Determining the Potential Benefit of Powered Prostheses: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Determining the Potential Benefit of Powered Prostheses (NCT02828982), a randomized na study. The primary outcome is Metabolic Cost of Walking.

2. Clinical Question

The registry describes a study of whether a powered ankle prosthesis differs from an unpowered prosthesis on measures related to walking. Its stated aim was to assess performance during short bouts of walking, extended walking to fatigue, and activity in the community. The primary question concerned metabolic cost during treadmill walking at a fixed speed after participants had worn each prosthesis for two weeks.

The population was people with unilateral traumatic amputation. The registry reports an enrollment of 12 and says each participant wore both study prostheses. The primary endpoint description notes that one participant was excluded from analysis because he could not reach steady-state energetics on a treadmill. The analyzed sample counts posted for the primary endpoint are 9 for each prosthesis condition; the reason for the difference between enrollment, the stated exclusion, and those counts is not reported in the registry record.

The intervention was the BiOM powered ankle prosthesis, described in the registry as a powered prosthetic ankle. The comparator was an unpowered prosthesis, identified in the intervention listing as a Dynamic Response Foot and in the results as the Unpowered Prosthesis. This within-participant comparison means participants experienced both conditions, although the registry record does not report the order or sequence used for each participant.

The posted primary result is a two-sided t-test with p = 0.585. The registry reports means and standard deviations for the two conditions but does not post a confidence interval for the primary comparison. The result therefore describes the observed data and the sponsor's reported test; it does not establish that the conditions are equivalent or that any particular difference is ruled out.

3. Trial Design

This was a randomized, unmasked, crossover study. The registry lists the trial as completed, with enrollment of 12. Its listed condition was unilateral traumatic amputation. In a crossover design, each participant receives or uses both conditions, so comparisons can be made within the same participant rather than relying only on comparisons between separate groups. The registry confirms that each participant wore both study prostheses, but the sequence, randomization procedure, and any period or carryover considerations are not reported in the registry record.

The two conditions were a powered ankle prosthesis and a dynamic response foot, described in the results as the BiOM Powered Prosthesis and the Unpowered Prosthesis. The primary metabolic-cost assessment took place after wearing the prostheses for two weeks. The registry also lists secondary measures of the Prosthetic Evaluation Questionnaire, lower-limb muscle activity, and time to fatigue. Time to fatigue was measured after the two-week wearing period as well.

The registry lists a primary outcome of metabolic cost of walking, measured with a portable metabolic system while participants walked on a treadmill at a fixed speed. It reports means and standard deviations for 9 participants in each condition. The endpoint description states that one participant was excluded because steady-state energetics could not be achieved. It does not explain all the differences between the enrolled count, this exclusion statement, and the posted endpoint counts.

The masking field is listed as none. The record does not report the sample-size justification, the planned statistical model for the crossover comparison, the intended analysis population, or whether period and sequence were incorporated into the analysis. The registry reports two-sided t-tests for the primary and time-to-fatigue comparisons, but does not give further detail about how the crossover structure was handled.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Participants were asked to complete both arms of the study. Before starting, participants were randomly enrolled in the dynamic response foot (unpowered prosthesis) or powered condition to start with. Some participants completed both arms while others only finished one of the arms.

First Intervention (3 Weeks)

Reasons for not completing this period:

Second Intervention (3 Weeks)

Analysis population for the primary outcome, as defined in the registry: Each participant wore both study prostheses. One participant was excluded from analysis as he was not able to reach steady-state energetics on a treadmill

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Metabolic Cost of Walking, measured over Testing took place after wearing the prostheses for two weeks.

The primary endpoint was metabolic cost of walking, reported in J/Nm. The registry describes this as a mean continuous measure. Metabolic cost was measured with a portable metabolic system while participants walked on a treadmill at a fixed speed, after wearing each prosthesis for two weeks.

Metabolic cost is used as a quantitative measure of the energetic demands associated with walking under the specified test conditions. A treadmill protocol at a fixed speed gives a defined setting for comparing measurements across prosthesis conditions. Because the study used a crossover design, the intended comparison could also draw on measurements from the same participant under both conditions. However, the registry does not report whether the sponsor's statistical analysis used a paired approach or otherwise accounted for the within-participant structure.

The posted results are means and standard deviations for the two conditions. The registry reports a two-sided t-test, with p = 0.585, but does not provide a primary comparison estimate or confidence interval. Without an interval estimate for this endpoint, the precision and range of differences compatible with the data cannot be assessed from the posted primary comparison. The p-value alone does not quantify the size of a difference, and it does not demonstrate equivalence.

