A Placebo-Controlled Study of Mirtazapine for PTSD: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Placebo-Controlled Study of Mirtazapine for PTSD (NCT00302107), a randomized phase 4 study. The primary outcome is Structured Interview of Posttraumatic Stress Disorder (SIP).

2. Clinical Question

3. Trial Design

The registry record describes a randomized, parallel-group, phase 4 trial comparing mirtazapine with placebo. The study summary calls it an 8-week, double-blind, placebo-controlled treatment trial. The registry’s masking field lists quadruple masking, while the summary describes double blinding and look-alike tablets. The record does not explain which roles were masked or how the two masking descriptions relate, so the masking details cannot be reconciled from the registry record alone.

The stated objective was to study the efficacy and tolerability of mirtazapine for PTSD. Participants were randomized to mirtazapine or placebo after meeting inclusion and exclusion criteria. The summary says symptoms, side effects, and compliance were assessed bi-weekly; pill counts were planned at week 4 and week 8. It also describes supportive clinical management during visits and the availability of an investigator by telephone for emergencies. The study description allows for more frequent visits if needed.

The registry reports enrollment of 78 and status as completed. The recorded start is 2006-04, and primary completion is 2011-05. The sponsor is the US Department of Veterans Affairs. The trial is listed under Anxiety and PTSD. The registry record does not report the trial’s original sample-size assumptions or explain how enrollment was selected.

The primary endpoint is a continuous measure, and the posted results are group means and standard deviations for change scores. The mirtazapine group has a reported mean change of -7.61 with standard deviation 9.89; the placebo group has a reported mean change of -5.83 with standard deviation 7.41. The reported comparison is unadjusted. The registry record does not report a statistical analysis plan or analysis details beyond the posted outcome information.

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: Prior to randomization, participants signed informed consent, were assessed to have met eligibility criteria including free of psychotropic medications for the previous two weeks (4 weeks for fluoxetine) prior to randomization.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Randomized, took at least one dose of study medication, and returned for at least one visit post-randomization

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.

Population: Seventy-three (94%) were male; mean age 38.1 +/- 9.5 (SD) years in placebo group and 37.7 +/- 10.3 (SD) in mirtazapine group.

6. Primary Endpoint

The registry lists the primary outcome as Structured Interview of Posttraumatic Stress Disorder (SIP), measured over Primary outcome is measured at baseline and week 8 (primary endpoint) with primary outcome change scores calculated as week 8 minus baseline score..

The primary endpoint is change in the Structured Interview of Posttraumatic Stress Disorder (SIP) score from baseline to week 8. The score is defined as week 8 minus baseline, so a negative change indicates a lower score at week 8 than at baseline. The SIP is described as a clinician-administered, 17-item scale for PTSD based on Diagnostic Statistical Manual-IV criteria. Its total score ranges from 0 to 68, with higher scores indicating more symptoms or a worse outcome. Individual items are scored from 0 to 4, and an item counts as positive when its score is at least 2, described as moderate.

A clinician-administered scale provides a structured way to assess symptoms using the same instrument across participants. The registry description gives reliability and validity information for the SIP, including comparisons with other measures and a diagnostic interview. Those properties help explain why the instrument was selected, but they do not establish that any particular change is clinically important for this trial. The registry record does not report a threshold for clinically meaningful change.

The posted group means are change scores rather than week 8 scores alone. The mean change was -7.61 in the mirtazapine group and -5.83 in the placebo group. The reported mean difference, calculated as mirtazapine minus placebo, is -1.78 units. These summaries describe average score changes; they do not show how each participant’s score changed or whether changes were similar across participants.

7. Results Posted in the Registry

• Mirtazapine: mean -7.61 versus -5.83 on control; difference -1.78 (95% CI -5.82 to 2.26), standardized effect d = -0.20; z-test p = 0.388.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: it is a numerical score, and the registry reports a mean and standard deviation for each treatment group. The outcome is not the score at a single time alone, but the change from baseline to week 8. The stated calculation is week 8 minus baseline. A negative change means the measured score is lower at week 8 than at baseline. Since higher SIP scores indicate more symptoms, a decrease corresponds to a lower symptom score on this scale.

What does the mean difference represent?

The posted mean difference is -1.78 units, with the comparison defined as mirtazapine minus placebo. This is a difference between group averages of change scores. It is not an average of individual treatment effects, and it does not imply that every participant’s change was close to the group mean. The posted group means are -7.61 for mirtazapine and -5.83 for placebo. The comparison condenses those group summaries into one estimate of separation on the SIP scale.

What does the standard deviation describe?

