Virtual Reality and Outpatient Hysteroscopy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Virtual Reality and Outpatient Hysteroscopy (NCT03699280), a randomized na study. The primary outcome is Worst Pain Scores.

2. Clinical Question

3. Trial Design

This was a randomized, single-masked, parallel-group trial comparing virtual reality distraction with the standard procedure during outpatient hysteroscopy. The registry lists the study as completed, with enrollment of 40. It records a start date of 2018-08-04 and primary completion on 2018-10-06. The sponsor was Imperial College London. The trial is listed as not applicable for phase.

The virtual reality arm received a video played during the procedure; the standard-procedure arm received the routine outpatient hysteroscopy protocol. Group summaries for the primary outcome report 20 participants in each arm. The registry does not describe the random sequence generation, allocation concealment, the specific party or parties masked, or the practical details of maintaining masking during a procedure in which one group viewed a video. Those omissions limit what can be assessed about implementation of randomization and masking from the record alone.

The registry describes a mixed-methods design with qualitative and quantitative aims. Its stated primary objective was feasibility of using virtual reality as a distraction technique for acute pain. Secondary aims included acceptability and perceived effectiveness, variation by patient demographics, willingness to participate in a future trial, implementation and content design, and exploring mechanisms that might contribute to pain relief and tailoring. However, no secondary results are posted in the record, and no analysis entries are listed. The quantitative result available for interpretation is the worst pain score during hysteroscopy. Thus, the reported randomized comparison informs the score outcome, rather than all of the broader feasibility and qualitative questions.

The registry gives the study duration as 6 months, but that is distinct from the listed calendar interval between study start and primary completion. Details of follow-up beyond the procedure, withdrawals, adherence to the assigned intervention, and deviations from protocol are not reported in the registry record. Such information would help readers judge how well the reported summaries reflect the intended randomized comparison.

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

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 Worst Pain Scores, measured over during the hysteroscopy procedure, an average of 10 minutes.

The primary endpoint was worst pain during the hysteroscopy procedure, summarized as a mean score with a standard deviation for each arm. Participants rated pain on an 11-point numeric scale running from 0, meaning no pain, to 10, meaning worst imaginable pain. The measure refers to the most pain experienced during the procedure, including pain that was momentary; it is not a rating of average pain over the whole procedure.

The registry gives the timeframe as during the hysteroscopy procedure, an average of 10 minutes. Focusing on the worst reported pain captures the peak experience, which may be important when pain fluctuates over a short procedure. It also has a specific interpretation: the arm mean summarizes participants’ peak ratings, not the duration of pain, the number of painful episodes, or how pain changed over time. A difference in average peak score does not reveal whether the experience was similar for all participants or varied substantially between them.

The registry frames this outcome as assessing feasibility of virtual reality as a distraction technique. Statistically, however, a pain score is a symptom measure rather than a direct measure of feasibility. The endpoint alone does not quantify successful use of the equipment, willingness to use it, acceptability, or ease of implementation. Those were among the broader stated aims, but their results are not reported in the registry record. The mean score comparison therefore addresses the recorded pain outcome, while leaving those other questions open.

7. Results Posted in the Registry

• Virtual Reality: mean 5.65 versus 7.85 on control; difference -2.20 (95% CI -3.74 to -0.66), standardized effect d = -0.89; z-test p = 0.005.

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the primary outcome as a continuous measure summarized by a mean and standard deviation. The posted arm summaries give a mean worst pain score of 5.65 with standard deviation 2.4 for virtual reality, and a mean of 7.85 with standard deviation 2.56 for standard procedure. A mean describes the central level of scores in a group, while the standard deviation describes how dispersed individual scores are around that group mean. Neither statistic shows the full shape of the score distribution.

What does the mean difference represent?

The reported contrast is virtual reality minus standard procedure. Its estimate is -2.2 points, so the sample mean in the virtual reality group is lower. The sign is a matter of the chosen subtraction order: reversing the order would reverse the sign without changing the underlying comparison. Because the outcome is a pain score with larger values representing more pain, the direction of this estimate corresponds to lower average worst pain scores in the virtual reality group. It does not mean that each participant’s score differed by this amount.

A mean difference is expressed in the outcome’s original units, which makes it interpretable on the 0-to-10 scale. The comparison uses the arm means and does not account for other variables. No adjustment for baseline characteristics or other covariates is reported in the registry record. The estimate should therefore be understood as an unadjusted group contrast, not as a covariate-adjusted treatment effect.

What does the p-value tell us?

