Robots Paired With tDCS in Stroke Recovery: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Robots Paired With tDCS in Stroke Recovery (NCT01726673), a randomized phase 2 study. The primary outcome is Median Change in Upper Extremity Fugl Meyer Assessment Score.

2. Clinical Question

3. Trial Design

The registry identifies this as a Phase 2, randomized, double-masked, parallel trial sponsored by Northwell Health. Its title concerns tDCS paired with robotic arm therapy for upper-extremity motor recovery after stroke. The registry lists the study status as completed, enrollment as 54, a start date of 2012-09, and primary completion as 2018-01. These dates describe the record and do not, by themselves, provide information about the duration of an individual participant’s follow-up.

The active condition combined tDCS at 2mA for 20 minutes with robotic arm therapy for 60 minutes. Stimulation was applied over the primary motor cortex in the affected hemisphere and was immediately followed by the robotic therapy. The schedule was 3x per week for 12 weeks, for 36 sessions total. The comparator used sham stimulation at 0 mA for 20 minutes at the same site and was followed by the same robotic therapy schedule. Thus, the intended contrast is the addition of active rather than sham stimulation while robotic therapy is part of both conditions.

The registry describes the study as randomized and double-masked, but the method used to generate the allocation sequence, the allocation concealment procedures, and who was masked are not reported in the registry record. The record also does not report whether participants could reliably distinguish active from sham stimulation. Those details matter when assessing how well randomization and masking might limit bias, but they cannot be inferred from the design labels alone.

The primary outcome is median change in Upper Extremity Fugl-Meyer Assessment score, measured from baseline to discharge at 12 weeks and again at 36 weeks. The efficacy description states that 45 patients completed 36 sessions and 6 month follow-up at 36 weeks and were included in the efficacy analysis. The registry record does not explain the reasons that other enrolled participants were not included in that analysis or report an intention-to-treat analysis. Its adverse-event record lists safety counts separately by treatment group.

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: Subjects were randomized to receive 36 sessions of either active tDCS immediately followed by robotic arm therapy or sham tDCS immediately followed by robotic arm therapy.

Overall Study

Analysis population for the primary outcome, as defined in the registry: 45 patients completed 36 sessions (3x/week for 12 weeks) of robotic arm training + sham or active tDCS, and 6 month follow-up at 36 weeks. These 45 participants were consequently included in the efficacy analysis.

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 Median Change in Upper Extremity Fugl Meyer Assessment Score, measured over baseline, discharge at 12 weeks (immediately following the intervention), and follow-up at 36 weeks (6 months after the intervention).

The primary endpoint is median change in the Upper Extremity Fugl-Meyer Assessment score. This scale assesses upper-extremity motor function, and the registry describes a total score range of 0-66 points, with higher values indicating better functional status. The endpoint is expressed as change from baseline, rather than simply the score at a single assessment. Change can describe how an individual’s measured function differs from their own starting point, while a comparison between treatment groups can address whether the pattern of change differs between conditions.

The registry specifies assessment from baseline to discharge at 12 weeks, immediately following the intervention, and again at 36 weeks, described as 6 months after the intervention. These time points address change at the end of the treatment period and change at follow-up. The registry lists the measure as a median and its dispersion as an inter-quartile range. The median describes the middle of the observed values, and the inter-quartile range describes the spread of the central portion of those values. These summaries can be useful when a distribution is skewed or includes values that make the mean less representative.

The registry record does not report the median changes, inter-quartile ranges, group contrast, confidence interval, or p-value. It therefore describes what was intended to be measured, but does not allow readers to determine the observed size, direction, or precision of the primary outcome difference.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes a randomized, double-masked, parallel design. In general, random allocation is intended to make treatment groups comparable on average with respect to factors known and unknown at the time of assignment. Masking is intended to reduce the risk that knowledge of treatment assignment influences care, assessment, or reporting. These are general purposes of the design features, not confirmation that the procedures were implemented successfully. The sequence-generation method, concealment procedures, and identities of those masked are not reported in the registry record.