The standard deviation describes how dispersed participant scores are around the group mean. The reported standard deviations are 9.89 for mirtazapine and 7.41 for placebo. A standard deviation is not the uncertainty interval for the group mean and is not a measure of the estimated difference between groups. It reflects variation in the observed change scores within each group. Greater dispersion generally makes it harder to estimate a group average or between-group contrast precisely, because the signal of a mean difference must be distinguished from the variation among participants.

General statistical explanation: the standard error describes the sampling uncertainty of an estimate, whereas the standard deviation describes variability among observations. The posted record gives standard errors for the group means as 1.626 for mirtazapine and 1.271 for placebo. The reported confidence interval for the between-group mean difference is the relevant interval for judging uncertainty in that contrast; the individual group standard errors should not be substituted for it.

How should the confidence interval be read?

The 95% confidence interval for the reported mean difference runs from -5.825 to 2.265. In general, a confidence interval represents a range of values reasonably compatible with the data and the method used to calculate the interval. Repeated use of the same procedure across comparable studies would produce intervals that cover the true value at the stated rate, under the method’s assumptions. For this result, the interval spans zero, so the data do not pin down a single direction for the difference. The interval also conveys uncertainty in the magnitude: it is not just a yes-or-no test result.

The unadjusted p-value is 0.388. General statistical explanation: a p-value is calculated under a reference model, often a model of no difference, and summarizes how unusual a result at least as extreme as the observed one would be under that model. It is not the probability that the treatment effect is zero and not a direct measure of clinical importance. The confidence interval and p-value are complementary summaries of the comparison, but neither establishes causation beyond the design and assumptions that support the analysis.

Why can over-dispersion matter?

General explanation: over-dispersion occurs when observed variability is greater than a model assumes. If a statistical analysis treats the data as less variable than they are, its standard errors may be too small and its confidence intervals too narrow. This outcome is a continuous score analyzed through group means and standard deviations, and the registry record does not report an over-dispersion assessment or a model-based adjustment. Over-dispersion should therefore be understood as a general issue in statistical modeling, not as a reported feature of this trial.

The comparison is described as unadjusted. That means the posted contrast is not reported as adjusted for other variables. The registry record does not state whether covariate-adjusted analyses were planned or performed, does not provide an analysis model, and does not give details of how the confidence interval or p-value was calculated. The reported mean difference, interval, and p-value are the available inferential summaries. No calculations beyond those posted are needed to interpret their basic meaning.

The outcome population is defined as randomized participants who took at least one dose of study medication and returned for at least one visit after randomization. This definition describes who was included in the posted outcome population; it is not the same as an explicitly reported intention-to-treat analysis. The registry record does not report how missing baseline or week 8 scores were handled, whether any participants were excluded after randomization, or whether the primary analysis used an imputation method. Those details matter because missing outcomes can affect the population represented by the estimated means and the assumptions supporting the comparison.

Finally, a statistically uncertain estimate is not proof of no difference. The point estimate is one estimate from the recorded data, while the confidence interval shows that a range of different underlying differences remains compatible with the analysis. The p-value does not replace consideration of the direction, scale, variability, outcome population, and missing-data information. Here, the registry record’s lack of further analysis detail limits how far the inferential interpretation can go.

10. Confidence Intervals and Precision

The reported mean difference in SIP change is -1.78 units for mirtazapine minus placebo. Its 95% confidence interval ranges from -5.825 to 2.265. The interval is the principal summary of precision for the between-group estimate: it indicates that the observed data do not identify the difference narrowly. Because the interval includes zero, it is compatible with no difference as well as with differences in either direction under the comparison’s sign convention.

The negative end of the interval represents a larger decrease in the mirtazapine group, while the positive end represents a larger decrease in the placebo group. The point estimate lies on the negative side, but the range extends across zero. The interval therefore cautions against treating the point estimate as a settled direction or as a precise estimate of magnitude. It does not mean that every value in the interval is equally likely, nor does it establish that effects outside the interval are impossible.

Precision is distinct from clinical importance. The registry record does not report a threshold for a clinically meaningful difference on the SIP. It also does not report a prespecified precision target or the original sample-size justification. The unadjusted p-value of 0.388 is consistent with the interval’s inclusion of zero, but a p-value alone does not express the range or clinical relevance of plausible effects.

Variation in change scores is reflected by the reported standard deviations, 9.89 for mirtazapine and 7.41 for placebo. The group outcome counts are 37 and 34, respectively. These summaries provide context for the estimate, but do not resolve uncertainty about missing outcomes or the analysis population. The registry record does not report sensitivity analyses that would show how the interval might change under alternative assumptions.

