Effects of Post-Stroke Upper Extremity Assistance: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effects of Post-Stroke Upper Extremity Assistance (NCT05036642), a na na study. The primary outcome is Wolf Motor Function Test - Change in Functional Score by Task (WMFT).

2. Clinical Question

The study’s stated aim is to quantify the short-term change in arm movement ability during and after robot-assisted therapy. Its background distinguishes the effects of therapy over weeks from effects over minutes or hours. The primary outcome therefore concerns task performance across a brief device-use interval, rather than a long-term rehabilitation outcome. The registry also lists a Participant Satisfaction Survey as a secondary outcome, but its results are not reported in the registry record.

3. Trial Design

This is a completed, single-group device study sponsored by Stanford University. The registry record identifies the intervention as Compliant Passive Arm Support and describes the participant group as stroke survivors with upper-extremity motor impairments. The study has Pilot and Evaluation phases. The primary outcome applies to the Evaluation phase only; the record does not state that the Pilot phase contributed to that outcome.

The registry lists an enrollment of 12. Allocation is recorded as NA, masking as none, and the model as single-group. The record explicitly cautions that this is not a randomized comparison and that the arms are not a like-for-like control comparison. Consequently, the baseline assessment without the device and the later assessment with the device should not be interpreted as if they were measurements from randomized treatment and control groups. They are assessments at different times in a single-group setting.

The study start date is 2022-07-26, the primary completion date is 2022-09-07, and the status is completed. The registry describes the primary assessment as occurring at baseline without the device and after approximately 30 minutes of use with the device. It does not report additional visit schedules or details about how the assessments were administered.

The primary endpoint is the Wolf Motor Function Test change in functional score by task. The record describes 15 tasks assessed for this outcome. Weight to box is reported in a separate analysis, and Grip strength was excluded because the necessary equipment could not be acquired. The WMFT consists of 17 tasks in the registry description, but the primary endpoint does not include all of them. The registry lists eight outcome measures and three analyses; it does not report a full statistical analysis plan or explain how those measures and analyses relate to one another.

The registry record does not report a randomized comparator, a sample-size justification, or a statistical plan for the trial’s own design. Planning illustrations based on an observed effect use assumptions of equal allocation between two arms, a normal approximation, and a two-sided significance level of 0.05. Those assumptions do not describe the reported single-group study design and should not be read as evidence of the trial’s planned power or as observed power.

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.

Pilot Phase

Evaluation Phase

Analysis population for the primary outcome, as defined in the registry: Data for this outcome were collected from participants in the Evaluation phase only.

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 Wolf Motor Function Test - Change in Functional Score by Task (WMFT), measured over Assessed at baseline (without the device) and after approximately 30 minutes of use with the device..

The primary endpoint is the Wolf Motor Function Test change in functional score by task. It describes the number of assessed tasks for which participants, on average, had increased functionality, no change, or decreased functionality between baseline without the device and assessment after approximately 30 minutes of use with the device. The registry states that lower functional scores indicate lower functional levels.

This kind of endpoint represents a task-based summary: it classifies the direction of average change for each task rather than reporting the size of a change in a single continuous scale. The number of tasks in each direction can indicate whether the observed pattern across the assessed tasks tended toward improvement, stability, or decline. It does not, by itself, show the magnitude of change for a task, whether individual participants changed in the same way, or how uncertain the classifications are.

The WMFT comprises 17 tasks in the registry description. Fifteen were assessed for this outcome. Weight to box was reported in a separate analysis, while Grip strength was excluded because the necessary equipment could not be acquired. These distinctions matter when interpreting the task count: it applies to the specified outcome analysis, not to every WMFT task.

The endpoint is relevant to the stated question because it focuses on functional task performance over the short interval surrounding device use. However, the registry record does not report the observed task counts, estimates of uncertainty, or the analysis details behind the categories. The endpoint’s description can therefore be interpreted, but its numerical result cannot be evaluated from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record does not report the statistical methods used for the primary endpoint. The following explanations are general, and do not claim that a particular method was used in this study.

What does a task-count endpoint represent?

