Repetitive Transcranial Magnetic Stimulation for "Voices": Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Repetitive Transcranial Magnetic Stimulation for "Voices" (NCT00004980), a randomized phase 2 study. The primary outcome is Hallucination Change Score (HCS) After 9 Active/Sham rTMS Sessions.

2. Clinical Question

The endpoint is intended to capture change in hallucination severity over the intervention period. Its scale anchors distinguish stopped hallucinations, no change from baseline, and hallucinations twice as severe as baseline. The primary comparison therefore describes group differences on that particular scale at the specified assessment point; it is not a direct measure of longer-term outcomes.

3. Trial Design

The registry record describes a completed phase 2, randomized, triple-masked, parallel-group trial sponsored by Yale University. Schizophrenia is the listed condition, and enrollment was 50. The record gives a start date of 1999-12 and a primary completion date of 2005-01.

Participants were assigned to active rTMS or sham stimulation. Active treatment was delivered to the left temporoparietal cortex at 1 Hz and 90% motor threshold, for 16 minutes per session. Both groups were scheduled for one session per day for nine consecutive days, with weekends excepted. The sham sessions also lasted 16 minutes. Further details about how randomization was generated, how allocation was concealed, or how masking was maintained are not reported in the registry record.

The primary endpoint was HCS after the ninth active or sham session, with the timeframe described as up to 2 weeks or the end of intervention. The listed analysis population used last observation carried forward (LOCF). The record also lists change from baseline in hallucination frequency, the Clinical Global Improvement (CGI) Scale, and responder status as secondary outcomes.

The record reports 27 participants in the active group and 23 in the placebo group for the primary outcome. Some secondary outcome counts differ: the CGI outcome includes 26 active-group and 22 placebo-group participants. The reasons for these differences, participant flow after randomization, and the handling of withdrawals are not reported in the registry record. The original protocol, details of the trial’s planned analysis, and the rationale used to choose the enrollment target are also 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.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Last Observation Carried Forward (LOCF)

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 Hallucination Change Score (HCS) After 9 Active/Sham rTMS Sessions, measured over After the 9th active/sham rTMS session (up to 2 weeks or end of intervention).

The primary endpoint is the Hallucination Change Score after the ninth active or sham rTMS session. It is reported as a mean on a scale whose stated anchors are 0 for hallucinations stopped, 10 for no change from baseline, and 20 for hallucinations twice as severe as baseline. Lower scores therefore correspond to less severe hallucinations relative to the baseline reference on this scale. The registry describes the outcome as change in hallucination severity relative to baseline.

The posted group means are 5.85 for active rTMS and 8.61 for placebo. The registry’s reported mean difference, active group minus placebo group, is -2.76. This contrast expresses the difference between group averages on the endpoint’s scale; it does not show how every individual participant changed. The standard deviations are 2.85 and 3.8, respectively, indicating variation in participant scores within each group.

A continuous score can retain gradations in response rather than reducing every participant to a responder or non-responder category. That is a general reason such outcomes may be useful. The endpoint is measured at a defined post-intervention assessment, so it describes that time point rather than persistence of a difference after treatment. The registry identifies LOCF as the population approach for this primary outcome. How missing observations were handled in detail, and why this endpoint was selected over other measures, are not reported in the registry record.

7. Results Posted in the Registry

• Active Repetitive Transcanial Magnetic Stimulation: mean 5.85 versus 8.61 on control; difference -2.76 (95% CI -4.65 to -0.87), standardized effect d = -0.83; z-test p = 0.004.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Change From Baseline in Hallucination Frequency After 9 Active/Shame rTMS Sessions

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Active Repetitive Transcanial Magnetic Stimulation: mean 1.42 versus 0.15 on control; difference 1.27 (95% CI 0.54 to 2.00), standardized effect d = 0.95; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Active Repetitive Transcanial Magnetic Stimulation than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Clinical Global Improvement (CGI) Scale After 9 Active/Shame rTMS Sessions

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Active Repetitive Transcanial Magnetic Stimulation: mean 2.84 versus 3.79 on control; difference -0.95 (95% CI -1.44 to -0.46), standardized effect d = -1.10; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Active Repetitive Transcanial Magnetic Stimulation than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Responder Status

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Active Repetitive Transcanial Magnetic Stimulation: 51.9% (14/27) versus 17.4% (4/23) on control; risk difference 34.5 percentage points (95% CI 10.1 to 58.9), risk ratio 2.98 (1.14–7.80); unadjusted two-proportion z-test p = 0.011.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Active Repetitive Transcanial Magnetic Stimulation than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary outcomes cover three different aspects: change from baseline in hallucination frequency, clinical global improvement, and responder status. The posted comparisons all favor lower scores or a greater responder count in the active group according to the stated outcome measures. This overall direction is compatible with the direction of the primary HCS comparison, but the measures are not interchangeable. They capture different constructs and use different scales or outcome formats.

