Comparing Virtual Reality Exposure Therapy to Prolonged Exposure: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Comparing Virtual Reality Exposure Therapy to Prolonged Exposure (NCT01459705), a randomized na study. The primary outcome is Clinician-Administered PTSD Scale (CAPS).

2. Clinical Question

3. Trial Design

The registry record describes a randomized, single-masked, parallel-group behavioral study with three arms: prolonged exposure therapy (PE), virtual reality exposure therapy (VRET) and waitlist. The reported enrollment is 162, with 54 participants in each arm. The study status is completed. The start is listed as 2011-10 and primary completion as 2014-05. The phase is listed as NA.

PE is described as following manualized procedures based on activating the fear structure associated with PTSD and pairing it with incompatible new information. VRET follows the same procedures, with imaginal exposure augmented by immersion in Virtual Iraq environments. Waitlist participants were asked to refrain from psychotherapy during the 5 weeks of study participation. These descriptions define the assigned interventions; the registry record does not report adherence, treatment completion, or the amount of therapy received by each participant.

The study summary describes a general hypothesis that 10 sessions of VRET or PE would treat PTSD and affect physiological arousal and perceptions of stigma toward behavioral health services. This is a statement of the study's hypothesis, not a reported result. The posted primary outcome is CAPS total score at the screening visit on Day 1, based on baseline scores with a last-week reference period. It is therefore a baseline measure rather than a post-treatment assessment.

The registry lists secondary measures covering PTSD symptoms, depression, attitudes toward mental health services, perceived stigma, suicide risk, anxiety and general behavioral symptoms. It reports 87 outcome measures and 82 posted secondary outcome measures. The registry record does not report who was masked, the trial's original sample-size justification, or the details of its allocation procedure beyond randomization.

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

Analysis population for the primary outcome, as defined in the registry: Baseline scores on the CAPS-W (last week reference)

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: All participants randomized to a treatment arm included in baseline data analysis

6. Primary Endpoint

The registry lists the primary outcome as Clinician-Administered PTSD Scale (CAPS), measured over Screening Visit (Day 1).

The primary endpoint is the mean total score on the Clinician-Administered PTSD Scale (CAPS), measured at the screening visit on Day 1. The registry describes the CAPS as a structured interview assessing PTSD diagnostic criteria through the frequency and intensity of symptoms. Total scores were used, with a possible range from 0 to 136; higher scores indicate higher symptom levels. The population description identifies baseline CAPS-W scores using a last-week reference.

A structured symptom measure can provide a consistent way to characterize PTSD symptoms across participants and groups. Its use as a baseline measure is useful for describing the groups' starting symptom levels and for assessing whether their reported mean scores were similar at screening. The registry also lists symptom and mental-health-related measures among the secondary outcomes.

The timing is essential to interpretation. The posted primary outcome is recorded at screening, not after the therapy period. Consequently, these means and between-group comparisons cannot establish symptom improvement, treatment response, or a difference in treatment effects. The registry record does not report a post-treatment CAPS result as the primary outcome in the information summarized for this comparison. The reported baseline scores should be read as a description of participants at entry, not as evidence about the efficacy of PE or VRET.

7. Results Posted in the Registry

• Virtual Reality Exposure Therapy (VRET): mean 80.44 versus 78.28 on control; difference 2.16 (95% CI -3.98 to 8.30), standardized effect d = 0.13; z-test p = 0.491.
• Waitlist: mean 78.89 versus 78.28 on control; difference 0.61 (95% CI -5.66 to 6.88), standardized effect d = 0.04; z-test p = 0.849.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: each participant's CAPS result is a score, and the arm summaries report means and standard deviations. A mean is the arithmetic average of the scores in a group. A standard deviation describes how spread out individual scores are around that group mean. A standard deviation is not the same as a standard error: the standard error describes uncertainty in an estimated mean, and it tends to be smaller when a group includes more participants, all else equal.

What does a mean difference represent?

A mean difference subtracts the reference group's mean from the comparison group's mean. Positive values in these reported comparisons mean that the comparison arm's baseline mean was higher than the PE mean; negative values would mean it was lower. The VRET mean difference is 2.16 units, based on means of 80.44 for VRET and 78.28 for PE. The waitlist mean difference is 0.61 units, based on means of 78.89 and 78.28. These are unadjusted comparisons of the posted summary statistics, not estimates of changes during treatment.