The unadjusted p-value is 0.005. In general, a p-value describes how unusual data at least as incompatible with equal group means would be under a specified null model and its assumptions. It is not the probability that the null model is true, nor the probability that the observed result arose “by chance” in an everyday sense. It also does not measure the size or practical importance of a difference. The size is described by the mean difference, while uncertainty about that estimate is conveyed by its confidence interval.

The registry does not report an analysis plan or multiplicity strategy. Consequently, the p-value should not be treated as evidence that all potential outcomes and analyses were accounted for in a prespecified testing framework. It is an unadjusted result for the posted primary outcome comparison. Whether other analyses were planned or conducted is not reported in the registry record.

Why use a confidence interval?

A confidence interval gives a range of effect estimates compatible with the data and the statistical procedure, under its assumptions. For this comparison, the interval is for the mean difference in pain score. It excludes zero, the value corresponding to no difference in arm means, and lies in the negative direction. The interval is not a range containing a fixed proportion of individual participant scores. It also does not guarantee that the true effect lies within its bounds. The width reflects both the observed variability and the amount of information available.

The calculation uses the reported arm means, standard deviations, and group sizes. The registry does not report an adjusted model or a different inferential analysis. The posted comparison is an unadjusted recalculation from those summary statistics. The record does not describe checks of model assumptions, distributional diagnostics, or any alternative analysis for the bounded rating scale.

Why does variability matter?

For a continuous outcome, variability affects the precision of a group comparison. When scores vary substantially within arms, the arm means provide less precise information about the difference than they would if scores were more tightly clustered, all else equal. The reported standard deviations, 2.4 and 2.56, indicate notable spread relative to the scale’s full range. The group means do not show whether the distribution is symmetric, whether scores cluster at particular values, or whether there are outlying observations. Those features are not reported.

General statistical explanation: a normal approximation can be useful for estimating uncertainty around a difference in means, but its adequacy depends on the data and design. For bounded ratings, especially with a small sample, the distribution may not resemble an unbounded normal distribution. That possibility does not by itself invalidate the reported comparison; it means that diagnostics or sensitivity analyses could help assess robustness. Such checks are not reported in the registry record.

What does randomization contribute?

Random allocation aims to make groups comparable on average, including for factors that were not measured. It supports a fair comparison of assigned groups when allocation is implemented properly and outcome data are handled appropriately. The registry records randomization, but does not describe sequence generation, concealment, adherence, or missing-data handling. Nor does it state whether the summary is based on all randomized participants or another analysis population. These details matter when assessing the strengths of a randomized estimate.

General statistical explanation: a randomized comparison estimates the contrast associated with assignment under the conditions of the trial. It is not automatically an estimate of the effect of receiving or adhering to the intervention if departures from assignment occur. The record does not report adherence or crossover information, so that distinction cannot be examined for this trial.

10. Confidence Intervals and Precision

The mean difference is -2.2 points for virtual reality minus standard procedure. Its confidence interval extends from a negative lower bound to a negative upper bound and excludes zero. This supports a difference in the direction of lower mean worst pain scores in the virtual reality group under the interval’s assumptions. The endpoints define a range of effect sizes compatible with the observed data and method; they do not show the range of scores experienced by individual participants.

The interval is not especially narrow relative to the scale and point estimate. It allows for a range of possible magnitudes, so the estimated direction is more clearly characterized than the exact size. The reported standard deviations—2.4 and 2.56—show that scores varied among participants in both arms. With 20 participants reported in each arm, that variability contributes to uncertainty in the difference. The interval should therefore accompany the point estimate rather than be replaced by a statement focused only on the p-value.

In general, a confidence interval describes the performance of a method over repeated studies under its assumptions; it is not a probability statement about the particular interval after the data have been observed. Its interpretation also depends on the analysis being appropriate. The registry does not report diagnostic checks, missing-data methods, or an adjusted analysis, so the interval’s precision should be read as conditional on the unadjusted calculation and its assumptions.

Precision and importance are separate considerations. The interval addresses statistical uncertainty, not whether a particular score difference would be meaningful to participants or clinicians. A threshold for a clinically important difference is not reported in the registry record. Nor does the score comparison reveal how many individuals had lower, unchanged, or higher scores. The result therefore gives a group-level estimate with a range of plausible values, but does not settle individual experience or practical importance.

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 event counts. They are illustrations based on the observed mean difference in this trial, rather than the trial’s own original design assumptions. The record does not report the assumptions used when the trial was originally designed, including a target effect or planned sample size justification.

The calculation assumes a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is expressed as a fraction of the observed effect. For planning for 100% of the observed effect, the figures are 20 participants per arm for 80% power and 27 participants per arm for 90% power. For planning for 75% of the observed effect, they are 36 participants per arm for 80% power and 48 participants per arm for 90% power. For planning for 50% of the observed effect, they are 80 participants per arm for 80% power and 107 participants per arm for 90% power.