The registry reports two-sided t-test results for the continuous secondary measures and a chi-squared result for responder status. The CGI analysis is described in the registry as a median difference, whereas its group summaries are reported as means and standard deviations. The record does not explain this difference in summaries or provide further analytic detail. For responder status, the record does not state the definition used to classify a participant as a responder, which limits interpretation of the category beyond the counts and test result.

Secondary outcomes generally carry less evidential weight than a prespecified primary outcome when many outcomes are examined. In general, considering multiple tests increases the chance that at least one result appears notable by chance if no adjustment or other plan accounts for multiplicity. The record lists multiple outcomes and analyses, but whether these secondary outcomes were prespecified, whether multiplicity was controlled, or how the results were prioritized is not reported in the registry record. Their p-values should therefore be read as unadjusted evidence for the individual comparisons, not as a single confirmatory test across all outcomes.

The secondary measures add breadth to the record by describing symptom frequency, a global rating, and responder classification. Their consistency in direction provides context for the primary comparison, but it does not remove the limitations of the sample, outcome reporting, or analysis information. No additional conclusion about durability or outcomes beyond the stated assessment timeframe follows from these secondary measurements.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Change From Baseline in Hallucination Frequency After 9 Active/Shame rTMS Sessions
• Clinical Global Improvement (CGI) Scale After 9 Active/Shame rTMS Sessions
• Responder Status

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous and is reported as a mean and standard deviation in each group. A mean describes the group’s average score; a standard deviation describes how spread out scores are around that average. Neither measure reveals the full distribution of individual scores. The posted summary gives 27 participants, a mean of 5.85, and a standard deviation of 2.85 for active rTMS, and 23 participants, a mean of 8.61, and a standard deviation of 3.8 for placebo.

What does the mean difference represent?

The primary mean difference is calculated as the active-group mean minus the placebo-group mean. The reported value is -2.76. Its sign gives the direction of the contrast: the active-group average was lower on the HCS scale at the assessment point. Because lower values on the specified scale indicate less severe hallucinations relative to baseline, the sign can be interpreted against those scale anchors. It does not mean that every person in the active group had a lower score than every person in the placebo group.

What does a two-sided t-test assess?

The registry reports a two-sided t-test for the mean difference. In general, a t-test evaluates whether the observed difference between group means is compatible with a null hypothesis of no mean difference, under the test’s assumptions. “Two-sided” means that departures in either direction are considered. The registry reports p = .005 for this analysis. A p-value is a measure of how unusual data at least as incompatible with the null hypothesis would be under the test’s assumptions; it is not the probability that the null hypothesis is true, and it does not measure the size or practical importance of a difference.

The registry record does not report the specific assumptions checks, any covariate adjustment, or further detail about the t-test implementation. The recalculation uses posted summary statistics for an unadjusted comparison and yields a mean difference of -2.76 with a 95% confidence interval from -4.649 to -0.871 and p = 0.004. It is not a replacement for a participant-level analysis or for the sponsor’s own reported analysis.

How should the scale be kept in view?

Statistical contrasts are meaningful only alongside the outcome’s measurement. Here, 0 is anchored as hallucinations stopped, 10 as no change, and 20 as hallucinations twice as severe as baseline. Those anchors establish direction and give context to group means and their difference. The registry labels the endpoint as a change score, while its description gives relative-to-baseline severity anchors. The record does not provide further material about the scale’s development or validation.

The primary outcome uses LOCF. In general, last observation carried forward assigns a participant’s last recorded value to a later assessment when the later value is missing. This approach can affect estimated group averages, and its suitability depends on the missingness and outcome process. The registry identifies LOCF but does not report the extent or pattern of missing primary outcome observations, the reasons for missingness, or sensitivity analyses using other missing-data approaches. It is therefore not possible from this record to assess how robust the primary estimate is to alternative missing-data assumptions.