The endpoint description says one participant could not reach steady-state energetics on a treadmill and was excluded from analysis. The registry does not report how that participant's measurements were handled beyond the stated exclusion, nor does it explain the posted counts of 9 in each condition.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Continuous outcomes are commonly described by a mean and a measure of spread. The mean gives the arithmetic average of the observed values, while the standard deviation describes how dispersed participant measurements are around that mean. The registry reports means and standard deviations for metabolic cost and time to fatigue. These summaries are useful for describing each condition, but by themselves they do not fully describe the uncertainty in a comparison between conditions.

What does a p-value say?

The registry reports two-sided t-tests for the primary endpoint and time to fatigue. A t-test evaluates how compatible the observed data are with a null hypothesis of no difference under the assumptions of the test. A two-sided test allows evidence in either direction. The p-value is not a measure of effect size, and it is not the probability that the null hypothesis is true. A p-value that is not small does not establish equivalence or rule out potentially important differences. Interpretation also depends on the analysis design, assumptions, sample size, and the estimate's uncertainty.

For the primary outcome, the posted sponsor test has p = 0.585. For time to fatigue, the sponsor reports p = 0.212. The registry does not report the test statistic, degrees of freedom, the exact form of the t-test, or a confidence interval for either sponsor comparison. It also does not report a primary treatment-effect estimate. As a result, the reported p-values cannot be used alone to assess the range of differences compatible with the data.

Why does crossover structure matter?

In a crossover trial, the same participant experiences both conditions. The measurements under the two conditions can be correlated because they come from the same person. A paired analysis uses the within-person differences and can account for this correlation. An analysis that treats the condition groups as independent does not use the pairing in the same way. The registry confirms that each participant wore both prostheses, but it does not specify whether the sponsor's t-test was paired, how the randomization sequence was handled, or whether period effects or carryover were assessed. These details are not reported in the registry record, so the analysis cannot be reconstructed fully from the posted summary.

The primary endpoint's two condition counts are both 9. Equal counts do not reveal whether the same 9 participants contributed measurements to both conditions, nor do they establish which statistical method was used. The registry says one participant was excluded because steady-state energetics could not be reached, but the posted primary counts and the full participant flow are not reconciled in the record.

How should the re-calculation be read?

The unadjusted re-calculation for time to fatigue uses the posted group means and standard deviations. It reports a mean difference of -1.5 minutes, an interval from -11.55 to 8.552, and an unadjusted p-value of 0.77. This is a simple summary-statistic comparison and does not incorporate the within-participant pairing. Its interval is an interval estimate, but the registry does not state a confidence level for it. The sponsor's own reported test is p = 0.212; the re-calculation is not a substitute for an analysis based on the participant-level crossover data.

When an interval estimate includes the null value of no difference, the data are compatible with effects on either side of that value under the interval's assumptions. The width of an interval conveys precision: wider intervals allow a broader range of possible effects. This general interpretation should not be mistaken for proof that every value in an interval is equally plausible, or that values outside it are impossible. For this trial, the crossover structure and the interval's unspecified confidence level limit what can be inferred from the unadjusted time-to-fatigue calculation.

What does over-dispersion mean?

Over-dispersion is a general issue that arises when observed variability is greater than a statistical model expects. It can make standard errors and p-values unreliable if the analysis does not accommodate the extra variability. The posted continuous-outcome summaries use standard deviations, and the record does not report an over-dispersion assessment or a model that would allow one to evaluate it. This concept is therefore relevant as a general statistical consideration, not as a finding about this trial.

What do the planning illustrations mean?

The time-to-fatigue planning illustrations use the observed effect as a starting point and show participant counts per arm for 80% and 90% power at different fractions of that effect. The registry states that these are illustrations, not the trial's own design or sample-size justification. They assume a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Their purpose is to show how planning counts depend on the effect size targeted. They are not observed power, do not describe the power of this completed trial, and do not establish that a future study would achieve its target under different design assumptions.