Because the outcome is measured at screening, an unadjusted difference describes the observed baseline contrast between assigned groups. Randomization is intended to create comparable groups on average, but it does not make every observed baseline measure exactly equal. A baseline contrast can therefore be present by chance. These values alone do not establish that a baseline difference caused any later outcome, and they do not estimate the effect of an intervention.

How should the confidence intervals be understood?

A confidence interval communicates uncertainty around an estimated difference. A wider interval indicates less precision than a narrower interval. The intervals reported for both comparisons include zero, the value corresponding to no mean difference. They also extend to positive and negative values, so the data are compatible with a range of possible baseline contrasts. The intervals should not be read as the range of individual participant scores or as the probability that the true difference falls within the particular interval.

For VRET compared with PE, the interval is -3.985 to 8.305. For waitlist compared with PE, it is -5.656 to 6.876. These intervals concern baseline mean differences. They do not quantify uncertainty in a treatment effect, because the endpoint is not a post-treatment score or a change score.

What does the p-value add?

A p-value describes how unusual results at least as inconsistent with a specified null model would be if that model were correct. It is not the probability that the null hypothesis is true, nor is it a measure of the clinical importance of a difference. The unadjusted p-values are 0.491 for VRET versus PE and 0.849 for waitlist versus PE. In conjunction with the intervals, they do not give strong evidence of a baseline mean difference. They also do not prove equivalence or show that the treatment groups have the same outcomes after therapy.

The word unadjusted matters: these comparisons are not reported as adjusted for other variables or for multiple testing. The registry record does not report the analysis model used in the trial's own primary analysis, any covariate adjustments, or a multiplicity procedure. The recalculated values are therefore best treated as straightforward comparisons based on the posted means, standard deviations and group sizes.

Why does variation matter?

People in the same group can have different CAPS scores. The reported standard deviations—16.35 for PE, 16.23 for VRET and 16.87 for waitlist—describe that variation. When there is substantial within-group variation relative to a group difference, it can be difficult to estimate the difference precisely. Confidence intervals capture uncertainty in the group contrast, while standard deviations describe variation among participants. Keeping those roles separate prevents an individual-level spread from being mistaken for the precision of a mean difference.

These summaries assume the posted means and standard deviations adequately describe the continuous outcome for the comparison. The registry record does not report diagnostic checks, transformations, or the detailed statistical model. No additional distributional assumptions or analysis choices should be inferred from the summaries alone.

What can these analyses answer?

They can describe baseline CAPS levels and quantify the unadjusted differences between the assigned groups at screening. They cannot answer whether participants' symptoms changed, whether PE or VRET altered symptoms, or whether the therapies differed in their effects. Those questions require outcome measurements at relevant follow-up times and an analysis comparing outcomes or changes. The registry record does not report those results in the primary-outcome information summarized for these comparisons.

10. Confidence Intervals and Precision

The precision of the baseline comparisons is expressed by the confidence intervals around the mean differences. For VRET versus PE, the estimated difference is 2.16 units and the interval ranges from -3.985 to 8.305. For waitlist versus PE, the estimated difference is 0.61 units and the interval ranges from -5.656 to 6.876. Both intervals include zero, and both allow differences in either direction.

The interval width conveys that the estimates do not pin down a single precise baseline contrast. The number at the center is the best estimate from the posted group summaries, while the interval indicates the range of values compatible with the data under the interval's statistical assumptions. It is not a range containing most individual CAPS scores, and it does not mean that every value in the interval is equally plausible.

Both intervals also need to be interpreted in light of timing: the outcome is a screening score, not a treatment response. Their precision concerns differences in baseline means only. They cannot be used to claim that a treatment benefit is small, large, absent or precisely estimated.

The registry record does not report a clinically important threshold for the CAPS difference or an equivalence margin. Without such a prespecified benchmark, an interval that includes zero cannot establish that the groups are equivalent. It says that the observed data do not distinguish a baseline mean difference from no difference at conventional levels of statistical evidence, while leaving uncertainty about the size and direction of the contrast.

