Ice Immersion Using Virtual Reality & Augmented Reality: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Ice Immersion Using Virtual Reality & Augmented Reality (NCT04196686), a randomized na study. The primary outcome is VR Effect on Pain Tolerence.

2. Clinical Question

The primary outcome was the mean duration of the ice bath, with longer duration defined in the study description as representing increased pain tolerance. The registry reports a mean of 135.0 seconds for control and 158.0 seconds for VR/AR. A simple, unadjusted comparison of those posted means gives a difference of 23.0 seconds, with a 95% confidence interval from 2.818 to 43.18 seconds. The registry also reports a separate Cox mixed effects analysis, with a hazard ratio of 0.36 and a 95% confidence interval from 0.25 to 0.51. These are different statistical summaries and should not be treated as interchangeable estimates. The event definition and further details needed to interpret the hazard ratio’s direction are not reported in the registry record.

3. Trial Design

This was a randomized, unmasked crossover study of VR/AR during an ice-bath cold pressor test. The registry reports an enrollment of 185 and identifies the study as completed. The study dates in the registry are 2019-07-24 to 2021-03-04 for start and primary completion, respectively. Stanford University is listed as the sponsor. The registry labels the phase as NA.

Participants were assigned to one of two sequences: ice-bath control followed by VR/AR, or VR/AR followed by control. The arm descriptions say that participants were randomized with a 1:1 allocation and that the hand used for the cold pressor test could be dominant or non-dominant. They describe crossing over between the conditions. The text of those descriptions is incomplete in the registry record, so further procedural details cannot be established from it. In particular, any washout procedure, period length, or handling of possible carryover effects is not reported in the registry record.

Masking is listed as none. The study description says participants would not be told the specifics of the time limit, to avoid competitiveness and expectations, but this does not amount to masking of the condition. The primary analysis population is described as participants who complete the respective interventions. The record reports 156 participants for the control outcome summary and 153 for the VR/AR summary; the registry does not explain the difference in those counts. The original sample-size justification, anticipated effect, and design assumptions are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: 185 participants signed informed consent, 156 participants were allocated to a study arm

First Treatment (4 Minutes)

Reasons for not completing this period:

Second Treatment (4 Minutes)

Analysis population for the primary outcome, as defined in the registry: Participants who complete the respective interventions are included in the analysis.

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: Participants who completed the protocol are included in the analysis

6. Primary Endpoint

The registry lists the primary outcome as VR Effect on Pain Tolerence, measured over Duration of ice bath (approximately 0 - 4 minutes).

The primary endpoint is the duration of the ice bath, measured in seconds. The study description treats a longer immersion as indicating greater pain tolerance during the cold pressor test. The registry describes the primary measure as a mean with standard deviation and gives the timeframe as the duration of the ice bath, approximately 0–4 minutes.

As a continuous outcome, duration preserves more information about how long immersion lasted than a simple yes-or-no outcome would. Comparing mean duration is one way to summarize the average difference between conditions. The posted group summaries are 135.0 seconds for control, with a standard deviation of 92.0, and 158.0 seconds for VR/AR, with a standard deviation of 89.0. The summaries describe the observed distributions using means and standard deviations; they do not show individual participant trajectories or how the paired crossover measurements were related.

The registry also reports a Cox mixed effects analysis for the primary outcome. A time-to-event approach can analyze the timing of an event, but the event definition and model specification are not reported in the registry record. Therefore, the hazard ratio cannot be translated from the posted record into a specific change in immersion duration. Duration in this experimental task is a measured endpoint, not a direct measure of every aspect of pain experience or of outcomes outside the cold pressor test.

7. Results Posted in the Registry

• Ice Bath With VR/AR: mean 158.00 versus 135.00 on control; difference 23.00 (95% CI 2.82 to 43.18), standardized effect d = 0.25; z-test p = 0.026.

8. Secondary Efficacy Endpoints

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

Skin Conductance Response Density (SCRD)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Ice Bath With VR/AR: mean 0.09 versus 0.10 on control; difference -0.01 (95% CI -0.02 to 0.00), standardized effect d = -0.15; z-test p = 0.177.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Ice Bath With VR/AR and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcome information does not form a single, consistent pattern that can be interpreted as an overall confirmation of the primary result. The registry reports a sponsor model and a separate simple recalculation for skin conductance response density, and the two analyses yield different inferential signals. The registry also lists perceived pain intensity as a secondary outcome, but a result for that measure is not reported in the registry record.