The 80% power column means that, if the assumed effect and calculation assumptions were correct, a study of the stated size would have an 80% chance of producing a statistically significant result under the specified procedure. The 90% column uses a higher chance of detecting that assumed effect, which requires more participants in these illustrations. These are hypothetical design calculations, not observed power for the completed trial.

Smaller effects require more participants to distinguish them from random variation with the same planned power and significance level. The figures should not be treated as a recommendation or as evidence that the trial was planned around its eventual estimate. Because the calculation uses the observed effect, it is sensitive to uncertainty in that estimate; the original design rationale is not reported in the registry record.

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 = -2.200; pooled standard deviation σ = 2.481.
2. n = 2 × (1.96 + 0.842)² × 2.481² / -2.200² = 20.0, rounded up to 20 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 <- -2.2                      # observed difference in means
sigma <- 2.48128998708333                     # 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(3699280)
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 = 20 (80% power), 27 (90% power)
 75% of observed effect: n per arm = 36 (80% power), 48 (90% power)
 50% of observed effect: n per arm = 80 (80% power), 107 (90% power)
power.t.test: 21 
Simulated power at n = 20 per arm: 0.784 (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 primary outcome, worst pain scores, and reports no secondary results. It records three outcome measures overall, but no results for the other measures are included in the posted record. The registry also lists no analysis entries. A multiplicity adjustment, hierarchy of outcomes, or other strategy for handling multiple comparisons is not reported in the registry record.

Multiplicity matters when researchers examine several outcomes, time points, subgroups, or analytic approaches. Each additional test can create another opportunity to observe a small p-value even when there is no underlying effect, so a prespecified strategy can help interpret a set of findings. The posted p-value of 0.005 is unadjusted. Without an analysis plan or a complete account of the analyses performed, it is not possible to determine whether it was intended as the sole confirmatory test or how it should be interpreted alongside other analyses.

Missing-data methods and the number of participants contributing to the outcome are not described beyond the arm summaries, each of which reports 20 participants. The record does not say whether any randomized participants lacked outcome data, whether all assigned participants were analyzed in their original groups, or how missing scores were handled. It also does not report an intention-to-treat, per-protocol, or other named analysis population. These omissions prevent assessment of possible bias from missing data or exclusions.

The registry describes the trial as mixed-methods, but qualitative analysis procedures and findings are not reported. It is also not reported whether subgroup comparisons by patient demographics were performed, despite the stated interest in whether acceptability or effectiveness might vary across demographic groups. No conclusions about subgroup effects or qualitative themes can be drawn from the posted primary score comparison. These gaps are limits of the registry record, not evidence that analyses were or were not conducted outside it.

14. Safety Results

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

Deaths

Other (non-serious) adverse events

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

The posted adverse-event figures list zero serious adverse events among 20 participants at risk in the virtual reality group and zero among 20 participants at risk in the standard-procedure group. The record also lists zero other adverse events in each group and zero deaths. These figures describe what was recorded for the groups in the registry record; they do not establish that no adverse events occurred outside the recorded definitions or reporting period.

Zero observed events should not be read as proof that an event is impossible or that the intervention is risk-free. With a small number of participants at risk, uncommon events may not appear during the study even if they could occur in a larger population or with longer observation. The registry does not report a detailed definition of adverse events, how events were solicited, the follow-up period for safety, or whether event attribution was assessed. It also lists no serious adverse-event terms.

These data are therefore a limited description of recorded events, not a precise estimate of safety for broader use. The absence of reported events should be interpreted alongside the trial’s size, duration, and the limited safety details available in the registry record.

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

The posted result is a lower mean worst pain score in the virtual reality arm than in the standard-procedure arm: 5.65 compared with 7.85. The unadjusted mean difference is -2.2 points, and the confidence interval excludes zero. The unadjusted p-value is 0.005. Together, these statistics describe a difference in the average peak pain ratings observed in this randomized comparison, with uncertainty about its exact magnitude.

The result does not mean every participant using virtual reality reported less pain, nor does it quantify how many participants’ scores differed in either direction. Group means and standard deviations cannot show the individual response pattern. The endpoint also captures the most painful moment, not the duration or trajectory of pain during the procedure. No threshold for a clinically important difference is reported in the registry record, so the statistical estimate alone does not determine practical importance.

The finding does not answer all of the trial’s stated feasibility questions. A lower pain-score mean is not itself evidence about whether the technology was acceptable, easy to implement, or suitable for different patient groups. The registry does not post secondary results or qualitative findings, so those aims cannot be evaluated from the pain-score comparison. It also does not report enough detail on missing data, adherence, or analysis populations to establish how these factors affected the result.

