Placebo Controled Clinical Trial Using Topiramate To Treat Posttraumatic Stress Disorder (PTSD) Patients.: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Placebo Controled Clinical Trial Using Topiramate To Treat Posttraumatic Stress Disorder (PTSD) Patients. (NCT00725920), a randomized phase 4 study. The primary outcome is Clinician Administered Posttraumatic Stress Disorder Scale.

2. Clinical Question

In statistical terms, this is a comparison of a continuous outcome between two randomized parallel groups. The posted group summaries describe the CAPS score distributions at the specified time point, while the sponsor’s reported analysis gives a two-sided t-test result. An unadjusted recalculation based on the posted means and standard deviations gives a different result; the significance of that discrepancy is discussed separately. The registry record does not report the original analysis protocol or details needed to determine why the two results differ.

3. Trial Design

This was a phase 4, randomized, parallel-group, placebo-controlled trial sponsored by Federal University of São Paulo. The registry lists the study status as completed, with a start date of 2007-01 and primary completion in 2009-11. Planned treatment and assessment lasted 12 weeks. The record reports enrollment of 35 participants: 17 in the topiramate group and 18 in the control group.

The study description says participants would receive topiramate or placebo, with the active-drug dose beginning at 25 mg/day and increasing by 25 mg weekly according to tolerance of side effects. Placebo pills were described as identical to the active-drug pills. Randomization and triple masking are recorded, but details such as the allocation-generation procedure, concealment process, and which people were masked are not reported in the registry record. The record also does not state whether treatment adherence, dose changes, or treatment discontinuation were assessed or how they were handled in analysis.

The stated assessment instruments included the CAPS, Beck Depression Inventory, Beck Anxiety Inventory, 36-Item Short Form Health Survey, and Social Adjustment Scale. The posted primary endpoint is the total CAPS score, summarized as a mean and standard deviation at 12 weeks. Although the study description mentions several clinical and quality-of-life domains, the registry record posts one outcome measure and does not post secondary outcome results.

The original sample-size justification and the trial’s planned analysis details are not reported in the registry record. The posted results include one sponsor-reported analysis and permit a simple unadjusted recalculation from the group summaries. That recalculation is not a substitute for the trial’s full analysis plan, which is not described 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:

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 Clinician Administered Posttraumatic Stress Disorder Scale, measured over 12 week.

The primary endpoint is the total score on the Clinician-Administered PTSD Scale (CAPS), assessed at 12 weeks. The registry describes CAPS as a structured interview used to diagnose PTSD and rate symptom severity. It includes 30 items addressing symptom frequency and severity, with a total score ranging from 0 to 136. The record classifies scores as subclinical from 0 to 19, mild from 20 to 39, moderate from 40 to 59, severe from 60 to 79, and extreme at 80 and above.

The registry also describes three CAPS symptom-cluster subscales: CAPS 1 for intrusive recollections, CAPS 2 for avoidance, and CAPS 3 for hyperarousal. The primary result, however, is the total CAPS score rather than a separate result for each cluster. The reported measure is continuous, and the posted results summarize it using a mean and standard deviation for each group. This allows a comparison of average scores, while the standard deviations describe the spread of scores among participants in each arm.

Using a structured symptom measure gives the trial a defined outcome for comparing groups at the specified assessment time. A group mean does not describe every participant’s experience, and the registry record does not report individual score distributions or changes from baseline. Therefore, the posted endpoint supports interpretation of the group averages at 12 weeks, but does not by itself show how many participants moved between the registry’s score categories or how scores changed for particular participants.

7. Results Posted in the Registry

• Topiramate: mean 30.41 versus 35.78 on control; difference -5.37 (95% CI -26.79 to 16.05), standardized effect d = -0.17; z-test p = 0.623.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a continuous measure, the total CAPS score at 12 weeks. The registry reports the mean and standard deviation in each randomized group. A mean gives the arithmetic average score within a group. A standard deviation describes how spread out individual scores are around that average; a larger standard deviation indicates more variability among participants. These summaries do not show the full shape of either group’s score distribution.

