days (one week) in two-sided testing, a sample size of 94 participants per group was calculated. Participants will be allocated to either the control group or the intervention group using a stratified cluster randomization according to the sealed opaque envelope principle. First, all rehabilitation clinics were divided into two groups depending on the inclusion rate: < 2 vs. ≥ 2 eligible participants per month. This division was made to avoid skewing of inclusion rate in both arms. Then in both groups, clinics were randomized by the principal investigator into the control or intervention group using the sealed opaque envelope principle. This resulted in six rehabilitation clinics in the control group and five rehabilitation clinics in the intervention group (Table 1). Randomization per center, or cluster randomization, was chosen for two reasons. First, individual selection bias by the physical therapist is minimized. Second, and more importantly, the physical therapists and participants within one center can now provide a single treatment, either care as usual or care as usual, with real-time visual biofeedback. Considering the constant interaction between physical therapists, as well as between participants, individual randomization would increase the risk of data collection errors.

Strict inclusion criteria will be used to introduce homogeneity of the groups; therefore, participants enrolled in the study meet the following inclusion criteria:

Participants with the following criteria will be excluded from the study:

Data will be gathered according to a strict study protocol (Fig. 1). Candidate patients who fulfill the inclusion criteria will be asked to participate in the study by the site investigator and receive an information letter with the study protocol and study objectives. If patients agree to participate in the study, informed consent will be signed and the cognitive score will be evaluated by the MMSE. The MMSE is an objective score to quantitatively asses the severity of cognitive impairment. A MMSE score < 18 indicates severe cognitive impairment [16] and this leads to exclusion of the patient in this trial. After informed consent and MMSE, data will be collected from participants according to the following trial protocol. The institutional protocol for physical therapy after proximal femur fractures will be followed in all participants. In addition, each participant receives daily force measurements using the SensiStep system in a 30-m walk to measure the gait parameters. Additional clinical tests will be executed to gain additional insight in the rehabilitation progress of each participant:

Finally, additional information of each participant will be documented, including the following:

Participants in the intervention group receive, in addition to the standard institutional protocol, real-time visual feedback about weight-bearing during the 30-m walk with SensiStep. The physical therapist has real-time visual insight in the weight-bearing as well by observing the tablet together with the participant. Both the biofeedback system (visually) and the physical therapist (verbally) assist and motivate the participant to adapt the weight-bearing to the optimal level. The target weight will be set at 100% bodyweight, as the participants have no restrictions in weight-bearing.

The primary outcome measures are the gait parameters maximum peak load (in %bodyweight) and step duration (in seconds), which will be analyzed in relation to the discharge date from the rehabilitation center. It is expected that the intervention group will show an improvement in gait parameters and earlier discharge date compared to the control group. The discharge date will be determined by a multidisciplinary team within the usual institutional regime, irrespective of the study protocol. Secondary outcomes include other spatiotemporal gait parameters and validated clinical tests, including the EMS, the FAC, and the VAS. The EMS is a standardized validated scale for assessment of frail elderly people. The FAC is used as a diagnostic and evaluative tool to assess the physical performance of patients. The VAS is a commonly used, subjective scale to assess the amount of pain patients experience. Spatiotemporal gait parameters measured by the SensiStep will be correlated to the clinical test scores (e.g. EMS, FAC, and VAS). Blinding the participants and/or physical therapist is not possible, as both users are aware of the biofeedback they receive in the intervention group. However, data will be encrypted and anonymously delivered by the Steering Committee to the data scientists, which makes the data scientists blinded for the analysis of gathered data.

Statistical analysis on outcome measures will be conducted using Statistical Package for the Social Sciences (SPSS). First, data distribution will be assessed. Second, missing data will be handled by a multiple imputation technique. The primary analysis will be based on an intention-to-treat principle. Also, a per-protocol analysis will be performed as sensitivity analysis to assess the robustness of the results to protocol deviations [17]. The primary goal is to find a relation between the gait parameters (maximum peak load and step duration) and the discharge date. In addition, other spatiotemporal gait parameters measured by the SensiStep will be analyzed for this relation as well. The secondary goal is to find a relation between the gait parameters and the clinical measurements (i.e. EMS, FAC, and VAS). Relations between all data will be examined using multiple regression analysis.

This trial will be conducted according to the principles of the Declaration of Helsinki (amended version by the 64th WMA General Assembly, Fortaleza, Brazil, October 2013) and in accordance with the Medical Research Involving Human Subjects Act. Gathered data will be encrypted and stored on a secure server and is accessible by the Steering Committee.

This trial will be conducted under strict supervision of an experienced Steering Committee from the University Medical Center Utrecht (UMCU), Maastricht University Medical Center+ (MUMC+), and Uppsala University (UU). Acting as a Steering Committee, the principal investigators from UMCU and MUMC+ will discuss the progress of enrollment at least once every two weeks. The Steering Committee will perform a blinded interim analysis on 30%, 60%, and 90% of the included participants. During these interim analysis, reasons for exclusion will be discussed with the site investigators. Other items will be discussed as well, including but not limited to, applied discharge criteria for each patient, adverse events, quality of the retrieved data, missing data, and necessity to include other participating centers.

A monthly newsletter will be written by the Steering Committee and distributed among the participating centers. In this newsletter, the inclusion rate will be shown per center, which is a common and effective way to increase awareness and readiness to include patients in multicenter trials. The trial progress and general results will be communicated by the Steering Committee upon request.