This prospective multisite randomized double crossover clinical trial study will follow an A1-B1-B2-A2 design that includes military veterans and civilians. Three collaborative sites will recruit and enroll participants: (1) University of Washington (Seattle, WA), (2) WillowWood (Mt. Sterling, OH), and (3) University of South Florida (Tampa, FL). Prior to data collection, the principal investigators and main study personnel from each site participated in a study protocol training session to ensure data collection at each site was the same. The study personnel meet every other week to discuss study procedures.

The study received approval (STUDY00017239) from the University of Washington’s institutional review board (IRB); the IRBs of the other sites and the US Army’s Office of Human Research Oversight reviewed the study protocol. The protocol is listed in the federal clinical trial registry (ClinicalTrials.gov NCT06214026).

All participants will provide informed consent. Study data will be confidential. There will be links between study data and participant identifiers that will be kept in separate secured areas. The link between participant identifiers and the research data will be destroyed after the records retention period required by state and/or federal law. All the information provided by the participants will be confidential.

After the completion of visits 2, 3, and 4, participants will be compensated US $200 per visit for a total of US $600. Parking, mileage, and transportation (within limits) will be paid for study visits.

The study includes three groups of individuals with amputations: (1) people with a unilateral above-knee amputation, (2) people with bilateral LEAs, and (3) people with a unilateral below-knee amputation who are categorized as having a K2 limited community ambulation level of mobility. All participants must be able to provide written informed consent independently and have the ability to read, write, and comprehend English.

Past research studies of people with amputations have historically underrepresented specific groups. Barriers to participation in clinical trials have been identified. This study targets individuals in these categories, and accommodations have been made. These included flexible payment schedules for participation, free rideshare transportation organized by study staff, free parking for participants who had their own transportation, and an on-call prosthetist. In addition, the study enrollment period was 3 years, enabling participants to delay protocol initiation due to socket changes, skin issues, or personal factors.

To be eligible for the study, participants must be 16 years or older and reside in an independent living environment within the community. They may use a walking aid but should not primarily rely on wheelchair mobility.

The weight limit is 165.6 kg (365 lbs). Participants must not have conditions, such as skin wounds, that would preclude the use of a prosthesis. Young children will not be included because they have different functional needs compared to adults. People will be excluded if they have fluctuating conditions that may significantly alter gait mechanics during the 10-week study. Examples include Parkinson disease, alcoholism, brain tumors, and hereditary cerebellar ataxias.

Participants will be discontinued from the study if they are unable to use a prosthesis. Examples include surgery or trauma. However, due to the potential for fluctuations in prosthesis use among this population, short-term issues will not disqualify individuals from continued participation.

The primary analysis will compare the META Arc foot in locked versus unlocked configurations. Statisticians will remain blinded to condition labels, analyzing outcomes as Condition X versus Y, with knowledge of the order (X→Y or Y→X), but not which corresponds to locked or unlocked.

Descriptive statistics (means and SDs for continuous variables; counts and percentages for categorical variables) will be used to summarize participant characteristics, stratified by amputation group. All analyses will be performed separately for each group.

Primary performance outcomes include the TUG, the 3XF8W, and the Four Square Step Test. Assumptions of normality and homogeneity will be tested. If met, a multivariate analysis of variance (MANOVA) will be used to evaluate the effect of foot configuration at α=.05. A permutational MANOVA will be used if assumptions are violated. If multivariate differences are significant, post hoc paired t tests (or Wilcoxon signed-rank tests if nonnormality is present) will be used to assess individual outcomes without correction for multiple comparisons. In order to help control the familywise type I error risk, pairwise comparisons of the primary outcomes will only occur if the multivariate difference comparison from the MANOVA is significant.

Satisfaction (PLUS-M) will be analyzed using paired t tests (or Wilcoxon tests) at α=.025 to maintain a familywise error rate of .05 across satisfaction outcomes.