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 illustrations based on the observed effect, not a report of the trial’s original design assumptions. For these continuous outcomes, the figures are participants per arm, not total events. The stated 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 as a stated fraction of the observed effect.

The 80% power column gives the planned participants per arm for a design that would have an 80% chance of detecting the chosen effect under the stated assumptions, if that effect were the true one. The 90% power column gives the corresponding planning figure for a 90% chance. These are design probabilities under assumptions, not guarantees about a future result. This is not observed power.

For planning at 100% of the observed effect, the listed figures are 383 participants per arm for 80% power and 513 participants per arm for 90% power. Planning for 75% of the observed effect gives 681 and 912 participants per arm, respectively. Planning for 50% of the observed effect gives 1,532 and 2,051 participants per arm, respectively. The smaller fractions represent smaller effects to detect. Smaller effects generally require more participants to distinguish them from random variation, which is why the listed planning figures increase as the target effect decreases.

These calculations are based on the observed effect and should not be read as a recommendation or as evidence about the trial’s original design. The registry record does not report the original sample-size justification or assumptions. The planning figures also do not account for any unreported features of the trial design or outcome-data availability beyond the stated assumptions.

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 = -1.780; pooled standard deviation σ = 8.792.
2. n = 2 × (1.96 + 0.842)² × 8.792² / -1.780² = 382.9, rounded up to 383 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 <- -1.78                      # observed difference in means
sigma <- 8.7916321381285                     # 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(302107)
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 = 383 (80% power), 513 (90% power)
 75% of observed effect: n per arm = 681 (80% power), 912 (90% power)
 50% of observed effect: n per arm = 1,532 (80% power), 2,051 (90% power)
power.t.test: 383.9 
Simulated power at n = 383 per arm: 0.797 (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. Multiplicity, Missing Data and Analysis Populations

The registry lists one posted outcome measure and no posted secondary outcomes. It lists no analyses. The primary result is reported with an unadjusted p-value of 0.388. The registry record does not report a multiplicity adjustment, a hierarchy of outcomes, or a plan for handling multiple statistical tests. With no secondary results posted and no analysis entries, it is not possible to determine whether additional analyses were conducted or how multiple testing was addressed.

The registry record also does not report a full statistical analysis plan. In particular, it does not state whether the primary analysis was prespecified in the form reported, which covariates might have been considered, or how the confidence interval and test were constructed. The available result is therefore best described as the posted unadjusted comparison rather than as a complete account of the study’s statistical analysis.

The outcome population is described as randomized participants who took at least one dose of study medication and returned for at least one post-randomization visit. That is the stated population for the primary outcome. The record does not say whether all randomized participants were included in the final analysis, whether participants without an eligible post-randomization visit were excluded, or how the analysis population relates to an intention-to-treat principle.

Missing-data methods are not reported in the registry record. There is no description of how missing baseline or week 8 SIP scores were handled, whether scores were imputed, or whether sensitivity analyses assessed the effect of missing outcomes. The registry record also does not report reasons for missing outcome data. Without those details, the extent to which the posted estimate depends on assumptions about unavailable scores cannot be assessed from the registry record.

These omissions do not change the numerical result that is posted, but they limit assessment of its robustness and the scope of inference. They also prevent a conclusion about whether the recorded outcome population fully represents everyone randomized. The count of 78 enrolled and the outcome-group counts are reported separately; the registry record does not explain their relationship.

14. 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 0% reporting threshold.

Most frequent non-serious adverse event terms

The posted serious adverse event figures report 0 affected participants among 39 at risk in the mirtazapine group and 0 among 39 at risk in the placebo group. The record lists no serious adverse event terms. These figures describe the registry’s recorded serious adverse event counts for the stated groups; they do not provide detail about event definitions, duration of observation, or how events were collected.

For other adverse events, the record reports 22 affected participants among 39 at risk with mirtazapine and 24 among 39 at risk with placebo. These are participant counts, not counts of individual events, and the record does not list the specific other adverse events in this summary. The posted figures alone do not indicate whether events were mild or severe, how long they lasted, or whether they led to treatment changes.

Deaths are not reported in the registry record. The adverse event counts should be interpreted as recorded safety summaries rather than as a complete assessment of all possible harms. They do not establish comparative safety beyond the reported categories, participant counts, and period covered by the registry data.

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

The result describes the average change in SIP score from baseline to week 8 in the outcome population recorded for the mirtazapine and placebo groups. The estimated mean difference is -1.78 units, with mirtazapine minus placebo as the comparison order. The confidence interval extends from -5.825 to 2.265, and the unadjusted p-value is 0.388. The interval includes zero and permits differences in either direction, so the posted data do not give a precise estimate of the direction or size of the between-group difference.