Skin conductance response density is a physiological measure, while the primary endpoint is duration of ice immersion. These outcomes capture different aspects of the experimental task, so a result for one need not match a result for the other. The secondary outcome summaries cannot establish whether a change in immersion duration was accompanied by a corresponding change in perceived pain or physiological response. The posted information also does not show participant-level associations among these measures.

Secondary outcomes generally carry less evidential weight than a prespecified primary outcome, especially when several outcomes or analyses are considered. Testing multiple outcomes increases the chance that at least one result will appear statistically notable by chance if no adjustment is made. The registry does not report a multiplicity adjustment or a prespecified hierarchy for interpreting secondary results. Consequently, unadjusted secondary p-values should be treated as exploratory signals rather than as stand-alone confirmation.

The sponsor’s secondary analysis uses a linear mixed effects model, while the recalculation is an unadjusted comparison of group means. As with the primary outcome, these approaches answer different statistical questions and are not interchangeable. The record does not describe enough model detail to reconcile their estimates. Overall, the secondary information is incomplete and mixed; it does not resolve the limits in interpretation of the primary result.

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

• VR's Effect on Perceived Pain Intensity
• Skin Conductance Response Density (SCRD)

9. Statistical Interpretation

10. Statistical Methods Explained

The registry presents two different kinds of analysis for the primary outcome: a sponsor-reported Cox mixed effects model and a simple recalculation from the posted group means and standard deviations. They answer related but not identical statistical questions. Their estimates are on different scales, and the available record does not provide the model details needed to reconcile them.

What does the mean difference describe?

A mean is the arithmetic average of a continuous measurement. In this study, the measurement is ice-immersion duration in seconds. The posted means are 135.0 seconds in control and 158.0 seconds with VR/AR. The unadjusted mean difference is reported as 23.0 seconds, with the VR/AR group higher. The recalculated confidence interval is 2.818 to 43.18 seconds. This comparison uses the group summaries rather than participant-level data or a model that explicitly describes the crossover design.

The standard deviation describes the spread of individual measurements around a group mean. The registry reports standard deviations of 92.0 seconds for control and 89.0 seconds for VR/AR. A standard deviation is not the same as a standard error: the standard deviation describes variation among observations, while a standard error describes uncertainty in an estimate such as a mean. The posted calculations also show standard errors for the group means, but they do not provide individual measurements.

What does a Cox mixed effects model do?

In general, a Cox model analyzes the timing of an event and expresses a contrast as a hazard ratio. A hazard is the instantaneous event rate among observations that have not yet experienced the event, conditional on the model. A hazard ratio compares hazards between conditions according to the model’s coding. A hazard ratio is not a difference in mean duration and should not be read as a percentage change in seconds. The interpretation of a hazard ratio below one depends on which event is defined and which condition is in the numerator or reference category.

The registry reports a hazard ratio of 0.36, with a 95% confidence interval from 0.25 to 0.51 and p < 0.001. However, the event definition and the model’s reference coding are not reported in the registry record. The record also does not give enough information about model terms to explain precisely how repeated measurements or crossover sequence were represented. A mixed effects model can, in general, include terms that account for correlated observations, but the registry’s model label alone does not establish which terms were used.

Why can the analyses differ?

The unadjusted recalculation compares means using summary statistics. The sponsor’s Cox mixed effects model analyzes a time-to-event quantity and may use a different model structure and assumptions. They can therefore produce different estimates and p-values without being direct replications of one another. The registry record does not report the specific reason for the discrepancy, such as event coding, model specification, or another analytical choice. No cause should be inferred from the estimates alone. Under the stated reporting convention for a discrepancy, the sponsor’s own analysis takes precedence as the trial’s reported analysis.

The crossover design also matters statistically. Each participant experiences both conditions in a crossover, so measurements from the same participant may be related. In general, analyses of crossover data can account for participant-level dependence and may consider period or sequence effects. The registry does not report enough detail to determine which of these features were modeled. The simple recalculation from separate group summaries should therefore be treated as an unadjusted descriptive comparison, not as a complete crossover analysis.

How should p-values be read?

A p-value describes how compatible data are with a specified null hypothesis under the assumptions of a particular analysis. It is not the probability that the null hypothesis is true, nor does it measure the size or importance of an effect. The sponsor’s reported primary p-value and the unadjusted recalculated p-value belong to different analyses. The registry’s primary result is the Cox mixed effects model result; the recalculation is a separate summary-statistic exercise. The record does not report whether or how the analyses were selected in advance.