With n=30 for the above-knee and K2 groups, the study is powered (>80%) to detect a standardized effect size of 0.53. In the bilateral group (n=16), the power is greater than 50% for the same effect size. Preliminary data show larger effects (eg, 0.98 for the 3XF8W), supporting adequate power. Although underpowered to detect a meaningful difference in the bilateral group, 16 subjects is in line with other gait studies that have published data on 10 [13], 15 [14], or 19 [15] subjects with bilateral amputations, a challenging population to recruit and enroll. The study is also powered (81%) to detect a 75% preference for one configuration (vs 50% under the null) using a sign test with n=30. Although no minimal clinically important difference has been established, it is predicted that a standardized effect of at least 0.50 will correspond to an impactful improvement in participants’ physical function.

Secondary outcomes (eg, 3XF8W step count, 10-meter walk, and PLUS-M subscales) will be analyzed using paired t tests or Wilcoxon tests at α=.05. Carryover effects will be assessed and, if present, accounted for in sensitivity analyses. As no multiple comparison corrections will be used for secondary analyses, resulting in a higher familywise type I error rate, all secondary results will be considered exploratory and hypothesis generating rather than confirmatory.

Logbook variables (eg, perceived support on challenging surfaces) will be compared using paired t tests or Wilcoxon tests. Fall counts will be analyzed with a sign test. Gait variables (eg, step width, double support time) from the Zeno walkway will be analyzed descriptively.

Interviews will be transcribed via Microsoft Teams or Zoom (Zoom Communications Inc) and analyzed using an inductive thematic approach. Open coding will identify themes, followed by focused coding to group codes into broader categories. Multiple codes may apply to individual responses. Qualitative analysis will follow current standards for rigor and reporting.

It is expected that the side-to-side adaptability of the META Arc prosthetic foot will help people with above-knee amputations, people with bilateral amputations, and people with lower levels of mobility manage their everyday activities better. It is expected that the results of this study will demonstrate that people with lower levels of mobility can complete non–straight gait tasks such as turning better and will participate in more challenging activities every day.

It is hypothesized that participants with bilateral amputations will report greater satisfaction, mobility, and comfort when using a prosthetic foot with frontal plane adaptability (unlocked condition) than when wearing a prosthetic foot without frontal plane adaptability (locked), as measured by the PLUS-M. It is also expected that participants will demonstrate improved performance during the 3XF8W using a prosthetic foot with frontal plane adaptability compared to a foot without frontal plane adaptability.

A cutoff for the TUG in determining fall risk of ≥8.17 seconds has been previously determined by Sawers and Hafner [16]. However, it is not clear what part of the TUG is causing the increased risk. By completing component analysis, this study will determine what part of the TUG is causing times above this fall risk threshold.

A previous study of the META Arc foot showed that optimal stiffness for walking on flat surfaces is not the same as for turning [17]. Preliminary findings have shown that there are differences in stride length and swing time when comparing turning with the prosthetic foot on the inside or the outside of the turn. This may have some implications in the development of gait training and fall prevention strategies.

During logbook and interview analysis, it is expected that participants will report their perceived changes in person-specific mobility during activities that require frontal plane adaptability. The logbook and interviews will also provide feedback on the investigational foot regarding what aspects the participants did or did not like. This information will be very useful for making improvements for the next version if needed. Preliminary logbook analysis has shown that participants have engaged in walking on uneven terrain and participated in activities such as walking the dog, hiking, and shopping with benefits while wearing the investigational foot.

The study populations are particularly relevant to veterans and service members with amputations resulting from combat-related trauma or disease. Prosthesis users should be challenged to move sideways, move backward, and turn. Biomechanical components are often conflated with typical summary scores, which are typically based on the time it takes to complete a task. By conducting component analysis, mobility challenges, prosthetic design, and training can be addressed to better fit the needs of the users.

There are some limitations to the protocol. The participants have a 2-week accommodation period for each phase of the A1-B1-B2-A2 crossover design. There is no structured training program. Although the order of locked and unlocked use of the study foot will be varied between individuals, participants may continue to become more accustomed to using the foot into the next period, with the possibility of bias in the results.

The study will use a design that allows participants to get used to using two forms of the investigational foot: the locked and unlocked versions. This unique study design feature allows for comparing the investigational foot to a standardized control foot in a randomized order. Accommodation for participants in understudied groups may facilitate recruitment. Findings from this study are expected to generate new knowledge on the functional benefits of frontal plane adaptability, informing prosthetic design and clinical practice.