In general, a count endpoint records how many units meet a stated condition. In this study, the units are WMFT tasks, and the conditions are average increased functionality, no change, or decreased functionality. Such a summary can make the direction of change across tasks easier to describe, but it compresses information. It does not preserve the size of each task’s change or the distribution of changes among participants. Two patterns can have the same task count while differing in the magnitudes and participant-level consistency behind that count.

The categories also depend on how change was determined. The registry gives the broad classification and explains that lower functional scores indicate lower functional levels, but it does not report the threshold, calculation, or rules used to assign task-level change to the three categories. The exact statistical procedure behind the classifications is therefore not reported in the registry record. A reader should not infer that a particular test or threshold was used.

What does a baseline-to-follow-up comparison show?

In general, repeated measurement follows a group over time, allowing the later assessment to be considered in relation to baseline. When the same participants contribute measurements at both times, their observations are related rather than independent. Analyses of repeated observations often account for that relationship. The registry does not state which participants had paired measurements, how incomplete pairs were handled, or whether the analysis modeled within-participant dependence.

A before-and-after comparison can describe change over the observed interval. It does not automatically identify the cause of that change. Without a concurrent randomized control group, influences such as practice with the assessment, natural fluctuation, or other changes over time may be difficult to separate from the intervention. These are general interpretive considerations, not claims that any particular influence occurred in this study.

Why does the scale of the endpoint matter?

In general, a count of tasks has a bounded range determined by the number of tasks assessed. Count outcomes may have distributions that differ from the bell-shaped pattern assumed by some common methods, particularly when values cluster near a boundary or vary more than a simple model expects. Over-dispersion means that observed variability exceeds what a model assumes; if present and not addressed, uncertainty estimates can be too small. Whether over-dispersion or another distributional issue arose in this study is not reported in the registry record.

Methods for count outcomes can use different assumptions and can target different quantities. Some estimate an average count, while others compare rates or model the probability of a count. The appropriate approach depends on the data structure, the estimand, and the analysis plan. No model, test, variance estimate, or effect measure for the primary endpoint is reported in the registry record, so the method cannot be identified from the endpoint description alone.

How should uncertainty be expressed?

In general, an estimate describes the data’s central result, while a confidence interval communicates the uncertainty around that estimate under the assumptions of the analysis. A narrow interval indicates greater precision than a wide interval, but neither interval width nor a statistical significance threshold measures clinical importance by itself. The registry record does not report confidence intervals or other uncertainty measures for the primary outcome. Precision therefore cannot be judged from the outcome description.

A p-value, when used, addresses how compatible the observed data are with a specified null model under the assumptions of a test. It does not give the probability that a treatment works, the size of an effect, or the probability that the null hypothesis is true. No primary-outcome p-value or test is reported in the registry record. The record’s planning assumptions include a two-sided significance level of 0.05 for illustrative calculations, but that assumption is not evidence that the trial’s own primary analysis used that threshold.

How do the planning assumptions relate to this design?

General sample-size illustrations can show how a hypothetical study might be planned under specified assumptions. The record describes illustrations based on an observed effect, equal allocation between two arms, a normal approximation, and a two-sided significance level of 0.05. They are planning illustrations, not observed power and not a reconstruction of the study’s design. The reported trial is single-group, and its allocation is recorded as NA. The two-arm assumptions therefore do not describe the design used for this endpoint. The registry record does not report a sample-size justification for the trial itself.

General power calculations depend on assumptions about the target effect, outcome variability, allocation, significance level, and analysis method. An illustration based on the observed effect can be especially sensitive to uncertainty in that observed effect. It should not be interpreted as proof that a study was adequately powered or as a measure of the probability that the results are true. No planning calculation can replace an account of the prespecified analysis and the actual observed estimates with their uncertainty.

What cannot be recovered from the record?