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 sample-size figures are planning illustrations based on the observed mean differences, not the trial's original design calculation. For these continuous-outcome comparisons, the figures are participants per arm, not total events. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms and a normal approximation. Each row uses a stated fraction of the observed effect as the effect to detect. 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 under those assumptions for an analysis to have an 80% chance of detecting the specified effect if that effect is real. The 90% power column gives the corresponding planning target for a 90% chance. These are design targets, not guarantees about a future result and not measures of the probability that a treatment works. The figures are not observed power.

The illustration makes the dependence on effect size clear: planning for a smaller effect requires more participants. For VRET compared with PE, planning for 100% of the observed effect gives 893 participants per arm for 80% power; planning for 75% gives 1,588, and planning for 50% gives 3,572. The corresponding 90% figures are 1,196, 2,125 and 4,782 participants per arm. The waitlist comparison has a smaller observed mean difference and consequently much larger planning figures, including 11,642 participants per arm for 80% power when planning for 100% of that observed effect.

The curve should be read as a visual representation of the same planning relationship, not as a separate result about the enrolled trial. These illustrations do not change the fact that the endpoint is a baseline measure. They do not show that the completed trial had adequate power for a treatment-effect question, and the registry record does not report its own design justification.

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 = 2.160; pooled standard deviation σ = 16.290.
2. n = 2 × (1.96 + 0.842)² × 16.290² / 2.160² = 892.8, rounded up to 893 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 <- 2.16                      # observed difference in means
sigma <- 16.2901104968628                     # 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(1459705)
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 = 893 (80% power), 1,196 (90% power)
 75% of observed effect: n per arm = 1,588 (80% power), 2,125 (90% power)
 50% of observed effect: n per arm = 3,572 (80% power), 4,782 (90% power)
power.t.test: 893.8 
Simulated power at n = 893 per arm: 0.802 (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 record lists 87 outcome measures and 82 posted secondary outcome measures, alongside 5 analyses. The secondary measures span PTSD symptoms, depression, attitudes toward seeking mental health services, perceived stigma, suicide risk, anxiety and behavioral symptoms. When multiple outcomes or comparisons are examined, the chance of obtaining at least one apparently unusual result can increase. Statistical approaches can account for this, but the registry record does not report a multiplicity adjustment or specify which outcomes were treated as confirmatory.

The two baseline comparisons described for CAPS use unadjusted p-values. No multiplicity-adjusted values are reported. These comparisons are also not post-treatment efficacy tests, so they should not be treated as evidence from a family of treatment-effect analyses. The registry record does not report a hierarchy of outcomes, a prespecified testing sequence, or interim analyses.

Missing-data methods and analysis populations are also not reported in the registry record for the endpoint. The arm summaries list 54 participants per group, but those counts alone do not establish whether every randomized participant contributed a CAPS score, how missing observations were handled, or whether the analysis used an intention-to-treat population. No conclusions about attrition, exclusions or imputation should be inferred from the group counts.

The record also does not report the trial's detailed statistical analysis plan for this endpoint, including any adjustment for baseline covariates. The displayed recalculated comparisons are unadjusted summaries. They are useful for describing the posted data, but they cannot be assumed to reproduce every feature of the sponsor's original analysis. More detail on prespecification, missingness and analysis populations would be needed to assess those aspects of the study.

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 serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: hospitalization - psychiatric (5.6%).

The registry record reports serious adverse events affecting 0 participants in PE, 3 in VRET and 0 in waitlist. These are counts among 54 participants at risk in each group. The record identifies “Hospitalization - psychiatric” as a leading serious adverse-event term and lists 5.6%; it does not attribute that percentage to a particular arm in the information. The registry record does not report deaths.

Other adverse events are reported as affecting 12 participants in PE, 15 in VRET and 0 in waitlist, with 54 at risk in each group. These figures are distinct from the serious adverse-event counts. The posted summaries do not establish the events' causes, timing, severity beyond the serious classification, or relationship to the assigned intervention. A count of affected participants is also not a complete description of event duration or clinical course.

These descriptive counts should be interpreted in light of the equal reported group sizes and the absence of further detail about event definitions and ascertainment. The registry record does not report additional analyses that would establish comparative safety. The figures alone should not be used to infer that an intervention caused an event or to make a treatment recommendation.