What does the mean difference describe?

The recalculated comparison is the topiramate mean minus the control mean, giving a mean difference of -5.37 CAPS points. A difference of zero would correspond to equal group means. A negative estimate indicates a lower observed mean in the topiramate arm, while a positive estimate would indicate a higher mean in that arm. This is a direct, unadjusted comparison of the posted summary values; it does not account for baseline scores or other covariates.

In general, a two-sample comparison of means assesses whether the observed separation between group averages is large relative to the uncertainty in that separation. A standard error represents the estimated sampling variability of an estimate. The recalculation uses the posted group sizes, means, and standard deviations to obtain an unadjusted standard error and test result. It assumes the group summaries correspond to the same outcome definition and analysis population, but the registry record does not establish that they match the sponsor’s analysis dataset.

How should the p-value be read?

In general, a p-value describes how unusual data at least as incompatible with a specified null hypothesis would be under the test’s assumptions, if the null hypothesis were true. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The recalculated unadjusted p-value is 0.623. Separately, the sponsor-reported analysis is identified as a two-sided t-test and reports a p-value of 0.008. Since those results disagree, they should not be merged or presented as if they came from one analysis.

The registry record does not report whether the sponsor’s test used a change score, an adjusted model, a different analysis population, or another specification. It also does not provide an analysis estimate or confidence interval for that test. Without those details, the sponsor result cannot be reproduced from the posted summaries, and the source of the discrepancy remains unknown. The sponsor’s own analysis takes precedence for the trial’s reported inferential result, while the recalculation is useful only as a transparent check on the summary statistics.

Why does variability matter?

In general, greater variability among individual outcomes makes it harder to estimate a group mean precisely when the sample size is unchanged. The CAPS standard deviations are 30.9 for topiramate and 33.76 for control, compared with a recalculated mean difference of -5.37. This context helps explain why the interval from the summary-based analysis is broad. It does not imply that the outcome is invalid; it indicates that individual scores vary considerably in relation to the observed separation of group averages.

The recalculation is described as unadjusted because it compares the groups without adjustment for other variables. Randomization is intended to support comparability between arms, but with a modest enrollment, chance differences in baseline characteristics can still occur. Baseline CAPS summaries and any baseline-adjusted analysis are not reported in the registry record. Consequently, the posted data do not show whether adjustment would change the estimate or its precision.

What can the registry record establish about analysis choices?

The record identifies the outcome as a mean of a continuous measure, gives a 12-week timeframe, and reports a sponsor t-test with two-sided testing. It does not report a detailed statistical analysis plan, the sponsor’s tested estimate, handling of missing endpoint values, analysis-set definition, or any adjustment for baseline score. The original sample-size justification, interim analyses, and multiplicity procedure are also not reported in the registry record. Those omissions matter because analysis choices can affect the estimate, uncertainty, and interpretation, especially when the sponsor result differs from a recalculation using the displayed summaries.

In general, a two-sided test allows evidence in either direction to count against the null hypothesis. A confidence interval complements such a test by showing a range of effect sizes compatible with the data under the analysis assumptions. The recalculated 95% interval runs from -26.79 to 16.05, and it includes zero. That interval and the recalculated p-value are internally consistent with a summary-based result that does not provide strong evidence against equal means under that particular calculation. They do not resolve the separate sponsor-reported result.

The outcome’s score categories provide descriptive context, but the posted analysis is based on mean total score rather than category counts. A mean may conceal variation in individual responses, including differences in the direction or size of change. The registry record does not report score distributions, baseline-to-follow-up changes, or the number of participants in each severity category at 12 weeks. Thus, interpretation should remain centered on the reported group means and the limits of the summary-based comparison, rather than treating the average as a complete description of all participants.