The registry does not report the original sample-size rationale, the planned primary analysis, a multiplicity strategy, or a complete missing-data plan. Those omissions limit evaluation of whether the analyses followed a prespecified plan and how uncertainty across endpoints was handled.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Unpowered Prosthesis", "BiOM Powered Prosthesis")
n   <- c(9, 9)
m   <- c(0.409, 0.397)
sp  <- c(0.04, 0.078)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Unpowered Prosthesis  9  0.409  0.040  0.378  0.440
CARM  BiOM Powered Prosthesis  9  0.397  0.078  0.337  0.457

11. Analyses Reported by the Sponsor

For the primary metabolic-cost outcome, the registry reports group means and standard deviations and the sponsor's two-sided t-test, p = 0.585. It does not post a primary effect estimate or confidence interval. No re-calculated primary comparison is available in the registry record, so there is no independent summary-statistic calculation to set alongside the sponsor's primary test.

For time to fatigue, the sponsor's reported two-sided t-test has p = 0.212. A simple unadjusted re-calculation from the posted means and standard deviations gives an unadjusted p-value of 0.77 and an interval estimate from -11.55 to 8.552 for the mean difference. This calculation treats the posted summaries as a two-group comparison and does not use the crossover pairing. It is therefore not a reconstruction of the sponsor's analysis and should not displace the sponsor's reported result.

The re-calculated result is useful as a transparent check based on the summary values, while the sponsor's analysis is the trial's own posted analysis. The registry does not provide enough methodological detail or participant-level data to determine why the results differ or whether the sponsor's test accounts for the crossover structure. The discrepancy flag is false, but the methods required to reproduce the sponsor's test are still not reported in the registry record.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists multiple outcome measures and analyses, including the primary metabolic-cost endpoint and secondary measures of prosthesis evaluation, muscle activity, and time to fatigue. It does not report whether a single primary hypothesis was formally designated for testing within a prespecified hierarchy, whether secondary tests were exploratory, or whether any multiplicity adjustment was planned or used. Without this information, the p-values should not be interpreted as if they necessarily came from one isolated test. When many outcomes are examined, the chance of at least one apparently unusual result can increase; this is a general statistical concern, not a reported finding about the trial.

Missing-data and analysis-population details are also limited. The primary endpoint description says one participant was excluded because he could not reach steady-state energetics on a treadmill. The record does not report whether this exclusion was specified in advance, whether any other participants lacked primary measurements, or how missing observations were handled in the analyses. Although enrollment was 12, the primary outcome counts are 9 in each prosthesis condition. The registry does not explain the full participant flow or reconcile those counts with the stated exclusion.

The registry confirms that each participant wore both prostheses, but it does not report whether all randomized participants completed both periods, whether participants were analyzed according to their assigned sequence, or what analysis population was used. It also does not state whether the primary analysis included only participants with complete paired observations. These omissions matter because missing one condition's measurement can affect a crossover comparison differently from missing an observation in a parallel-group trial.

The record does not report a missing-data strategy, sensitivity analyses, a multiplicity plan, or the prespecified statistical analysis plan. These details are not reported in the registry record, so it is not possible to assess the impact of missing observations or multiple testing beyond noting the stated exclusion and the posted results.

13. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The adverse-event record reports zero serious adverse events and zero other adverse events in each of the two prosthesis groups. It lists 12 participants at risk in each group and zero deaths in each group. The registry does not list serious adverse-event terms.

These figures describe the events recorded in this study and the denominators shown in the registry. They do not establish that no events could occur, and they do not by themselves provide a precise estimate of event risk. The study enrolled 12 participants, and the registry describes a crossover in which each participant wore both prostheses; it does not explain how the adverse-event denominators were constructed across the two conditions or how event attribution was handled.

The absence of recorded events in this small study is therefore a limited observation about the trial record, not a basis for broad comparisons or conclusions about safety. The registry does not provide additional event detail, definitions, or follow-up information in the posted adverse-event figures. These limitations should be kept in view when interpreting the zero counts.

14. What the Result Does, and Does Not, Mean

The results describe measurements collected under the study's specific conditions: participants with unilateral traumatic amputation wore both prostheses, and metabolic cost was assessed on a treadmill at a fixed speed after two weeks of wearing each prosthesis. The registry reports condition-specific means and standard deviations and a two-sided sponsor test for the primary outcome. It also reports a time-to-fatigue comparison and limited adverse-event figures.

The primary p-value does not show that the two prostheses are equivalent, and it does not rule out a difference. The record lacks a primary effect estimate and interval estimate, which makes it difficult to judge the precision of the comparison. The time-to-fatigue summaries and sponsor test provide another outcome-specific result, but the analysis details are limited and the simple re-calculation does not account for paired observations.