The registry lists a count-based primary endpoint, the assessment times, and the Evaluation-phase population. It does not report the numerical endpoint results, task-specific estimates, participant-level data, missing-data approach, model or test, multiplicity strategy, confidence intervals, or a prespecified statistical plan. Those omissions prevent a methods-focused reader from checking assumptions, reproducing estimates, or judging precision. They also mean that no statistical method should be attributed to the study beyond the outcome and design descriptions stated in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists eight outcome measures and three analyses. The record does not explain which outcomes or analyses were prespecified as primary beyond identifying the WMFT change in functional score by task as the primary outcome. It does not report whether secondary outcomes were considered confirmatory or exploratory, whether hypotheses were tested across multiple tasks, or whether any adjustment for multiple comparisons was planned.

In general, evaluating multiple outcomes or analyses can increase the chance of finding an apparently notable result by chance if each is interpreted as a separate confirmatory test. Adjustment procedures can address some forms of this issue, but they depend on which comparisons were planned and how they relate. The registry record does not provide enough detail to determine whether multiplicity was a concern for the reported analyses or how it was handled. No multiplicity correction should be assumed.

The primary endpoint is stated to use data from the Evaluation phase only. The registry does not report the number of Evaluation-phase participants with complete primary-outcome data, how missing task assessments were handled, or whether participants who did not complete both assessments were included in any analysis. A method such as complete-case analysis, imputation, or a model using partial observations is not specified. Missingness and its potential effect on the endpoint therefore cannot be assessed from the record.

The registry also does not define the analysis population beyond noting that the primary outcome data were collected from participants in the Evaluation phase. It does not report whether all such participants were analyzed, whether any were excluded, or how protocol deviations were handled. The absence of these details limits assessment of both representativeness and potential bias. No particular missing-data method, analysis set, or multiplicity strategy is reported in the registry record.

11. 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 registry lists adverse-event figures separately for the Pilot Phase and the Evaluation Phase. For the Pilot Phase, the number at risk for serious events and other events is 1, and the recorded counts affected are 0 for serious events and 0 for other events. For the Evaluation Phase, the number at risk for each category is 11, with 0 recorded as affected by serious events and 0 by other events. Deaths are recorded as 0 in both phases.

These entries describe what the registry reports for the listed groups; they do not establish that no events of any kind occurred outside the categories or reporting period represented in the record. The registry does not provide event descriptions, exposure duration, collection procedures, or additional safety context. The top serious-event terms are not listed.

Zero recorded affected participants is not equivalent to proof of zero risk. Safety interpretation depends on the completeness and duration of monitoring, event definitions, and the amount of exposure, details that are not reported in the registry record. The phase-specific figures should also remain separate: the primary outcome population is Evaluation phase only, while the adverse-event information includes both phases. They should not be combined to infer a single safety rate or to compare phases, particularly because the record does not describe a controlled safety comparison.

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

The registry record describes a study focused on short-term arm function during and after use of Compliant Passive Arm Support. Its primary endpoint organizes average change across assessed WMFT tasks into increased functionality, no change, or decreased functionality. This framing is useful for understanding what the investigators intended to measure: task performance over an interval of approximately 30 minutes, rather than a long-term clinical outcome.

The primary endpoint’s numerical result is not reported in the registry record. It is therefore not possible to state how many tasks fell into any category, which direction was most common, or how certain the observed classifications were. The record does not report effect estimates, confidence intervals, or a primary-outcome p-value. A description of the planned measurement should not be mistaken for evidence of a particular result.

Even if the primary outcome values were available, the single-group design would constrain what they could establish. Baseline and follow-up assessments can describe observed change over time, but the absence of a concurrent randomized control group means the change cannot be attributed uniquely to the device from this comparison alone. The registry explicitly notes that this is not a randomized comparison. It does not report a longer-term follow-up outcome that would answer whether performance changes persisted beyond the short assessment interval.

The task-count summary also does not quantify the size of each change or show whether participants had similar responses. The record does not report task-level numerical results or participant-level variation. Consequently, the endpoint’s scope should not be expanded into claims about an individual participant’s response, the magnitude of functional change, or broader effects beyond the measured tasks.

The findings that can be stated from the registry are limited to the study design, measurement description, timing, and reported adverse-event counts. The record does not establish an efficacy conclusion, a causal effect, or a long-term outcome. Its value lies in the specific short-term measurement question it poses, while the missing outcome values and limited design leave that question unanswered in the registry record.

