Finding an Optimal Latency for Paired Associative Stimulation in People With Chronic Stroke: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Finding an Optimal Latency for Paired Associative Stimulation in People With Chronic Stroke (NCT02188420), a randomized na study. The primary outcome is Change in Cortical Excitability Using Single TMS Pulses.

2. Clinical Question

The question is about the immediate change in a physiological measurement after stimulation, not a reported measure of functional recovery. The posted primary outcome is the change in the average size of motor evoked potentials elicited by single transcranial magnetic stimulation pulses. Measurements were taken before and after paired associative stimulation at each session, with the post-test occurring during the 60 minutes after application.

The registry’s rationale describes paired associative stimulation as a peripheral nerve stimulus followed shortly afterward by a cortical stimulus. The investigators sought to identify an interpulse interval before determining whether the approach would translate to people with stroke. The record does not report an outcome about symptoms, daily function, or longer-term health.

3. Trial Design

The registry describes a randomized, single-masked crossover study and records an enrollment of 4. The study status is completed. It lists University of Minnesota as sponsor, with a start date of 2014-09 and primary completion in 2015-05. The masking field says single, but the identity of the masked party is not reported in the registry record. The method used to generate or conceal the randomized order is also not reported in the registry record.

In a crossover design, participants receive conditions in sequence, so comparisons may use measurements from the same person under different conditions. The registry lists sequences containing the four stimulation conditions: latency minus 3ms, latency minus 5ms, latency minus 7ms, and latency plus 100ms. Its arm descriptions state that a washout period of 1 week followed a condition before the next level was applied. The record does not describe how sequence order was assigned beyond identifying the study as randomized, or report whether all participants completed every condition.

The four condition-level outcome entries each report n=3. The overall enrollment is 4, while safety entries list 4 at risk in each condition. These figures describe different registry entries; the record does not explain the difference between the enrollment and the outcome counts. They should not be treated as interchangeable denominators. In particular, the outcome entries do not establish how many people contributed paired measurements across all conditions.

The study concerns people with stroke, and its title and description specify chronic stroke. Further participant characteristics, recruitment details, and inclusion and exclusion criteria are not reported in the registry record. The record does not report a sample-size justification, planned stopping rule, or other details of the study’s operational design.

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: There were no significant events or excluded participants prior to being assigned an intervention order

Overall Study

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 Change in Cortical Excitability Using Single TMS Pulses, measured over Change from pretest (immediately prior to PAS application) to posttest which will occur over the 60 minutes that follow PAS application..

The primary endpoint is titled “Change in Cortical Excitability Using Single TMS Pulses.” It is measured in Log10(microvolts), and the registry describes the summary as a mean with standard deviation. At each session, investigators assessed the average size of 20 motor evoked potentials using an electromyography signal in response to single transcranial magnetic stimulation pulses to the motor cortex. The endpoint is the change from the pretest, taken immediately before paired associative stimulation, to the post-test. The post-test occurred during the 60 minutes after stimulation.

In general, a physiological endpoint can help quantify an immediate response to an experimental intervention, while remaining distinct from a measure of how a person feels or functions. This measure was relevant to the study’s stated question because the investigators were testing whether different timing intervals were associated with changes in cortical excitability. The registry describes the research aim in terms of inhibition of the nonstroke hemisphere, and states a hypothesis favoring latency minus 5ms.

The record does not explain how to interpret the sign of a positive or negative change in terms of more or less inhibition. It also does not report a clinically meaningful threshold for this measure. Therefore, the direction and magnitude of the posted changes should be understood as changes in the recorded endpoint, without assigning a clinical interpretation that the registry does not specify.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies a continuous primary outcome summarized by a mean and standard deviation. The outcome is change in cortical excitability measured in Log10(microvolts). The reported values are condition-level summaries. They give the average recorded change and the spread of measurements within each condition, but do not provide a formal comparison among conditions.

What does a mean describe?

In general, a mean is a measure of the center of a set of observations. Here it summarizes the recorded changes in motor evoked potential size for a stimulation condition. A mean can describe the observations without showing how reliably it estimates an underlying average. That reliability depends on such things as the number of observations and their variability. The registry gives a condition-level count and standard deviation, but no confidence interval for the mean.

What does the standard deviation describe?

In general, a standard deviation measures how dispersed observations are around their mean, on the scale of the measured values. A larger standard deviation indicates more spread in the observations; it is not a measure of uncertainty in a difference between conditions. The reported standard deviations vary across the four conditions, with greater spread reported for latency minus 7ms and latency plus 100ms than for latency minus 3ms and latency minus 5ms. The small reported condition counts mean that these dispersion estimates themselves may be unstable.

Why does the crossover design matter?

