Rehabilitating and Decelerating Language Loss in Primary Progressive Aphasia With tDCS Plus Language Therapy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Rehabilitating and Decelerating Language Loss in Primary Progressive Aphasia With tDCS Plus Language Therapy (NCT03728582), a randomized na study. The primary outcome is Change in Accuracy of Naming Trained Verbs (0 to 15, Higher Scores Reflect Better Naming Abilities) as Assessed by Object and Action Naming Battery (OANB).

2. Clinical Question

3. Trial Design

This completed study was randomized, triple masked, and used a crossover model. It enrolled 8 people with primary progressive aphasia. The registry lists Johns Hopkins University as the sponsor. The recorded study start was 2018-10-10, and primary completion was 2022-03-30. The phase is listed as NA.

Participants received active tDCS paired with speech-language therapy and sham tDCS paired with speech-language therapy. In the active-condition description, tDCS began at the start of a 45min therapy session and lasted for 20 min. Therapy targeted verb naming in sentence context. The sham-condition description includes oral and written naming and also specifies verb naming in a sentence context. The record describes a 2 month washout period before participants switched conditions.

A crossover design means that participants receive more than one study condition in sequence, so comparisons can use within-participant information. The effect of treatment period, order, and any carryover from an earlier condition can matter when interpreting such a design. The registry record does not report the sequence allocation details, the number of periods or sessions, or how period and carryover effects were addressed in the analysis. Although the study is identified as randomized, the specific randomization procedure is not reported in the registry record.

The primary endpoint is described as change in accuracy for trained verbs on the OANB, from baseline to 4-7 weeks after treatment begins. The registry lists 7 observations for each condition. Those arm-level counts should not be treated as a participant flow account: the record does not explain how the posted result counts relate to enrollment or the crossover sequence. Reasons for any missing outcome observations are not reported in the registry record.

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.

Intervention Period 1 (up to 6 Weeks)

Reasons for not completing this period:

Washout (8 Weeks)

Reasons for not completing this period:

Intervention Period 2 (Up to 6 Weeks)

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 Accuracy of Naming Trained Verbs (0 to 15, Higher Scores Reflect Better Naming Abilities) as Assessed by Object and Action Naming Battery (OANB), measured over Baseline, and 4-7 weeks after treatment begins.

The primary endpoint measures accuracy when naming trained verbs, using the Object and Action Naming Battery (OANB). The trained-verb component contains 15 items, and higher scores indicate better naming ability. The endpoint is described as change from baseline, with assessment 4-7 weeks after treatment begins. The broader OANB includes 262 items: 100 verbs, comprising 15 trained and 85 untrained verbs, and 162 nouns. The primary question focuses on the trained verbs rather than the full battery.

Training-specific naming is relevant to the study question because the language therapy targeted verb naming in sentence context. A score on trained items describes performance on the items targeted in therapy; it does not by itself establish whether naming improved for untrained words or everyday communication. The registry lists other outcomes addressing untrained verbs, nouns, and functional communication, but those outcomes are not substitutes for the primary endpoint.

For a continuous outcome such as a naming score, a group mean describes average performance, while a standard deviation describes variability among the observations contributing to that mean. The posted result lists the active condition with a mean of 2.3 and standard deviation of 0.285, and the sham condition with a mean of 2.05 and standard deviation of 0.15. The registry does not clarify whether these posted values are baseline-to-follow-up changes or another summary of the endpoint. It also does not report an estimated treatment contrast or its uncertainty.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What do the posted means and standard deviations describe?

A mean is the arithmetic average of the observed scores contributing to a summary. A standard deviation describes how dispersed those observations are around the mean. The active-condition mean is 2.3 with a standard deviation of 0.285, and the sham-condition mean is 2.05 with a standard deviation of 0.15. These are descriptive statistics. They do not alone show whether the observed difference is compatible with chance variation, nor do they quantify the uncertainty of a contrast between conditions.

The standard deviation is different from a standard error. A standard deviation describes variability among observations; a standard error describes uncertainty in an estimated quantity, such as a mean or treatment contrast. No standard errors or confidence intervals for the primary comparison are reported in the registry record. It would be inappropriate to infer them from the arm summaries without the necessary analytic information, especially given the crossover design.

Why does the crossover design matter?

In a crossover trial, each participant receives multiple conditions in sequence. The central comparison can therefore use information about how each person performed under the different conditions. Scores from the same participant are usually correlated: a person's scores may tend to be relatively high or low across conditions. A comparison that treats those observations as unrelated can misrepresent uncertainty.

Period and sequence also matter. A period effect is a change associated with when a condition is received, rather than with the condition itself. A carryover effect occurs when the experience of an earlier condition continues to influence outcomes during a later condition. A washout period is intended to allow earlier effects to diminish, but its presence alone does not demonstrate that carryover was absent. The registry describes a 2 month washout period, but does not report how period, sequence, or carryover effects were evaluated or incorporated in the statistical analysis.

The record gives condition-specific means and standard deviations, but not participant-level paired results, the standard deviation of within-person differences, or a model-based treatment contrast. Those missing details prevent reconstruction of a crossover comparison from the posted summaries.