The registry also reports a linear mixed effects model for skin conductance response density. In general, a linear mixed effects model can describe a continuous outcome while including fixed terms for conditions or other covariates and random terms for correlated observations. The registry provides the model label and estimate but not its full specification. The simple recalculation for this endpoint is a separate mean comparison and does not reproduce the sponsor’s model.

Finally, the registry lists three outcome measures and three analyses, but detailed model specifications, missing-data methods, multiplicity procedures, and the original sample-size plan are not reported in the registry record. Without these details, the analyses can be described as posted, but they cannot be fully audited or reconstructed from the registry summaries alone.

11. Confidence Intervals and Precision

The recalculated unadjusted primary mean difference is 23.0 seconds, with a 95% confidence interval from 2.818 to 43.18 seconds. The interval describes a range of values compatible with the data and the assumptions of this calculation. Its width reflects uncertainty in the estimated difference. It does not mean that 95% of individual participants had a difference within that range, and it does not capture uncertainty about design or model choices that are not represented in the calculation.

The interval does not include zero, which is consistent with the unadjusted p-value of 0.026 for a test of no mean difference under the calculation’s assumptions. This is not the same inferential result as the sponsor’s Cox mixed effects analysis. The sponsor reports a hazard ratio of 0.36 with a 95% confidence interval from 0.25 to 0.51. That interval is on the hazard-ratio scale and cannot be compared directly with a confidence interval measured in seconds. The record does not report the event coding needed to give the hazard ratio a more specific directional interpretation.

Precision is limited by the variability in immersion duration: the posted standard deviations are 92.0 seconds for control and 89.0 seconds for VR/AR. These spreads are substantial relative to the group means, indicating that individual durations varied. Group-level estimates summarize those observations but do not show how much participants’ responses changed between their two crossover conditions.

The registry does not report a prespecified threshold for a practically important difference, so the interval cannot be classified as establishing a meaningful benefit. The model specification, handling of missing observations, and details of any adjustment for crossover features are also not reported in the registry record. These omissions limit how fully the precision of the sponsor’s model estimate can be assessed from the posted information.

12. Planning a Similar Trial: Sample Size

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

The planning figures are illustrations based on the observed effect, not the trial’s original design assumptions. For continuous outcomes, the figures are participants per arm, not total event counts. The assumptions stated for the illustration are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Each planning scenario uses a specified fraction of the observed effect. Smaller effects require more participants to detect under the same assumptions.

The 80% power column describes the planned number of participants per arm for a design with an 80% chance of detecting the stated effect if that effect is real and the planning assumptions hold. The 90% power column gives the corresponding planning target for a higher chance of detection. These are design calculations, not probabilities that the observed finding is true. For time-to-event planning calculations, the registry’s stated convention is to report the total number of events; the primary outcome’s displayed planning illustration is a continuous-outcome calculation and uses participants per arm.

For the primary comparison, the planning illustration gives 244 participants per arm for 80% power and 326 participants per arm for 90% power when planning for 100% of the observed effect. Planning for 75% of the observed effect gives 433 and 579 participants per arm, respectively; planning for 50% gives 973 and 1,303. The rising targets show how detecting a more modest effect requires a larger planned sample under the same assumptions. A curve based on such calculations should be read as a planning relationship between assumed effect size and sample requirement, not as observed trial performance.

This is not observed power. The original trial’s sample-size justification and assumptions are not reported in the registry record, and the illustration uses the observed effect rather than the trial’s own prospective design target.

13. How the Sample Size Is Derived

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

Formula

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

Worked calculation for this trial

1. Observed difference Δ = 23.000; pooled standard deviation σ = 90.527.
2. n = 2 × (1.96 + 0.842)² × 90.527² / 23.000² = 243.2, rounded up to 244 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 <- 23                      # observed difference in means
sigma <- 90.5270848666991                     # 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(4196686)
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 = 244 (80% power), 326 (90% power)
 75% of observed effect: n per arm = 433 (80% power), 579 (90% power)
 50% of observed effect: n per arm = 973 (80% power), 1,303 (90% power)
power.t.test: 244.2 
Simulated power at n = 244 per arm: 0.799 (20000 replicates)

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

14. Analyses Reported by the Sponsor

The registry’s primary analysis is a Cox mixed effects model. It reports a hazard ratio of 0.36, a 95% confidence interval from 0.25 to 0.51, and p < 0.001. The registry does not report the event definition or enough model details to translate this hazard ratio into a difference in mean immersion duration.