In general, crossover studies can compare conditions within the same participant. Such a comparison accounts for the fact that each person may have a different underlying level of response. An appropriate analysis needs to reflect the repeated measurements and the order in which conditions were received. If the same people contribute measurements to multiple conditions, those measurements are related rather than independent. The registry describes a randomized crossover design and a washout period of 1 week between levels, but does not report the statistical model, the handling of sequence or period effects, or whether carryover was assessed.

For a paired comparison, the uncertainty depends in part on how strongly the measurements from the same participant move together across conditions. The condition means and standard deviations alone do not reveal that relationship. They are therefore insufficient to reconstruct a valid crossover test or its confidence interval. The registry reports no such test or interval.

Why is over-dispersion relevant in general?

In general, over-dispersion means that observed variation is greater than a statistical model expects. It can make model-based standard errors too small if the model fails to allow for the extra variation. The primary outcome is continuous, and the registry does not report an analysis model or an assessment of over-dispersion. The differing standard deviations show that spread was not identical across condition summaries, but they do not by themselves establish over-dispersion or identify a suitable model.

What can a confidence interval add?

In general, a confidence interval expresses uncertainty around an estimated quantity, such as a mean or a difference. It can show which values are reasonably compatible with the data under the method used. The registry’s primary outcome entries have no reported confidence limits, and there is no confidence interval for a comparison. Consequently, the posted summaries do not show the precision of a contrast between stimulation intervals.

The registry’s sample-size assumptions describe planning illustrations based on the observed effect, using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The illustration varies the fraction of the observed effect to detect. These are planning assumptions, not evidence of observed power and not necessarily the study’s design rationale. The record does not report the trial’s own sample-size justification, and no participant-per-arm planning results are described in the registry entries used for this interpretation.

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("Latency Minus 3ms", "Latency Minus 5ms", "Latency Minus 7ms", "Latency Plus 100ms")
n   <- c(3, 3, 3, 3)
m   <- c(-0.07667, 0.11, -0.05667, 0.1233)
sp  <- c(0.06184, 0.06164, 0.2625, 0.15173)
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  Latency Minus 3ms  3  -0.077  0.062  -0.230  0.077
CARM  Latency Minus 5ms  3  0.110  0.062  -0.043  0.263
CARM  Latency Minus 7ms  3  -0.057  0.263  -0.709  0.595
CARM  Latency Plus 100ms  3  0.123  0.152  -0.254  0.500

11. Multiplicity, Missing Data and Analysis Populations

The primary outcome concerns four stimulation conditions. Comparing several conditions can create a multiplicity issue: in general, making multiple statistical tests increases the chance of at least one apparently notable result when all comparisons are considered together. A statistical plan may address this by specifying a primary contrast, adjusting for multiple comparisons, or using another prespecified approach. The registry record does not report a multiplicity strategy, a hierarchy of comparisons, or a prespecified contrast.

No primary-endpoint analyses are reported, and the record does not provide p-values or confidence intervals for comparisons. It is therefore not possible to assess how the investigators may have handled multiplicity in an analysis, or whether any adjustment was used. The registry also lists no posted secondary outcomes. The analysis plan, including any interim analyses, is not reported in the registry record.

The outcome counts are n=3 for each condition, while the enrollment is 4. The record does not explain whether this reflects missing measurements, incomplete participation in a condition, or another reporting difference. It does not report an imputation method or describe how missing outcome data were handled. No analysis population is specified, such as an intention-to-treat or per-protocol population, and the record does not state which participants or sessions were included in the posted outcome summaries.

In general, missing-data methods and analysis populations matter because they define which observations contribute to an estimate and what assumptions are being made about absent measurements. Without those details, the condition summaries should be read as the registry’s reported descriptive results, rather than as estimates with a fully described analysis basis.

12. 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 entries list 4 participants at risk in each of the four conditions. For every condition, the registry reports 0 participants affected by serious adverse events, 0 affected by other adverse events, and 0 deaths. No serious adverse-event terms are listed.

These entries describe what the registry reports for the recorded safety categories and their stated at-risk counts. A zero affected count means that the registry records no affected participants in that category; it does not establish that an event could never occur or that the intervention has no risk. The record does not provide event descriptions beyond the categories, additional severity details, or a narrative account of safety monitoring.

The safety counts use 4 at risk per condition, while the primary outcome entries report n=3 per condition. The registry does not explain the difference, so the safety and efficacy denominators should not be assumed to represent the same set of observations. The crossover design also means that condition entries are not necessarily independent groups. The posted adverse-event figures are best read as condition-level registry counts, not as a comparative safety analysis.

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

The record describes a small randomized crossover investigation of timing intervals for paired associative stimulation in people with chronic stroke. It reports mean changes in a physiological measure for four conditions, along with standard deviations and condition-level outcome counts. It also records no affected participants in the listed serious or other adverse-event categories and no deaths.