The randomized design is relevant because it is intended to support comparison of assigned groups, but unreported details of allocation and implementation limit a full assessment of potential bias. The observed result pertains to outpatient hysteroscopy under the conditions of this trial. It should not be generalized automatically to other procedures, settings, or patient groups. Nor do the zero recorded adverse-event figures establish an absence of risk; they are limited observations from this trial’s recorded follow-up.

In short, the record supports a group-level difference in worst pain scores in this small randomized study, while leaving uncertainty about the difference’s precise size, its practical meaning, feasibility outcomes, and how broadly the result applies.

16. Why This Trial Matters Statistically

This trial illustrates how a randomized comparison can be paired with a patient-reported continuous outcome. The mean difference retains the outcome’s original score units, while the standard deviations and confidence interval add information about variation and uncertainty. Reporting all three helps avoid reducing interpretation to a single p-value. Here, the direction of the estimate and the interval’s exclusion of zero are informative, but the interval still allows a range of possible magnitudes.

The trial also demonstrates why a statistically detectable group contrast does not answer every research question. Its stated primary objective concerned feasibility, while the posted quantitative endpoint was a pain score. The score is relevant to the stated interest in acute pain, but feasibility, acceptability, implementation, and willingness to join future research require their own measures and analyses. A pain comparison cannot substitute for those outcomes.

Small trials can produce useful early estimates, but those estimates may be imprecise and can be sensitive to participant variability. Planning illustrations based on an observed effect can show how assumptions about the target effect relate to sample-size requirements. They are not a substitute for the original design rationale, which is not reported in the registry record, and they are not evidence of the trial’s observed power.

Finally, transparent reporting of allocation, analysis populations, missing data, prespecified outcomes, and multiplicity is important for evaluating a result. Those details are not reported in the registry record. Their absence does not negate the posted comparison, but it constrains how confidently readers can judge its robustness and how it fits within the study’s broader mixed-methods aims.

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 arm summaries are based on 20 participants in each group, and the trial enrollment was 40. This is a small study for estimating a mean difference, particularly when pain scores vary between participants. The confidence interval indicates that the exact magnitude remains uncertain. The record does not provide individual-level data or the score distribution, so it is not possible to assess clustering at particular scale points, outliers, or the proportion of participants with different response patterns.

The endpoint is a retrospective rating of each participant’s worst pain during the procedure. That peak-focused measure does not capture pain duration or change over the procedure, and it may not represent the overall experience in the same way as a measure of average pain. The timeframe is stated as the procedure, averaging 10 minutes, but details of when and how ratings were collected are not reported in the registry record.

There is no posted information on the randomization sequence, allocation concealment, masking procedures, adherence, crossovers, or deviations from the assigned approach. Missing-data handling and the analysis population are also not reported. These limitations matter even in a randomized study because they affect whether the group summaries represent the intended comparison.

The registry posts no secondary results despite the broader stated aims, and it lists no analysis entries. It does not report a multiplicity strategy, an analysis plan, demographic subgroup results, or qualitative findings. Consequently, the available result cannot establish acceptability, feasibility of implementation, or variation across patient groups. The registry also does not report a clinically important difference threshold, detailed safety definitions, or a longer-term safety assessment. These limitations constrain interpretation to the posted unadjusted pain-score comparison and recorded adverse-event figures.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Virtual Reality and Outpatient Hysteroscopy — NCT03699280.
• Registry record last updated 2023-02-02; status last verified 2022-05.
• Results first posted 2023-02-02.
• 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

NCT03699280, titled “Virtual Reality and Outpatient Hysteroscopy,” was a completed randomized, single-masked, parallel-group study sponsored by Imperial College London. Enrollment was 40. The recorded study start was 2018-08-04, and primary completion was 2018-10-06. The intervention was a virtual reality video during hysteroscopy; the comparator was the routine outpatient hysteroscopy protocol.

The primary outcome was mean worst pain score during the procedure on an 11-point scale from 0 to 10. The virtual reality group mean was 5.65, compared with 7.85 for standard procedure; the unadjusted mean difference was -2.2. Its confidence interval excludes zero, and the unadjusted p-value was 0.005. The posted comparison is limited by the small group summaries and the absence of detail on analysis populations, missing data, and other methods.

The record lists zero serious adverse events, zero other adverse events, and zero deaths in each group. It reports no secondary results and no analysis entries. Further details on the original sample-size rationale, multiplicity, and qualitative findings are not reported in the registry record.