A separate recalculation from the posted group means and standard deviations estimates an unadjusted mean difference of 23.0 seconds, with a 95% confidence interval from 2.818 to 43.18 seconds and an unadjusted p-value of 0.026. This is a simple comparison on the seconds scale, rather than a replication of the sponsor’s time-to-event model. The results differ in both their statistical scale and reported p-value. The reason for that discrepancy is not reported in the registry record.

Where the recalculation differs from the sponsor’s own analysis, the sponsor’s analysis takes precedence as the trial’s reported inferential result. The recalculation remains useful as a transparent, unadjusted comparison of the posted means, but it should not be presented as the primary model result or used to explain the sponsor’s hazard ratio. The record does not provide enough information to determine which modeling choices account for the difference.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists multiple outcomes and analyses, including the primary ice-immersion duration and secondary outcomes. When several outcomes or statistical tests are examined, the chance of obtaining at least one apparently notable result can increase. Statistical methods can address this issue through prespecified testing hierarchies or adjustments, but the record does not report whether either approach was used. The status of the p-values as adjusted or unadjusted is not reported for the sponsor analyses; the simple recalculations are explicitly unadjusted.

The registry does not report an analysis plan that identifies the order of testing, defines which comparisons were primary, or states how secondary outcomes were to be interpreted. This makes it difficult to assess how much weight to place on isolated p-values, particularly for secondary outcomes. The registry’s listing of an outcome does not by itself establish that its analysis was prespecified in the detailed statistical plan.

The primary analysis population is described as participants who complete the respective interventions. That description does not explain how incomplete records were handled, whether all randomized participants were included in any analysis, or whether observations were excluded for other reasons. The registry reports different outcome-summary counts for control and VR/AR, but the reasons for the counts and any missing measurements are not reported in the registry record. No imputation method, sensitivity analysis, or missing-data assumption is described.

Because this is a crossover study, interpretation also depends on how within-participant measurements, condition sequence, and possible period effects were handled. The sponsor model is labeled mixed effects, but the terms included and the treatment of incomplete crossover data are not reported. There is not enough information to assess whether the analysis fully represented the study design or how sensitive the findings are to missing data or modeling choices. These gaps do not establish that a particular procedure was omitted; they mean that it is not reported in the registry record.

16. 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 registry records zero serious adverse events in the control group and zero in the VR/AR group. It also records zero other adverse events in each group and zero deaths in each group. The adverse-event risk sets are reported as 156 for control and 153 for VR/AR.

These figures describe what the registry reports for the listed groups; they do not establish that no adverse events occurred outside the recorded categories or observation period. The registry does not provide further detail about adverse-event definitions, collection procedures, timing, or completeness of reporting. It also does not explain whether the adverse-event population corresponds exactly to the outcome-analysis population.

Zero recorded events should not be interpreted as proof of no risk. In a study with a task-specific exposure and a particular observation period, event counts can be limited, and the record does not provide a detailed safety analysis. These figures are descriptive registry data, not a basis for medical advice or a broad safety conclusion.

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

The posted results describe performance in an ice-water cold pressor test. The primary endpoint is the time a participant kept a hand immersed, which the study used as an indicator of pain tolerance. The results do not directly measure all dimensions of pain, nor do they establish what would happen in other clinical or everyday settings.

The group summaries show longer average immersion duration in the VR/AR condition than in control. A simple unadjusted recalculation estimates a mean difference of 23.0 seconds, with a 95% confidence interval from 2.818 to 43.18 seconds. The sponsor’s reported primary result is a Cox mixed effects model with a hazard ratio of 0.36, a 95% confidence interval from 0.25 to 0.51, and p < 0.001. These analyses use different statistical scales, and the event definition needed to interpret the hazard ratio is not reported. The sponsor’s analysis takes precedence as the trial’s reported inferential result.

Neither estimate establishes the size of an effect in a broader population or in a different type of pain experience. The registry does not report a threshold for a meaningful change in seconds. The primary result also does not separate the different immersive content or coaching approaches described in the study summary, so it cannot identify which component may be associated with the contrast.