These results do not establish which interval is best, whether any interval produces a meaningful change, or whether one condition differs from another. The registry reports no statistical comparison, confidence interval, or p-value for the primary outcome. The sign of a change is not linked in the record to a stated interpretation of more or less cortical inhibition. The stated hypothesis about latency minus 5ms therefore cannot be evaluated from these summaries alone.

The results also do not provide evidence about functional outcomes or longer-term effects. The primary endpoint concerns the immediate change in motor evoked potential size after stimulation, with post-testing during the 60 minutes after application. The registry does not report outcomes on recovery, disability, quality of life, or persistence of response.

In general, a small exploratory study can describe observations and inform questions for later research, but its estimates can be imprecise and sensitive to individual observations. Here, the condition outcome entries each report n=3, and the registry does not explain how those counts relate to the enrollment of 4. The record does not report a sample-size justification or an analysis plan that would support a more definitive interpretation.

The appropriate reading is therefore descriptive and limited to the posted registry results. They show condition-level changes in a physiological endpoint, but do not establish a reliable ranking, a clinically relevant effect, or a safety conclusion beyond the reported event counts.

14. Why This Trial Matters Statistically

This trial is statistically notable because it asks a within-person timing question using a crossover design, but the posted results are condition-level summaries rather than a reported crossover analysis. In general, a crossover design can make efficient use of repeated measurements, yet valid inference depends on preserving the pairing and accounting for sequence and time-related features. The registry does not report how these features were handled analytically.

The distinction between descriptive and comparative results is central. A mean for each condition can show the direction of the recorded change, and a standard deviation can describe its spread. Neither alone demonstrates that the conditions differ. In this record, there is no reported contrast, uncertainty interval, or hypothesis test, so the numerical ordering of means should not be taken as evidence of a statistically established winner.

The condition-level counts also deserve attention. The outcome entries report n=3 for each condition, while enrollment is 4 and each safety entry lists 4 at risk. The record offers no reconciliation. Transparent reporting of the analysis population, missing measurements, and paired data would help readers understand the basis of the estimates and the precision of a condition comparison.

Finally, the research question is mechanistic and narrowly focused on a short-term physiological endpoint. That can be useful for generating hypotheses about stimulation timing, but it does not answer whether a timing choice leads to functional improvement. The record’s limitations make it especially important to distinguish a hypothesis-generating description from a confirmatory statistical conclusion.

15. 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 main limitation is the small reported outcome count: each condition has n=3, while the enrollment is 4. The registry does not explain the difference or state how many participants contributed complete repeated measurements. This prevents a clear assessment of the analysis population and limits interpretation of the condition summaries.

The record provides no statistical comparison for the primary outcome. It reports no confidence intervals, p-values, adjusted or unadjusted estimates, or analysis model. Because this is a crossover study, the absence of paired estimates and information about within-person relationships is particularly important. The registry also does not report whether sequence, period, or carryover effects were considered.

The endpoint is an immediate physiological measure, assessed during the 60 minutes after stimulation. The record does not report longer-term follow-up or functional outcomes. It does not define the sign of change as an increase or decrease in inhibition, nor does it specify a threshold that would indicate a meaningful change.

Further limitations are that the identity of the masked party, randomization procedure, missing-data approach, multiplicity strategy, and trial-specific sample-size justification are not reported in the registry record. The adverse-event entries report no affected participants, but provide limited event detail. These omissions do not show that the methods were absent; they mean that the registry record does not describe them.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Finding an Optimal Latency for Paired Associative Stimulation in People With Chronic Stroke — NCT02188420.
• Registry record last updated 2025-02-06; status last verified 2025-02.
• Results first posted 2025-02-06.
• 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.

19. Record Summary

The registry describes a completed randomized, single-masked crossover study sponsored by University of Minnesota. It records enrollment of 4 and dates of 2014-09 for study start and 2015-05 for primary completion. The study examined paired associative stimulation at latency minus 3ms, latency minus 5ms, latency minus 7ms, and latency plus 100ms in people with chronic stroke.

The primary endpoint was change in cortical excitability using single transcranial magnetic stimulation pulses, summarized as a mean and standard deviation. The posted condition-level outcome entries each report n=3. No primary-outcome analyses or secondary outcomes are reported. The safety entries report 0 serious adverse events, 0 other adverse events, and 0 deaths in each condition, with 4 at risk per condition.

The registry states a hypothesis favoring latency minus 5ms for inhibition of the nonstroke hemisphere, but does not report a statistical comparison or define how the direction of endpoint change maps to inhibition. The findings therefore remain descriptive in the registry record.