The result does not show that the groups were identical. A non-significant test result is not proof of equivalence, and the registry record does not report an equivalence margin or non-inferiority margin. Nor does the observed mean change alone establish how individual participants changed. The group averages and standard deviations describe aggregate results, not individual responses.

The outcome is a clinician-administered PTSD symptom scale with higher values indicating more symptoms. The score’s direction helps interpret negative change, but the registry record does not provide a threshold for a clinically meaningful change. Statistical uncertainty and clinical importance are related but distinct questions; the posted result does not answer the latter with a stated threshold.

The findings are limited to the primary SIP change endpoint and the outcome population defined in the record. No secondary outcomes or analyses are posted. The record does not report missing-data handling, an adjusted analysis, the original sample-size rationale, or a full statistical analysis plan. These gaps restrict assessment of robustness and generalizability. The result should therefore be understood as the recorded unadjusted comparison for this endpoint, rather than as a comprehensive account of all outcomes or all possible effects.

16. Why This Trial Matters Statistically

This trial illustrates why an estimated difference should be read together with its uncertainty. The mean difference is -1.78 units, but the confidence interval extends from -5.825 to 2.265. The point estimate gives a concise summary of the observed separation in group mean changes; the interval shows that the data leave substantial uncertainty about both direction and magnitude. A single estimate should not be mistaken for a precise or definitive account of the underlying difference.

The distinction between within-group change and between-group comparison is also important. Both reported group means are negative, but the primary comparative question concerns the difference between those averages, not whether either group changed from baseline in isolation. Changes over time within a group do not, by themselves, establish that the groups differ. The placebo group provides the comparator needed for the stated contrast.

The SIP is a continuous scale, and the report gives means and standard deviations. That makes the numerical interpretation accessible in scale units, while also highlighting the importance of variability and outcome measurement. The standard deviations of 9.89 and 7.41 provide context for the range of observed participant changes around the means. They are not substitutes for the confidence interval on the group difference.

Finally, the registry record demonstrates the limits of interpreting a result without accompanying analysis details. The primary outcome is posted, but no secondary outcomes or analyses are listed. Missing-data methods, multiplicity handling, the original sample-size rationale, and a full analysis plan are not reported in the registry record. These details matter to assessing robustness, not just to reading a p-value. A careful interpretation keeps the conclusion tied to what is recorded while recognizing what the record does not establish.

17. 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 registry record leaves several trial-specific details unresolved. It reports enrollment of 78 and outcome groups of 37 for mirtazapine and 34 for placebo, but does not explain the relationship between those counts. The outcome population is restricted in its description to randomized participants who took at least one dose and returned for at least one post-randomization visit. The record does not state whether all other randomized participants were included in any analysis or why they might not be represented in the posted outcome groups.

The study summary describes double blinding, while the registry masking field lists quadruple masking. The record does not identify which participants, investigators, assessors, or other roles were masked, or clarify the difference between the descriptions. The original sample-size justification is also not reported, so the planning illustration based on the observed effect cannot be treated as the trial’s design rationale.

The registry record does not describe missing-data handling, sensitivity analyses, covariate adjustment, multiplicity control, or a full statistical analysis plan. No secondary outcomes or analyses are posted. These gaps limit assessment of the primary result’s robustness and prevent a fuller account of the trial’s statistical approach.

The primary outcome is a symptom scale, but the record does not report a threshold for clinically meaningful change. The posted group means do not reveal the distribution of individual changes, and the safety summary does not provide details of the other adverse events. Deaths are not reported in the registry record. These limitations concern the information available for interpretation and do not replace the results that are posted.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: A Placebo-Controlled Study of Mirtazapine for PTSD — NCT00302107.
• Registry record last updated 2014-11-26; status last verified 2014-11.
• Results first posted 2014-11-26.
• 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.

21. Record Summary

The registry record identifies a completed, randomized, parallel-group phase 4 study of mirtazapine versus placebo for PTSD in OIF/OEF veterans. The study summary describes an 8-week treatment period. The sponsor is the US Department of Veterans Affairs, and the record lists Anxiety and PTSD as conditions.

The primary outcome is week 8 minus baseline change in SIP score. The posted mean changes are -7.61 for mirtazapine and -5.83 for placebo; the mean difference is -1.78, with a confidence interval from -5.825 to 2.265 and an unadjusted p-value of 0.388. No secondary outcomes or analyses are posted. The registry record does not report the original sample-size justification, missing-data methods, or a full statistical analysis plan.