13. Why This Trial Matters Statistically

This trial is statistically interesting because it asks about performance over a short interval rather than only change across a longer course of therapy. The registry’s stated rationale is that researchers know robot-assisted therapies improve motor performance over weeks, while the effects over minutes or hours are less clear. A brief assessment interval raises measurement questions that differ from those in a long-term treatment comparison: whether the task score is sufficiently responsive to short-term change, how repeated assessment is handled, and how much uncertainty surrounds the observed pattern.

The endpoint’s task-based count is also a useful example of how outcome definitions shape interpretation. Counting tasks with average increased functionality, no change, or decreased functionality gives a compact description of direction across tasks. It does not communicate the magnitude of change, participant-to-participant variation, or the precision of those categories. Understanding those trade-offs is important when interpreting count outcomes in rehabilitation and device research.

The design helps clarify a separate statistical lesson. A baseline-to-follow-up assessment can describe a temporal pattern, but a single-group study does not supply the randomized control comparison needed to isolate a treatment effect from other time-related influences. The registry itself cautions that the study is not a randomized comparison. Clear separation between descriptive change and causal inference is essential in interpreting such work.

Finally, the record illustrates the importance of complete reporting. It gives the outcome definition, timing, and target phase, but not the numerical primary result or its uncertainty. Without those values and a statistical analysis plan, readers cannot assess the observed pattern, precision, or handling of missing data and multiple analyses. The distinction between a focused research question and a fully interpretable numerical result is central to sound statistical reading.

14. 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.

The most important limitation for interpreting the primary endpoint is that its numerical result is not reported in the registry record. The task counts for increased functionality, no change, and decreased functionality are absent, as are uncertainty measures. This prevents assessment of the result’s direction, size, and precision.

The single-group design and absence of a randomized comparator limit causal interpretation. The baseline assessment without the device and the assessment after approximately 30 minutes of use offer a temporal comparison, but the record does not establish that any change was caused by device use. The registry also does not report whether other factors affecting performance were assessed or controlled.

The primary endpoint is limited to the Evaluation phase, but the number of participants contributing complete primary-outcome data is not reported. The record does not explain missingness, exclusions, or the analysis population in further detail. This limits the ability to judge whether the analyzed observations represent all participants intended for the Evaluation phase.

The endpoint is based on task classifications and does not, as described, report the magnitude of change for each task or variation across participants. The record does not state the method or threshold used to assign the change categories. It also excludes Grip strength because necessary equipment could not be acquired and treats Weight to box in a separate analysis, so the primary outcome does not cover every WMFT task.

The registry does not report the trial’s statistical plan, model or test, confidence intervals, multiplicity approach, or missing-data method. The planning illustrations use assumptions for a two-arm design that differ from the reported single-group model, and they are not evidence of observed power. Adverse-event figures are reported by phase, but event descriptions and further monitoring details are not reported in the registry record.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Effects of Post-Stroke Upper Extremity Assistance — NCT05036642.
• Registry record last updated 2023-10-25; status last verified 2023-10.
• Results first posted 2023-10-25.
• 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.

18. Record Summary

NCT05036642, Effects of Post-Stroke Upper Extremity Assistance, is a completed Stanford University study of Compliant Passive Arm Support in stroke survivors with upper-extremity motor impairments. The registry lists enrollment of 12, a single-group model, no masking, and allocation recorded as NA. The study dates are 2022-07-26 to 2022-09-07 for start and primary completion.

The primary outcome is WMFT change in functional score by task, assessed at baseline without the device and after approximately 30 minutes of device use. Data for this outcome were collected in the Evaluation phase only. The registry describes 15 assessed tasks, with Weight to box in a separate analysis and Grip strength excluded because necessary equipment could not be acquired. Numerical primary-outcome results and the statistical analysis plan are not reported in the registry record.

The registry lists a Participant Satisfaction Survey as a secondary outcome. Adverse-event counts are recorded separately for Pilot and Evaluation phases, with zero affected participants and zero deaths reported in each phase. The study is not a randomized comparison, so the design supports description of the stated short-term assessment, not a like-for-like controlled comparison.