The findings should not be extended automatically beyond the participants, testing conditions, and outcomes represented in the registry. The registry's stated aims include short bouts of walking, extended walking to fatigue, and community performance, but the posted information does not give a complete account of results for every aim. In particular, the record does not provide full numerical results for all listed secondary measures. The two-week wearing period and fixed-speed treadmill assessment also define the context of the primary endpoint; the registry does not establish that these measurements capture all aspects of day-to-day use.

The record does not report the planned sample-size rationale, full statistical plan, multiplicity approach, or detailed missing-data procedures. These gaps, together with the small enrollment and the stated primary exclusion, limit how confidently the results can be generalized or how precisely differences can be assessed. The reported statistics are evidence about this trial's observed data, not a standalone determination of equivalence, effectiveness, or performance in other settings.

15. Why This Trial Matters Statistically

This trial illustrates why design and analysis details are especially important in crossover studies. Each participant wore both prostheses, so the measurements are linked within individuals. The statistical method should reflect that structure when estimating and testing a condition difference. The registry identifies the analysis as a two-sided t-test but does not say whether it used paired observations or how sequence and period were handled. That uncertainty affects how readily readers can interpret or reproduce the reported comparison.

It also illustrates the difference between a p-value and an estimate of precision. The primary result includes a p-value but no posted treatment-effect estimate or confidence interval. Without an interval, readers cannot directly assess the range of primary differences that may be compatible with the data. The group means and standard deviations describe each condition, but they do not replace an interval for the within-participant contrast.

The time-to-fatigue result provides a further lesson: a simple re-calculation based on group summaries can differ from the sponsor's reported test, particularly when the summary calculation does not use the crossover pairing. Such a calculation is a limited check, not a replacement for an analysis of participant-level data that reflects the design. Planning illustrations based on an observed effect are also distinct from the trial's original sample-size rationale and are not observed power.

Finally, the record highlights the importance of reporting participant flow, missing-data handling, multiplicity decisions, and prespecified analysis methods. These details allow clinical researchers and biostatisticians to evaluate not only the numerical findings but also the path from design to analysis. Their absence does not invalidate the posted results, but it leaves important aspects of their interpretation unresolved.

16. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several limitations are specific to interpreting this trial's registry record. Enrollment was 12, and the primary outcome counts are 9 in each condition. The endpoint description identifies one exclusion due to inability to reach steady-state energetics, but the record does not reconcile the enrollment, exclusion, and posted counts. The full participant flow and the handling of any other unavailable measurements are not reported in the registry record.

The crossover design means the two condition measurements are linked within participants. The registry confirms that each participant wore both prostheses but does not report the sequence, period structure, or whether the sponsor's t-test accounted for paired data. It also does not describe assessment of carryover or period effects. These omissions make it difficult to judge whether the reported test matches the design and to reproduce the comparison.

The primary comparison is reported without an effect estimate or interval estimate. This prevents direct assessment of the primary estimate's precision from the registry results. For time to fatigue, a summary-statistic re-calculation is available, but it is unadjusted for within-participant pairing and its interval's confidence level is not stated in the registry record. The sponsor's own test is reported separately.

The record does not report the sample-size justification, a detailed statistical analysis plan, the analysis population, a multiplicity strategy, or a missing-data plan. Although the stated study aims include several types of walking performance, complete numerical results are not posted for every listed secondary measure. The results therefore provide a restricted account of the trial and should be interpreted within those reporting limits.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Determining the Potential Benefit of Powered Prostheses — NCT02828982.
• Registry record last updated 2022-01-18; status last verified 2022-01.
• Results first posted 2020-11-13.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

20. Record Summary

The registry record describes a completed randomized, unmasked crossover study with enrollment of 12 participants with unilateral traumatic amputation. Each participant wore a powered ankle prosthesis and an unpowered prosthesis. The primary outcome was metabolic cost of walking on a treadmill at a fixed speed after two weeks of wearing each prosthesis.

The primary outcome summaries are reported for 9 participants in each condition, and the sponsor reports a two-sided t-test with p = 0.585. The record notes that one participant was excluded because steady-state energetics could not be reached, but does not reconcile this with the posted counts. Time to fatigue is a listed secondary measure; its sponsor test has p = 0.212. The registry also lists prosthesis evaluation and lower-limb muscle activity.

The record does not report the full statistical plan, sample-size rationale, multiplicity strategy, or detailed missing-data procedures. It reports zero serious adverse events, zero other adverse events, and zero deaths in each listed group.