10. Confidence Intervals and Precision

The unadjusted mean difference is -5.37 CAPS points, with a 95% confidence interval from -26.79 to 16.05. The interval is wide relative to the point estimate and includes zero. In general, a confidence interval describes the precision of an estimate: a narrower interval indicates less sampling uncertainty, while a wider interval indicates more. The interval is compatible with a lower mean in the topiramate group, little or no difference, or a higher mean in that group under the assumptions of this recalculation.

The interval is based on an unadjusted analysis of the posted group summaries, not on the sponsor’s reported test. The sponsor analysis does not have a reported effect estimate or confidence interval in the registry record, so its precision cannot be assessed from the posted inferential result alone. The discrepancy between the sponsor analysis and the recalculation is an important limitation: a confidence interval from one analysis cannot be used to characterize the other.

The group sizes are 17 and 18, and the standard deviations are 30.9 and 33.76. In general, a small sample combined with substantial outcome variability can yield a broad confidence interval. The registry record does not report baseline scores, missing endpoint counts, or an adjusted model that might clarify the estimate. The posted confidence interval therefore describes uncertainty for the unadjusted comparison of the available summary values, rather than all possible analyses of the trial.

A confidence interval is not a guarantee that the true effect lies within its endpoints, and it does not summarize the probability that one treatment is better. It is a measure of statistical uncertainty conditional on the model and data used. Here, that conditional interpretation is particularly important because the sponsor’s reported p-value differs from the summary-based recalculation and the record does not explain why.

11. 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 are participants per arm, not total events. They are illustrations based on the observed effect, not the trial’s original design justification. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set at the stated fraction of the observed effect. The trial’s own original sample-size assumptions are not reported in the registry record.

The 80% power column gives the planned number of participants per arm for an illustrative design with an 80% chance of detecting the selected effect under the stated assumptions, if that effect is the true one and the model assumptions hold. The 90% power column gives the corresponding per-arm number for a design with a 90% chance. These are design-planning quantities, not probabilities that the completed trial has established an effect. They are not observed power.

The rows use planning for 100% of the observed effect, planning for 75% of the observed effect, and planning for 50% of the observed effect. For 100% of the observed effect, the figures are 572 participants per arm for 80% power and 766 participants per arm for 90% power. For 75% of the observed effect, they are 1,017 and 1,361 participants per arm, respectively. For 50% of the observed effect, they are 2,287 and 3,061 participants per arm, respectively.

Smaller effects require more participants to detect under the same planning assumptions, because a smaller separation is harder to distinguish from sampling variation. These figures are planning illustrations calculated from the observed effect, not the enrolled sample size or evidence about the trial’s actual power. They should not be read as a recommendation or as a reconstruction of the original protocol’s design decisions.

12. 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 = -5.370; pooled standard deviation σ = 32.405.
2. n = 2 × (1.96 + 0.842)² × 32.405² / -5.370² = 571.6, rounded up to 572 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 <- -5.37                      # observed difference in means
sigma <- 32.404872061672                     # 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(725920)
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 = 572 (80% power), 766 (90% power)
 75% of observed effect: n per arm = 1,017 (80% power), 1,361 (90% power)
 50% of observed effect: n per arm = 2,287 (80% power), 3,061 (90% power)
power.t.test: 572.6 
Simulated power at n = 572 per arm: 0.798 (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.

13. Analyses Reported by the Sponsor

The registry reports the sponsor’s analysis as a two-sided t-test comparing the groups and gives a p-value of 0.008. It does not provide the sponsor analysis’s estimated effect, confidence interval, analysis population, or additional model details. The posted group summaries yield a separate unadjusted recalculation: a mean difference of -5.37, a 95% confidence interval from -26.79 to 16.05, and a p-value of 0.623.

These results disagree. The recalculation is a simple check based on the displayed means, standard deviations, and group sizes; it is not a correction or replacement for the sponsor’s analysis. Because the discrepancy is flagged, the sponsor’s own analysis takes precedence when describing the trial’s reported inferential result. At the same time, the missing estimate and analysis details mean that the reported test cannot be independently reconciled with the summary-based calculation from the registry record.