What would an inferential comparison add?

An estimated contrast would quantify the difference between conditions using a specified statistical method. Its confidence interval would describe the precision of that estimate, with narrower intervals generally indicating greater precision under the model. A hypothesis test could also report a p-value, which measures how incompatible the observed data are with a stated null model, under the test's assumptions. A p-value is not the probability that a treatment is effective, and statistical significance does not by itself establish practical importance.

No primary contrast, confidence interval, or p-value is reported in the registry record. The record lists no analysis records and says that values were missing or not estimable for a two-arm comparison of the continuous endpoint. Therefore, the posted descriptive summaries do not support a statement about statistical significance or the precision of the condition difference.

How should the endpoint scale be understood?

The primary outcome is trained-verb naming accuracy on a measure with a possible trained-item score from 0 to 15. Higher scores reflect better naming ability. The primary outcome description identifies a baseline assessment and follow-up 4-7 weeks after treatment begins. A change score ordinarily represents a participant's follow-up score relative to their baseline score. However, the registry does not clarify whether the listed arm means are calculated change scores, means at a particular assessment, or another summary. This ambiguity limits interpretation of the numerical values as change.

The outcome is continuous in the registry's classification, but it is bounded by the scoring scale. Statistical methods should be suitable for the outcome's distribution and design. General statistical practice may consider whether assumptions such as approximate normality are reasonable, and whether an analysis accounts for repeated observations. The registry record does not report diagnostic checks, model assumptions, or any transformation or alternative method.

What does over-dispersion mean in general?

Over-dispersion is a general issue in statistical modeling when observed variability is greater than a model assumes. If a model underestimates variability, its standard errors may be too small and its confidence intervals too narrow. The primary endpoint is reported as a continuous score, and the registry does not state that over-dispersion was assessed or that it was an issue. This concept should not be taken as a description of a finding in this trial.

What can the planning assumptions tell readers?

The record describes a sample-size planning illustration based on the observed effect, using a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. It also states that the effect to detect is a stated fraction of the observed effect, and that non-time-to-event planning results are expressed as participants per arm. No planning sample-size estimates are reported in the registry record. Such estimates, when available, would be planning illustrations rather than observed power. They would not replace an analysis of the trial outcomes or establish the study's actual power after results are known.

The registry does not report a trial-specific sample-size justification, statistical analysis plan, or information about interim analyses. The planning assumptions alone do not establish that they were used to design this study.

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("Active tDCS Plus Speech-Language Therapy", "Sham Plus Speech-Language Therapy")
n   <- c(7, 7)
m   <- c(2.3, 2.05)
sp  <- c(0.285, 0.15)
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  Active tDCS Plus Speech-Language Therapy  7  2.300  0.285  2.036  2.564
CARM  Sham Plus Speech-Language Therapy  7  2.050  0.150  1.911  2.189

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 5 secondary outcomes in addition to the primary outcome. They cover naming accuracy for trained and untrained verbs and nouns, and functional communication assessed by the Functional Assessment of Communication Skills for Adults (ASHA FACS). Some outcome descriptions are repeated or closely overlap in the record. The registry does not report whether the outcomes were arranged in a testing hierarchy, whether any multiplicity adjustment was applied, or how secondary findings were selected for emphasis.

Multiplicity arises when several outcomes or statistical tests are considered. In general, testing more hypotheses can increase the chance of at least one nominally significant result if no adjustment or other prespecified strategy is used. That is a general statistical concern, not evidence that a particular error occurred in this trial. Because the record reports no inferential analyses, it provides no basis for evaluating the handling or consequences of multiple testing.

Missing-data methods and reasons for missing observations are not reported in the registry record. The record does not describe an intention-to-treat population, a per-protocol population, or another analysis population. It also does not explain whether every participant completed both crossover conditions or both outcome assessments. The enrollment is listed as 8, while each condition's reported result has 7 observations; the record does not explain that accounting, so it should not be interpreted as a complete participant-flow description.

In crossover studies, incomplete participation can affect whether paired condition comparisons are available. General approaches to missing data depend on why observations are missing and on the analysis model. Without participant flow, missingness reasons, and analysis details, the potential effect of missing data on the primary comparison cannot be assessed.

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.

Most frequent non-serious adverse event terms

The registry reports no serious adverse events among the 7 at risk in the active tDCS plus speech-language therapy condition, and no serious adverse events among the 7 at risk in the sham condition. It reports no deaths in either condition. No serious adverse-event terms are listed.

For other adverse events, the registry reports 0 affected in the active condition and 1 in the sham condition, with 7 at risk in each condition. The record does not provide details of the other event in the information reported for these results. These counts describe what the registry records; they do not establish that an event was caused by a particular condition or that no unreported event occurred.

Safety counts should be interpreted in light of the study size and design. The record describes a crossover, so condition-level counts may reflect exposure periods rather than independent groups. It does not explain whether the same participant could contribute to event reporting in both conditions, how event timing was assigned, or how adverse events were monitored. The counts are descriptive and do not provide a comparative safety analysis.