How does the binary outcome differ?

Responder status is a binary outcome: participants are classified into categories rather than analyzed as a continuous score. The registry reports 14 responders among 27 active-group participants and 4 among 23 placebo-group participants, with a chi-squared analysis and p = .01. In general, a chi-squared test assesses whether observed category counts differ from those expected under no association between group and outcome. The registry record does not report the responder definition, so the clinical meaning of that classification cannot be fully evaluated from the record alone.

Different outcome types call for different summaries and analyses. A mean difference is expressed in the endpoint’s scale units; responder counts describe how many participants meet a stated classification. The estimates should not be interchanged or interpreted as if they measured the same feature. The record also reports a median difference for the CGI outcome, even though group summaries for that outcome are reported as means and standard deviations. The full analysis details and rationale for the reported statistic are not reported in the registry record.

Finally, the recalculations are unadjusted and use summary statistics. General statistical practice distinguishes such a recalculation from an analysis planned in a protocol or conducted on participant-level data. The original statistical plan, any covariates or stratification factors, and any interim analyses are not reported in the registry record. Those omissions limit what can be inferred about the intended analysis and its relationship to the posted results.

11. Confidence Intervals and Precision

A confidence interval describes a range of values compatible with the data and the analysis method, at the stated confidence level. It communicates uncertainty about the estimated group contrast; it does not mean that the true value has a particular probability of falling inside this specific interval. The registry reports a 95% confidence interval for the primary mean difference from -4.65 to -.86. Its limits are on the HCS scale, and both are on the negative side of no difference.

The recalculated 95% interval is from -4.649 to -0.871. The registry’s reported interval and the recalculated interval are close, although their limits are not identical. The registry reports p = .005; the unadjusted recalculation reports p = 0.004. Small differences in reported inferential quantities can arise from differing analysis details or rounding, but the record does not specify the source of this difference. The registry’s analysis is the trial’s reported result.

Interval width is one way to assess precision: a narrower interval indicates less uncertainty in the estimate under the analysis, while a wider interval indicates more. Precision depends on the amount and variability of the data as well as the analysis. The primary groups contain 27 and 23 participants, and their standard deviations are 2.85 and 3.8. These features help describe the data available, but they do not by themselves determine the confidence interval.

The interval concerns an average difference, not the range of individual participant outcomes. Individual scores can vary within each group, as the standard deviations indicate. The interval also does not account for possible limitations that are not represented in the analysis, such as uncertainty about missing-data assumptions or the choice of endpoint. The record identifies LOCF for the primary outcome but does not report sensitivity analyses. Thus, the interval expresses statistical precision within the reported analysis, not certainty about every aspect of the trial or about outcomes beyond the stated assessment timeframe.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures give participants per arm for the continuous and binary outcomes in this trial; they are not total event counts. The calculation assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Each planning scenario uses a stated fraction of the observed effect as the effect to detect. For time-to-event outcomes, a different presentation would give total events, but none of the listed outcomes is a time-to-event outcome.

The 80% power column gives the estimated participants per arm under which the calculation would have an 80% chance of detecting the specified effect, if the assumptions and effect size used in the calculation applied. The 90% power column gives the corresponding estimate for a 90% chance. These are planning probabilities for a hypothetical study under the calculation assumptions, not probabilities that the observed result is correct.

Rows that plan for a smaller fraction of the observed effect require more participants per arm than rows using the full observed effect. For example, the primary outcome’s planning figures increase from 23 participants per arm for 80% power at 100% of the observed effect, to 41 at 75%, and 91 at 50%. The analogous 90% figures are 31, 55, and 122. This pattern reflects a general statistical principle: smaller effects are harder to distinguish from random variation, so more participants are needed to attain the same planned power.