The registry record does not explain whether the two analyses differ because of outcome handling, data, an adjustment, or some other specification. None of those possible explanations can be assumed. The appropriate reading is to report both results with their distinct sources and limitations, and to avoid treating the recalculated result as if it were the sponsor’s primary analysis.

14. Multiplicity, Missing Data and Analysis Populations

The registry record describes several assessment domains, including PTSD symptoms, depression, anxiety, quality of life, and social adjustment. It posts one outcome measure, the total CAPS score, and no secondary outcome results. The record does not report a multiplicity procedure or explain how the several mentioned domains were ranked or analyzed. Without that information, it is not possible to determine whether the sponsor’s reported test was part of a larger family of planned comparisons or how any multiplicity was handled.

Missing-data methods are not reported in the registry record. The posted group sizes for the primary outcome are 17 and 18, matching the reported randomized arm counts, but those counts alone do not establish whether every randomized participant contributed an observed 12-week CAPS score or whether missing values were imputed. The record does not describe withdrawals, exclusions, or the handling of participants without an endpoint measurement.

The analysis population is also not defined in the registry record. It is therefore unclear whether the sponsor’s result is based on all randomized participants, participants who completed treatment, participants with available endpoint data, or another set. The registry does not provide a participant-flow account or explain how protocol deviations and treatment discontinuation were handled.

The original statistical analysis plan, interim analyses, and adjustment choices are not reported in the registry record. These omissions limit assessment of whether the posted result follows a prespecified plan, how the multiple stated outcomes were treated, and whether missingness could affect the comparison. The discrepancy between the sponsor’s test and the unadjusted recalculation makes the absence of those details especially relevant, but it does not establish a particular cause for the disagreement.

15. 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 no serious adverse events among participants at risk in either group: 0 of 17 in the topiramate group and 0 of 18 in the control group. This is a count of participants affected by a serious adverse event in the posted record. No serious adverse-event terms are listed. Death counts are not reported in the registry record.

The record also lists other, non-serious adverse events affecting 12 participants in the topiramate group and 11 in the control group. These counts are separate from the serious-event figures and should not be read as counts of serious events. The registry does not provide event descriptions, severity details, timing, or information about whether individual participants experienced more than one event.

In general, an absence of recorded serious events in a small trial does not establish that serious events cannot occur. These counts describe what is posted for this study and do not provide a complete assessment of safety beyond the reported follow-up and data. The registry record does not report additional safety analyses or the method used to collect adverse-event information.

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

The posted primary outcome compares average total CAPS scores between randomized groups at 12 weeks. The observed means were 30.41 for topiramate and 35.78 for control. A summary-based, unadjusted calculation estimates a mean difference of -5.37 points, with a 95% confidence interval from -26.79 to 16.05 and an unadjusted p-value of 0.623. This estimate is uncertain and the interval includes zero.

The registry also reports a sponsor-run two-sided t-test with p = 0.008. That result differs from the recalculation using the posted group summaries. The sponsor’s own analysis takes precedence for the trial’s reported inferential result, but the registry record does not provide enough information to reproduce or explain it. The two findings should therefore be kept distinct rather than blended into one conclusion.

Neither the means nor the p-values describe every participant’s outcome, and the mean difference does not show how many participants improved, worsened, or remained unchanged. The record does not report baseline CAPS values, individual changes, endpoint missingness, or analysis-population details. It also does not post results for the other outcome domains described in the study summary.

Statistical results alone do not establish the practical or clinical importance of a difference. The registry’s CAPS score categories provide a framework for describing score levels, but the posted results do not report category transitions or participant-level responses. The evidence available in the registry record is limited to the stated group summaries and the sponsor-reported test, with an unresolved discrepancy between them. It should be interpreted as a trial-specific statistical record, not as medical advice or a definitive statement about treatment effects in other settings or populations.