What does a median change represent?

The primary measure is the median change in Upper Extremity Fugl-Meyer Assessment score. A change score compares a participant’s score at a specified assessment with their baseline score. The median is the middle value when the changes are ordered. Unlike the mean, it is less influenced by very large or small values. The registry also specifies the inter-quartile range, which describes the spread of the central portion of the observed values. In general, reporting a median with an inter-quartile range is informative when a distribution may be asymmetric or when extreme observations could pull the mean away from a typical value.

The registry defines the total scale as ranging from 0-66 points and says higher scores indicate better functional status. That direction applies to the score itself. For a change score, interpretation also requires clarity about the subtraction convention: whether the value is follow-up minus baseline or baseline minus follow-up. The convention is not reported in the registry record. The record does say that change was calculated from baseline at the specified assessment times, but does not provide the observed values needed to understand the direction of the group results.

Why distinguish the time points?

The registry identifies discharge at 12 weeks, immediately following the intervention, and follow-up at 36 weeks, described as 6 months after the intervention. These assessments address distinct questions. An end-of-intervention measurement describes the observed outcome after the scheduled treatment period. A later follow-up describes the outcome at a subsequent assessment. The two time points should not be treated as interchangeable, because the outcome at follow-up may not match the outcome at discharge. The registry provides no results for either time point, so it is not possible to determine whether changes were maintained, increased, or diminished.

What would a between-group estimate add?

A comparison between the active and sham conditions would quantify the contrast in outcome changes between the treatment groups. In general, an estimate gives the size and direction of that contrast, while a confidence interval conveys its precision. A p-value can address compatibility with a statistical null hypothesis under the chosen model, but it does not communicate the size or clinical importance of a difference. For a median-based endpoint, the exact method used to compare groups matters. The registry record does not report the analysis method, the group estimates, the contrast, a confidence interval, or a p-value. No such statistics can be reconstructed from the descriptions alone.

For the same reason, the outcome label does not establish whether an analysis compared change scores directly, adjusted for baseline scores, or used another approach. These methods can answer related but not identical statistical questions and can rely on different assumptions. The statistical model, assumptions, and any adjustments are not reported in the registry record. General statistical explanations of these methods should not be mistaken for a description of what the trial actually did.

How should planning illustrations be understood?

The stated planning assumptions describe an illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between the treatment conditions, and a normal approximation. The effect to detect is expressed as a stated fraction of the observed effect. In general, sample-size planning asks how many participants per arm would be required under chosen assumptions to detect a specified effect with a target level of power. Such calculations are sensitive to the assumed effect size, variability, significance level, allocation, and analysis method.

These are planning illustrations, not the number of participants actually analyzed and not observed power. Observed power is not a substitute for reporting an effect estimate and its uncertainty. The record does not give numerical planning outputs in the information available for this explanation, and the stated assumptions are not identified as the trial’s original design justification. The sample-size justification used when the trial was planned is not reported in the registry record.

In general, over-dispersion means that observed variability is greater than a model assumes. If a statistical model underestimates variability, its uncertainty estimates may be too small and its inference may be unreliable. The registry does not identify an over-dispersion issue or report a model for the primary endpoint; this concept is general and is not a claim about the trial’s data.

What can the safety counts establish?

The record lists serious adverse events, other adverse events, and deaths by treatment group. These counts are descriptive. To compare safety fully, readers would also need clear event definitions, observation periods, participant-level exposure, details about the events, and the method used to assess possible treatment relationships. Those details are not reported in the registry record. The listed counts should therefore be read as the registry’s recorded figures, not as a complete safety analysis or a test of whether the groups differ.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome and identifies secondary outcomes for median change in WOLF Motor Function Test and median change in Motor Power Manual Muscle Test Score for the Upper Extremity. It also reports 3 outcome measures and 6 analyses. The primary outcome description includes assessment at discharge at 12 weeks and follow-up at 36 weeks. These multiple outcomes, analyses, and assessment times raise a general multiplicity issue: when many statistical questions are examined, the chance of at least one apparently notable result can increase if no adjustment or prespecified hierarchy is used.