The original trial’s own sample-size assumptions and justification are not reported in the registry record. These calculations are retrospective planning illustrations based on the observed effect and are not observed power. They should not be read as evidence that the trial was designed with these particular targets or assumptions.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -2.760; pooled standard deviation σ = 3.319.
2. n = 2 × (1.96 + 0.842)² × 3.319² / -2.760² = 22.7, rounded up to 23 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -2.76                      # observed difference in means
sigma <- 3.31934042142913                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(4980)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 23 (80% power), 31 (90% power)
 75% of observed effect: n per arm = 41 (80% power), 55 (90% power)
 50% of observed effect: n per arm = 91 (80% power), 122 (90% power)
power.t.test: 23.7 
Simulated power at n = 23 per arm: 0.789 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry’s primary analysis reports a mean difference of -2.76, a 95% confidence interval from -4.65 to -.86, and p = .005 from a two-sided t-test. The unadjusted recalculation from the posted group summaries gives the same mean difference, a 95% confidence interval from -4.649 to -0.871, and p = 0.004. The point estimates agree, while the confidence limits and p-values differ slightly.

The registry record does not identify why the inferential values differ. The recalculation is a simple comparison from posted summary statistics; it cannot reproduce details that may depend on participant-level data or analysis choices not described in the record. No discrepancy is flagged in the record, and there is no basis to attribute the small differences to a particular cause. The sponsor’s reported analysis is the trial result to cite, while the recalculation is best treated as a check based on the posted summaries.

For the secondary outcomes, the registry also reports its own analyses, including two-sided t-tests for the continuous outcomes and a chi-squared test for responder status. Recalculated comparisons are unadjusted. A numerical comparison between analyses does not establish that the same analysis procedure was used in each case, and the record does not provide a full statistical analysis plan.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome and three secondary outcomes, with four outcome measures and four analyses posted. Multiple outcomes can create a multiplicity issue: as the number of statistical tests increases, the chance of at least one apparently notable result can increase even when the underlying effects are absent. General approaches to this issue include prespecifying a hierarchy or adjusting inference across tests. The registry record does not report whether a multiplicity strategy was planned or applied, or whether the secondary analyses were considered confirmatory or exploratory.

The primary endpoint is identified as HCS, and the registry reports its analysis separately from the secondary outcomes. The primary result should be distinguished from the secondary p-values, which do not have the same status simply because they are listed in the record. Without a protocol or statistical analysis plan, it is not possible to determine whether outcome selection, analysis methods, or timing followed a prespecified plan. Interim analyses and any stopping rules are not reported in the registry record.

For the primary measure, the record identifies a LOCF population. In general, LOCF carries the last observed value forward when a later measurement is missing. It can yield biased estimates when the carried-forward value does not represent the participant’s unobserved later outcome. Whether that concern applies depends on the reasons and pattern of missing data. The numbers of missing primary assessments, reasons for missingness, and sensitivity analyses are not reported in the registry record.

The record reports 27 and 23 participants for the primary outcome, while the CGI outcome has 26 and 22. It does not explain these differing counts. Detailed participant flow, analysis populations for each outcome, and whether all randomized participants were included in each analysis are not reported in the registry record. As a result, the effect of exclusions or missing observations cannot be assessed from the posted summaries alone. These uncertainties concern the interpretation of the analyses; they do not change the numerical results reported for the specified groups.

16. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent non-serious adverse event terms

The registry reports zero serious adverse events among 27 participants at risk in the active rTMS group and zero among 23 participants at risk in the placebo group. The posted serious-event counts therefore contain no events in either group during the reported safety accounting. The record lists no serious adverse event terms.

Zero reported events in these groups is not proof that serious events cannot occur, and the counts do not by themselves establish comparative safety. The number of participants and the trial’s observation period constrain what can be learned from a trial of this size. The registry record does not provide further detail about the event definitions, monitoring procedures, or duration of safety follow-up.

The record separately lists other adverse events affecting 16 active-group participants and 3 placebo-group participants. Those counts are distinct from the serious adverse event figures. Death counts are not reported in the registry record. These entries should be read as the posted safety information, without inferring causation or making a clinical judgment about an individual event.

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

The posted primary result is a lower mean HCS score in the active rTMS group than in the placebo group after the ninth session. The registry reports a mean difference of -2.76 and a 95% confidence interval from -4.65 to -.86. On the scale as described, lower values indicate less severe hallucinations relative to baseline. This is a comparison of group averages at the reported assessment point.

The result does not imply that every participant in the active group experienced a reduction, nor that every participant in the placebo group did not. Group means can conceal variation, and both groups have nonzero standard deviations. The confidence interval describes uncertainty about the average contrast under the reported analysis; it is not a prediction interval for individual outcomes.