17. Why This Trial Matters Statistically

This trial illustrates why a reported p-value should be read alongside the effect estimate, its uncertainty, and the analysis details. The posted group means and standard deviations allow an unadjusted comparison, but that recalculation does not match the sponsor’s reported test. When an analysis discrepancy is present, it is important to distinguish the sponsor’s result from a calculation based on summary statistics, rather than silently choosing one or presenting them as equivalent.

The wide confidence interval for the recalculated mean difference also shows why the point estimate alone is not enough. A single estimate can suggest a direction, but the interval communicates that the summary-based data are compatible with a range of differences. In general, precision depends on the amount of information in the data and the variability of the outcome. The posted arm sizes and standard deviations provide context for the broad uncertainty, while the missing analysis details limit further assessment.

The record is also a useful reminder that a trial description can mention several outcome domains while posting results for only one measure. Multiplicity procedures, missing-data handling, the analysis population, and the original sample-size justification are all relevant to evaluating a clinical trial’s statistical evidence, yet they are not reported in the registry record.

Finally, the planning illustrations show how required sample size depends on the effect chosen for planning and the desired power, under stated assumptions. They are not observed power and do not reconstruct the trial’s own design. Read together, the outcome summaries, sponsor analysis, recalculation, and unreported methods demonstrate the importance of transparent estimates and analysis reporting.

18. 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 trial enrolled 35 participants, with 17 assigned to topiramate and 18 to control. This limited enrollment is important when interpreting the precision of group comparisons, particularly because the posted CAPS standard deviations are 30.9 and 33.76. The unadjusted confidence interval is broad. The registry record does not report the original sample-size justification, so it is not possible to assess the planned precision or assumptions behind the trial’s design.

The primary outcome is reported at 12 weeks, but baseline CAPS summaries and participant-level changes are not posted. The registry record does not report missing endpoint counts, the analysis population, or methods for handling missing data. It also does not give information about adherence, dose attained, treatment discontinuation, or protocol deviations. Those omissions constrain interpretation of the group summaries.

A further limitation is the difference between the sponsor’s reported p-value and the recalculation based on the posted group means and standard deviations. The sponsor’s analysis takes precedence, but its estimate, confidence interval, and detailed specification are not reported. The reason for the discrepancy is unknown from the registry record.

Although the study description names depression, anxiety, quality of life, and social adjustment outcomes, no secondary results are posted. The registry also does not report a multiplicity plan, interim analyses, or a complete statistical analysis plan. Finally, adverse-event reporting is limited: the record gives affected-participant counts, but does not report deaths, serious-event terms, event timing, or detailed event descriptions. These limitations apply to interpretation of the registry record and do not establish what was or was not collected outside the posted results.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Placebo Controled Clinical Trial Using Topiramate To Treat Posttraumatic Stress Disorder (PTSD) Patients. — NCT00725920.
• Registry record last updated 2014-08-26; status last verified 2014-08.
• Results first posted 2014-08-26.
• 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.

22. Record Summary

The registry describes a completed, randomized, triple-masked, parallel-group phase 4 trial of topiramate versus placebo for posttraumatic stress disorder. Enrollment was 35, with 17 participants in the topiramate group and 18 in the control group. The planned treatment period and primary assessment timeframe were 12 weeks.

The primary outcome was mean total CAPS score. The posted group means were 30.41 for topiramate and 35.78 for control. A simple unadjusted recalculation gives a mean difference of -5.37 and a 95% confidence interval from -26.79 to 16.05. The sponsor-reported two-sided t-test gives a different result, and the registry record does not explain the discrepancy; the sponsor’s own analysis takes precedence.

No secondary outcome results are posted. The original sample-size justification, missing-data methods, analysis population, and detailed statistical plan are not reported in the registry record. The safety record lists no serious adverse events in either group and does not report death counts.