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

The posted primary-outcome results describe CAPS scores at screening, before the study's treatment comparison can be evaluated. The mean scores were 78.28 for PE, 80.44 for VRET and 78.89 for waitlist. The unadjusted baseline differences were 2.16 units for VRET versus PE and 0.61 units for waitlist versus PE; both confidence intervals include zero. These results indicate no statistically clear difference in the reported baseline means, while leaving uncertainty about the size and direction of the contrasts.

They do not show that PE or VRET reduced PTSD symptoms, changed physiological arousal, or affected perceptions of stigma. Those aims appear in the study summary, but the primary measure reported in this comparison was taken at the screening visit. A baseline comparison cannot determine whether scores changed after treatment or whether change differed across groups.

Likewise, the results do not establish equivalence among the groups. Failure to detect a baseline difference is not a formal equivalence test, and the registry record does not report an equivalence margin. The confidence intervals permit a range of differences and should be considered alongside the scale's purpose and timing.

The serious and other adverse-event counts provide descriptive safety information, but they do not establish causation or comparative risk. Further details about event timing, attribution and analysis are not reported in the registry record. Overall, the posted numbers support a limited statistical reading of baseline symptom scores and reported event counts. They are not a verdict on treatment efficacy, clinical benefit or comparative safety.

16. Why This Trial Matters Statistically

This trial is statistically instructive because the study's broad therapeutic question and its posted primary endpoint do not have the same timing. The study summary describes comparing VRET, PE and waitlist for PTSD, while the reported CAPS endpoint is measured at screening. Recognizing the assessment time is essential: an accurately calculated difference can still answer a different question from the one a reader expects.

The randomized, parallel design assigns participants to distinct groups, and the reported allocation is balanced at 54 participants in each arm. That design provides a framework for comparing assigned groups, but baseline summaries remain descriptive. The reported means and unadjusted intervals concern starting scores, not symptom change or a causal treatment effect.

The confidence intervals also demonstrate why a non-significant comparison should not be translated into proof of equality. Both include zero while allowing differences in either direction. Precision and the clinical meaning of a difference require attention to the interval, the outcome scale and the assessment time—not just the p-value.

Finally, the record illustrates the distinction between trial design calculations and retrospective planning illustrations. The sample-size figures use fractions of observed baseline differences and specified planning assumptions; they are not the original study's justification and do not describe observed power. Separating those quantities from the enrolled sample is important when interpreting what the completed study can 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 main limitation for interpreting the primary result is its timing: CAPS was recorded at screening on Day 1. The posted comparisons therefore describe baseline scores and cannot quantify change during or after therapy. Although the study summary describes treatment hypotheses, post-treatment CAPS results are not reported in the primary outcome information.

The recalculated comparisons are unadjusted and use posted group summaries. The registry record does not report the original analysis model, covariate adjustment, multiplicity procedures, handling of missing observations, or analysis population. Those omissions limit assessment of how closely the simple comparisons align with the study's prespecified analysis.

The record reports 54 participants per arm for this outcome but does not state whether all randomized participants contributed data or how incomplete measurements were addressed. It also does not report adherence, treatment completion, or the number of sessions actually received. The summary's reference to 10 sessions is a general study hypothesis, not a record of delivered treatment.

The registry record does not report a clinically important difference threshold or an equivalence margin for CAPS. Thus, the intervals cannot be judged against a stated clinical benchmark, and the lack of a clear baseline difference is not evidence of equivalence. Safety summaries are also limited: event details, attribution and timing are not reported, and the listed serious-event term is not assigned to an arm in the information available. These limitations constrain conclusions to the posted baseline summaries and descriptive event counts.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Comparing Virtual Reality Exposure Therapy to Prolonged Exposure — NCT01459705.
• Registry record last updated 2015-12-14; status last verified 2015-11.
• Results first posted 2015-12-14.
• 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 describes a completed randomized, single-masked, parallel-group study of PE, VRET and waitlist in active-duty soldiers with PTSD and combat-related trauma. Enrollment was 162, with 54 participants in each arm. The primary outcome was mean CAPS total score at the screening visit on Day 1, using the last-week reference.

Baseline means were 78.28 for PE, 80.44 for VRET and 78.89 for waitlist. The recalculated unadjusted mean differences versus PE were 2.16 for VRET and 0.61 for waitlist; both confidence intervals included zero. These results describe starting scores, not treatment effects. The record also lists secondary mental-health measures and descriptive adverse-event counts, but does not report the detailed analysis, missing-data methods or original sample-size justification.