The record concerns a specific intervention schedule, comparator, population, endpoint, and assessment timeframe. It does not establish persistence of an observed difference after the intervention period, and it does not provide results for outcomes beyond those listed. The primary measure is a scale with stated anchors, not a direct account of all dimensions of hallucination experience.

The record’s secondary findings provide additional outcome measurements, but they do not eliminate uncertainty about multiplicity, missing data, or analysis choices. The registry identifies LOCF for the primary endpoint, while its detailed missing-data handling and sensitivity analyses are not reported. The original analysis plan, randomization details, masking procedures, and basis for the enrollment target are also not reported.

Accordingly, the numerical results describe what was posted for this completed trial; they do not by themselves establish a broader treatment conclusion. The trial’s scale, sample, timeframe, and available registry entries define the scope of what these results can support.

18. Why This Trial Matters Statistically

This trial illustrates why the endpoint’s scale anchors matter when interpreting a mean difference. The primary contrast is negative, and the record anchors lower HCS values to less severe hallucinations relative to baseline. Without that direction-of-scale information, the sign of the contrast could be misread. Statistical reporting is most useful when an estimate is paired with a clear account of what the measurement means.

It also illustrates the distinct roles of a point estimate, confidence interval, and p-value. The mean difference conveys the size and direction of the group-average contrast in scale units. The confidence interval conveys its statistical precision under the analysis, and the p-value describes compatibility with a null hypothesis under the test assumptions. None of these alone communicates individual variation, data quality, or clinical importance.

The available summaries support a straightforward unadjusted recalculation, but that calculation is not the same as access to the original analysis plan or participant-level data. Small differences between the recalculated and registry-reported inferential quantities underline the importance of identifying which result is the trial’s own analysis and which is a summary-statistic check.

The record also shows how additional outcomes can add context while raising questions about multiplicity and interpretation. Continuous scores and responder counts represent different kinds of evidence. Without a reported responder definition or a stated multiplicity plan, secondary results should not be given the same confirmatory weight as a clearly designated primary result. Finally, incomplete reporting of missing-data patterns and analysis populations limits assessment of robustness. These are general lessons in reading clinical trial results, grounded in the features and omissions of this record.

19. 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 specific to this record constrain interpretation. The primary outcome uses LOCF, but the number and pattern of missing values, reasons for missingness, and sensitivity analyses are not reported. The primary outcome group counts are 27 and 23, while the CGI counts are 26 and 22; the record does not explain the differing counts or provide a full participant flow account.

The registry identifies randomization and triple masking, but the randomization procedure, allocation concealment, and masking implementation are not reported in the registry record. It is therefore not possible to assess those procedures in detail from this information. The full protocol and statistical analysis plan, including any prespecified hierarchy among outcomes, covariate adjustments, interim analyses, or stopping rules, are also not reported.

The HCS scale has stated anchors, but further details about its development, validation, and measurement properties are not reported. The responder definition is not reported, limiting interpretation of that secondary endpoint. The CGI analysis is reported as a median difference while group summaries are means and standard deviations; the analytic rationale is not reported.

The registry gives a post-session timeframe of up to 2 weeks or the end of intervention. Longer-term outcomes are not listed in these results, so durability cannot be assessed from this record. The enrollment of 50 and the outcome group sizes also limit the precision attainable, as reflected in the confidence interval. The rationale for the original enrollment target and the trial’s original sample-size assumptions are not reported. These limits should be considered alongside, rather than substituted for, the posted estimate.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Repetitive Transcranial Magnetic Stimulation for "Voices" — NCT00004980.
• Registry record last updated 2013-08-07; status last verified 2013-08.
• Results first posted 2012-07-09.
• 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.

23. Record Summary

The registry describes a completed phase 2 randomized, triple-masked, parallel-group trial sponsored by Yale University, with enrollment of 50 participants and schizophrenia listed as the condition. It compared active rTMS with sham stimulation, using nine scheduled sessions and assessing the primary HCS endpoint after the ninth session.

The primary registry analysis reports an active-versus-placebo mean difference of -2.76, with a 95% confidence interval from -4.65 to -.86 and p = .005. The recalculation from posted summaries gives the same estimate and a close confidence interval and p-value. Three secondary outcomes are listed, along with serious and other adverse event counts.

Important details—including the original analysis plan, missing-data patterns, and the rationale for the enrollment target—are not reported in the registry record. The posted estimates therefore describe the recorded outcomes and analyses within the limits of that reporting.