The findings should be understood alongside the randomized crossover design, the absence of masking, and incomplete reporting of analysis details. The registry does not describe the original sample-size plan, missing-data procedures, multiplicity handling, or the full model specification. These gaps make it difficult to assess the robustness of the estimates or to determine how all design features entered the analyses.

In short, the registry reports a difference in an experimental behavioral measure and a model-based time-to-event result. It does not show that VR/AR is effective for treating pain, and it does not support medical advice. Generalization beyond the specific cold pressor task and study conditions would require evidence beyond these registry results.

18. Why This Trial Matters Statistically

This trial illustrates why the statistical scale and design matter when interpreting a result. A simple difference in mean duration is easy to express in seconds, while the sponsor’s Cox mixed effects result is a hazard ratio. These are not interchangeable quantities. If the event definition and model specification are not reported, readers cannot fully translate the model result into the practical duration contrast that the endpoint description emphasizes.

The crossover design is also important. Participants experienced both conditions, so their paired observations may be statistically related. A comparison based only on separate group summaries cannot represent all information from such pairing. Mixed effects methods can, in general, model correlated observations, but the registry does not report enough specification to establish exactly how the sponsor’s model handled participant-level dependence, sequence, or period.

Comparing a sponsor model with a transparent unadjusted recalculation helps identify that their results are not the same analysis. It discourages treating a summary-statistic calculation as a substitute for the trial’s own analysis and makes the importance of reporting model details visible. The discrepancy is not explained in the registry record, so it is not possible to determine which analytical feature produced it.

The secondary information and missing methodological details further show why a statistically notable result must be interpreted in context. Multiple outcomes can raise multiplicity concerns, and a physiological secondary measure is not the same construct as immersion duration. Without a reported analysis plan, complete outcome reporting, and sufficient model detail, readers cannot evaluate every analytic choice. The study therefore offers a useful example of how to distinguish a measured task outcome, a model-based estimate, and the limits of what registry reporting can establish.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The registry does not report the event definition for the primary Cox model or enough detail about the model’s coding and terms to interpret its hazard ratio fully. It also does not explain why the sponsor’s primary model result differs from the unadjusted recalculation. The sponsor analysis takes precedence as the reported trial analysis, but the discrepancy remains unexplained.

The study is described as a randomized crossover, yet the registry record does not report whether the analysis accounted for participant-level pairing, period effects, sequence effects, or possible carryover. The arm descriptions are incomplete. Details such as a washout period and the specific implementation of the different VR/AR content variants are not reported in the registry record.

Masking is listed as none. The study description says participants were not told the specifics of the time limit, but participants could experience whether VR/AR was present. Expectations could therefore be relevant to a behavioral endpoint such as immersion duration. The registry does not report a method for assessing expectations or separating their influence from other aspects of the intervention.

The primary analysis population is described as participants completing the respective interventions. The reason for the different outcome-summary counts, the extent of missing measurements, and any approach to missing data are not reported. The original sample-size justification, multiplicity procedures, and a prespecified analysis plan are also not reported in the registry record.

Finally, the endpoint is an experimental cold pressor measure, and the registry does not report a threshold for a practically meaningful difference. The posted primary summary does not distinguish among the different VR/AR content and coaching modifications described in the study summary. These factors limit conclusions about which component is associated with the observed result and how broadly it may apply.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Ice Immersion Using Virtual Reality & Augmented Reality — NCT04196686.
• Registry record last updated 2022-03-25; status last verified 2022-02.
• Results first posted 2022-03-25.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The registry describes a completed, randomized, unmasked crossover study of VR/AR during an ice-water cold pressor test. The primary endpoint is ice-immersion duration in seconds, used as an indicator of pain tolerance. The posted group means are 135.0 seconds for control and 158.0 seconds for VR/AR.

The sponsor’s primary analysis is a Cox mixed effects model, reporting a hazard ratio of 0.36, a 95% confidence interval from 0.25 to 0.51, and p < 0.001. A separate unadjusted recalculation estimates a mean difference of 23.0 seconds, with a 95% confidence interval from 2.818 to 43.18 seconds and p = 0.026. The analyses differ in scale and result; the sponsor’s own analysis takes precedence. The event definition and reason for the discrepancy are not reported in the registry record.

Secondary findings, analysis details, missing-data methods, multiplicity procedures, and the original sample-size rationale are incompletely reported. These results concern the cold pressor task and do not establish effects in other pain settings.