The statistical plan, prespecified hierarchy, and any multiplicity adjustment are not reported in the registry record. It is therefore not possible to determine which tests were planned as confirmatory, whether the two primary assessment times were treated as separate comparisons, or how the secondary analyses were interpreted. The outcome and analysis counts alone do not establish that every listed analysis was performed or that any particular adjustment was used.

The registry states that 45 patients completed the intervention and 36-week follow-up and were included in the efficacy analysis. Enrollment is listed as 54. The participant flow, reasons for non-completion or exclusion, and the number with available data at each assessment are not reported in the registry record. Nor are imputation methods, sensitivity analyses, or rules for handling missing outcomes reported. Without these details, the effect of missing data on the estimated results cannot be assessed.

The registry record does not state whether the primary analysis followed intention-to-treat principles, used a per-protocol population, or applied another analysis-population definition. The stated efficacy population is the group described as completing treatment and follow-up; it does not by itself establish the analysis strategy for all randomized participants. These gaps limit evaluation of both multiplicity and the potential impact of missing data.

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.

Most frequent non-serious adverse event terms

The registry lists serious adverse events for the active tDCS plus robotic arm therapy group as 0 among 30 at risk, and for the sham tDCS plus robotic arm therapy group as 0 among 24 at risk. It records other adverse events for 2 among 30 at risk in the active group and 1 among 24 at risk in the sham group. Deaths are recorded as 0 in each group. These figures are registry counts, not an adjusted comparison between treatment conditions.

No serious adverse-event terms are listed. The registry does not report descriptions of the other adverse events, their severity, timing, duration, or whether investigators considered them related to stimulation or robotic therapy. The observation period and the procedures used to collect and adjudicate adverse events are not reported in the registry record. Consequently, the counts do not indicate whether events were alike in nature or clinical importance, or whether all groups had equivalent opportunities for events to be detected.

Zero recorded serious adverse events means none are listed for the stated at-risk groups in this record; it does not establish that the treatment conditions have no serious risks. Likewise, the other-event counts do not establish a safety difference. Without further event detail and an inferential analysis, these figures are best read descriptively and kept separate from conclusions about the primary motor-function outcome.

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

The registry record establishes that a randomized, double-masked, parallel Phase 2 trial compared active tDCS plus robotic arm therapy with sham tDCS plus robotic arm therapy in people with stroke. It describes the stimulation conditions, the shared robotic therapy schedule, and a primary outcome based on median change in Upper Extremity Fugl-Meyer Assessment score. It also identifies assessment at discharge and follow-up, and states that 45 patients who completed the treatment course and follow-up were included in the efficacy analysis.

Those details clarify the intended comparison, but they do not show what the primary outcome results were. The medians, inter-quartile ranges, between-group estimate, confidence interval, and p-value are not reported in the registry record. The posted information therefore cannot establish the direction or magnitude of an outcome difference, its precision, or whether a statistical test supported a difference. It also cannot establish whether any observed change would be clinically important.

The treatment contrast is specifically active versus sham tDCS when both conditions are paired with robotic arm therapy. It cannot be interpreted as a comparison of robotic therapy against no therapy, or as a comparison of active stimulation in isolation from the shared therapy. The registry record also does not provide the methodological details needed to assess randomization and masking procedures, the statistical analysis plan, or the handling of missing data and multiple analyses.

The safety counts are likewise limited: they report no serious adverse events and no deaths in either listed group, alongside other adverse-event counts, but do not describe those events or their relationship to treatment. These data do not support broad claims about safety. The appropriate interpretation is therefore narrow: the record documents the study design, intended measurements, stated efficacy population, and recorded safety counts, while the primary efficacy result and important analytical details are not reported in the registry record.