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

The registry records a randomized, triple-masked crossover study of active versus sham tDCS, each paired with speech-language therapy for people with primary progressive aphasia. The primary endpoint focuses on accuracy for trained verbs on the OANB. The posted condition summaries are a mean of 2.3 with standard deviation 0.285 for active treatment and a mean of 2.05 with standard deviation 0.15 for sham treatment.

Those values are descriptive summaries, not a reported inferential comparison. The registry provides no treatment contrast, confidence interval, or p-value for the primary outcome. It also does not clarify whether the posted means represent baseline-to-follow-up changes as the endpoint title states. As a result, the summaries do not establish the direction, magnitude, or precision of a treatment effect in the intended crossover comparison.

The results do not show whether any score change was maintained beyond the recorded assessment window, whether naming generalized to untrained items, or whether functional communication changed. Those are separate questions, addressed by secondary outcomes in the registry, but no inferential analyses for them are reported. The registry's outcome descriptions do not establish the clinical importance of any score pattern.

These data also do not establish how well the results apply to people with PPA beyond the enrolled study population. The record does not report participant characteristics that would help evaluate representativeness. The listed enrollment is 8, and the condition summaries each have 7 observations; the registry does not explain the relationship between those figures.

Accordingly, the posted results are best read as limited descriptive evidence about recorded naming scores in this study. They do not establish that active tDCS is more effective than sham tDCS, nor do they establish that the conditions are equivalent. The absence of an inferential comparison is not proof of no difference; it means the registry record does not report a statistical estimate that would answer that question.

14. Why This Trial Matters Statistically

This study raises statistical questions that are especially important in small crossover trials. Because participants receive both conditions, the analysis needs to preserve the within-participant structure rather than treating each condition as an unrelated group. The registry identifies the crossover model and a washout period, but does not report a crossover-specific analysis, paired treatment estimate, or evaluation of period and carryover effects. This limits assessment of whether the primary question was analyzed in a way that reflects the design.

The record also illustrates why descriptive summaries and inferential results serve different purposes. Means and standard deviations convey observed scores and variability. They do not show the uncertainty in a condition contrast, particularly when repeated observations are related. A clear estimate with an uncertainty interval would make the precision of the comparison easier to assess. Neither is reported for the primary outcome.

Outcome definition and reporting clarity matter as well. The endpoint is described as a change from baseline, but the registry does not explain whether the posted condition means are change scores. That distinction affects what the reported values represent. Clear identification of the analysis population, missing observations, and crossover periods would also help readers understand how the reported results relate to enrollment and the study design.

Finally, the trial lists several secondary outcomes involving trained and untrained naming and functional communication. With multiple outcomes, readers need to know which analyses were prespecified and how multiplicity was handled. Those details are not reported in the registry record. The study is therefore statistically informative as an example of the distinction between a well-defined clinical question and the incomplete reporting needed to assess its estimate, precision, and robustness.

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.

Several trial-specific limitations affect interpretation. The registry lists an enrollment of 8 and 7 observations for each condition, but does not report participant flow or explain the relationship between enrollment and condition-level results. It also does not provide participant characteristics, so the characteristics of the analyzed population and the applicability of the findings cannot be assessed from the record.

The primary endpoint is described as change from baseline, yet the registry does not clarify what the posted condition means represent. No comparison estimate, confidence interval, p-value, or analysis record is reported. The record indicates that values were missing or not estimable for a two-arm continuous comparison. Without paired outcomes or the model details, a crossover treatment contrast cannot be reconstructed.

Details needed to evaluate the crossover analysis are not reported: the sequence allocation procedure, period-specific outcomes, and the handling of period or carryover effects. The study description includes a 2 month washout period, but the registry does not report evidence about carryover or the rationale for the washout duration. The number and timing of treatment sessions are also not reported in the registry record.

The record does not describe missing-data methods, analysis populations, an analysis plan, multiplicity handling, or interim analyses. The primary endpoint's recorded time frame is stated, but further follow-up results are not reported. Secondary outcomes are listed, but inferential results are not available in the registry information. These reporting gaps prevent a fuller evaluation of the treatment comparison, robustness, and outcome pattern.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Rehabilitating and Decelerating Language Loss in Primary Progressive Aphasia With tDCS Plus Language Therapy — NCT03728582.
• Registry record last updated 2023-04-27; status last verified 2023-04.
• Results first posted 2023-04-19.
• 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, triple-masked crossover study sponsored by Johns Hopkins University. It enrolled 8 people with primary progressive aphasia and compared active tDCS plus speech-language therapy with sham tDCS plus speech-language therapy. The study record gives a 2 month washout period between conditions.

The primary endpoint is change in accuracy for trained-verb naming on the OANB, with higher scores indicating better naming ability. The registry lists a mean of 2.3 and standard deviation of 0.285 for the active condition, and a mean of 2.05 and standard deviation of 0.15 for the sham condition; each condition has 7 observations. No inferential comparison, confidence interval, or p-value is reported, and the meaning of the posted means relative to the stated change endpoint is not clarified. Serious adverse events and deaths are recorded as zero in each condition. The registry record does not provide enough analysis detail to determine the precision or statistical significance of a condition difference.