13. Why This Trial Matters Statistically

This trial raises several important statistical questions because its central comparison is an add-on contrast. Both treatment conditions include robotic arm therapy, while active or sham tDCS differs between them. That structure determines what a group difference would mean: it would address the added stimulation condition within the robotic-therapy schedule, rather than the effect of robotic therapy alone. Keeping the comparison precise prevents overstatement about which component is being evaluated.

The primary endpoint is a median change score measured at discharge and follow-up. This makes it important to distinguish the assessment times, understand the change-score convention, and examine both the group summaries and an appropriate between-group estimate. Medians and inter-quartile ranges describe the observed distribution, but they do not alone quantify the uncertainty or the contrast between groups. The registry record does not report the outcome values or inferential results, so these statistical questions cannot be answered from the record’s outcome description.

The stated efficacy population also makes analysis-population and missing-data reporting consequential. The record says that 45 patients completed treatment and follow-up and were included in efficacy analysis, while enrollment is reported as 54. Without participant flow, reasons for non-completion, and the analysis strategy, readers cannot assess how well the efficacy analysis represents the randomized participants or how missing outcomes may have affected estimates.

Finally, the registry identifies multiple outcomes and analyses, but does not report the statistical plan or multiplicity handling. That matters when interpreting any collection of outcome tests: prespecification and adjustment can affect how much confidence to place in isolated findings. Here, the absence of posted primary results prevents evaluation of the treatment contrast, while missing methods limit assessment of how it would have been analyzed. These are reporting limitations, not evidence that a particular result occurred.

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 primary limitation of the registry record for interpreting efficacy is that it describes the endpoint without reporting its results. There are no posted medians or inter-quartile ranges for either group at either assessment time, and no group contrast, confidence interval, or p-value. Consequently, the record cannot support an assessment of effect direction, size, precision, or statistical evidence for a difference.

The record identifies a completed randomized, double-masked, parallel study, but does not report the randomization method, allocation concealment, who was masked, or how masking was maintained. It does not provide the full eligibility criteria or characteristics of the analyzed population. These omissions limit assessment of recruitment, representativeness, and protections against bias.

The efficacy description says that 45 patients completed 36 sessions and follow-up at 36 weeks and were included in the efficacy analysis, whereas enrollment is listed as 54. Participant flow, reasons for non-completion or exclusion, availability of outcome measurements by time point, and the handling of missing values are not reported in the registry record. The record also does not state whether all randomized participants were included in the primary analysis.

The statistical analysis plan, model, assumptions, baseline adjustment, multiplicity strategy, and sample-size justification are not reported in the registry record. Although planning assumptions are described for an illustration based on the observed effect, those assumptions do not establish what was used to design or analyze the trial. The record also does not provide enough detail about adverse-event definitions, timing, clinical descriptions, or relatedness assessments to support a comprehensive safety interpretation.

These limitations concern what can be concluded from the registry record. They do not imply that the study did not collect the missing information or that any particular result was obtained.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Robots Paired With tDCS in Stroke Recovery — NCT01726673.
• Registry record last updated 2021-05-14; status last verified 2021-05.
• Results first posted 2018-08-16.
• 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

The registry record describes NCT01726673, a completed Phase 2 randomized, double-masked, parallel trial sponsored by Northwell Health. It compared active tDCS plus robotic arm therapy with sham tDCS plus robotic arm therapy in people with stroke. Enrollment is listed as 54.

The primary outcome was median change in Upper Extremity Fugl-Meyer Assessment score, assessed from baseline to discharge at 12 weeks and at follow-up at 36 weeks. The record says that 45 patients completed treatment and follow-up and were included in the efficacy analysis, but does not report the outcome values or statistical comparison. It lists no serious adverse events and no deaths in either group, and records other adverse events for 2 participants in the active group and 1 in the sham group. The statistical plan and further participant-flow and safety details are not reported in the registry record.