Office-sclerotherapy for Epistaxis Due to Hereditary Hemorrhagic Telangiectasia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Office-sclerotherapy for Epistaxis Due to Hereditary Hemorrhagic Telangiectasia (NCT01408732), a randomized phase 1, phase 2 study. The primary outcome is Severityof Epistaxis.

2. Clinical Question

3. Trial Design

The registry describes a completed phase 1/phase 2 randomized crossover trial, with no masking. The record lists an enrollment of 18. In a crossover design, participants receive more than one treatment condition in sequence, so treatment comparisons may involve observations from the same participants across conditions. The record does not report how the analysis accounted for this within-participant structure.

The two sequences began differently. One group continued standard treatment during the first six weeks, then received sclerotherapy during the second six weeks, with standard treatment available for breakthrough epistaxis. The other group received sclerotherapy to visible nasal lesions during the first six weeks, then continued standard treatment during the second six weeks, again with needed standard treatment for breakthrough episodes. The arm descriptions do not provide further detail about the standard-treatment methods.

The primary outcome is described as change from baseline to 14 weeks, while the treatment sequences are described in six-week periods. The registry does not explain the timing of all assessments or how the stated endpoint relates to the treatment periods. It also does not report a washout period, sequence-specific outcome results, or an analysis of possible period or carryover effects.

The record lists the University of Minnesota as sponsor and gives the study start as 2011-02 and primary completion as 2014-09. The original sample-size rationale and the trial's planned statistical methods 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: Each group will get the same therapy in a different order Total number of subjects 18 with 2 incompletes. First period is 6 weeks with a 2 week washout period and 2 period is 6 weeks

Overall Study

Reasons for not completing this period:

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Severityof Epistaxis, measured over Change from Baseline to 14 weeks.

The primary endpoint is severity of epistaxis, measured with the Epistaxis Severity Score. The registry describes the ESS as a recently developed online questionnaire that grades epistaxis severity. It consists of questions with a scale ranging from 0-5, and an average score is calculated for a final assessment. A higher score indicates more severe nosebleeds.

The endpoint is expressed as change from baseline to 14 weeks, with the outcome summarized as a mean and standard deviation. A change score focuses on how the measured severity differs from the participant's baseline value rather than relying only on an endpoint score. In general, using a standardized score can help describe symptom severity consistently across participants and treatment conditions.

The registry does not report the exact questionnaire items, scoring details beyond this description, or how missing questionnaire responses were handled. It also does not state whether the 14-week assessment was intended to represent the end of both treatment periods or how the crossover sequence was incorporated into the endpoint analysis. Those details matter when interpreting a change score in a study where treatment conditions occur in sequence.

7. Results Posted in the Registry

• Sclerotherapy Intervention: mean -0.46 versus -1.44 on control; difference 0.98 (95% CI -1.52 to 3.48), standardized effect d = 0.27; z-test p = 0.442.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: it is a numerical severity score summarized with means and standard deviations. The posted comparison reports a mean difference between sclerotherapy and standard treatment, along with a confidence interval and an unadjusted p-value. These statistics answer related but distinct questions. The estimate describes the size and direction of the observed difference; the interval describes its uncertainty; and the p-value measures how compatible the data are with a specified no-difference model under the assumptions of the analysis.

What does the mean difference represent?

A mean difference subtracts one condition's mean from the other's, following a stated ordering. Here, the registry gives a mean of -0.458 for sclerotherapy and -1.438 for standard treatment, and reports a mean difference of 0.980. The sign depends on which condition is subtracted from which. A positive value therefore needs to be read alongside the treatment labels and the score's direction. Since higher ESS values indicate more severe nosebleeds but the record does not explain its change-score sign convention, the estimate should not be translated into a treatment-benefit claim.

For an ordinary parallel-group comparison, the standard error of a mean difference reflects the uncertainty in the estimated means. In a crossover study, observations from the same participant across conditions can be related. That relationship can affect the uncertainty of the within-participant contrast. The registry does not report the analysis model or enough detail to determine whether the posted comparison used paired data, accounted for treatment sequence, or addressed period effects. The displayed comparison should consequently be interpreted as a registry-posted estimate, not as a reconstruction of the trial's full analysis.

How should the confidence interval be read?

A confidence interval gives a range of values compatible with the data and the statistical method used, under that method's assumptions. It is not a range containing a stated proportion of individual participants' scores, and it does not show the probability that a particular treatment effect is true. The reported interval runs from -1.519 to 3.479. Because it includes zero, the summary does not rule out no mean difference; it also includes values in both directions. Its breadth indicates that the estimate is not precise enough to settle the direction or size of the difference.

The registry does not state the confidence level for this interval. It also does not report whether the interval accounts for repeated observations, sequence, or other design features. Those methodological details affect how an interval should be interpreted, particularly in a crossover trial.

What does the p-value add?

The unadjusted p-value is 0.442. In general, a p-value describes how unusual the observed data, or data more extreme, would be if a specified null model were true and the model assumptions held. It is not the probability that the null hypothesis is true, nor does it measure the clinical importance of a difference. The word unadjusted indicates that this reported value is not described as adjusted for multiple comparisons. The record does not report the planned testing hierarchy or a broader analysis plan.

There is no contradiction between the p-value and the interval: both indicate substantial uncertainty around the estimated difference. A p-value alone does not tell readers whether an effect is meaningful, and failure to rule out no difference is not proof of equivalence. Establishing equivalence would require an appropriate design, a prespecified equivalence margin, and analysis details; these are not reported in the registry record.

Why do crossover details matter?

In a crossover study, treatment condition and time or sequence can be entangled. Participants may change over time, and an effect of an earlier treatment could potentially persist into a later period. These are general considerations, not findings that such effects occurred in this study. Random assignment of treatment sequence can help balance some sources of variation, but analysis still needs to reflect the design. The registry does not report a washout, sequence-specific results, or methods for addressing carryover and period effects. Without those details, the posted arm summaries cannot show whether the comparison fully accounts for the crossover structure.

The record reports standard deviations of 2.894 for sclerotherapy and 4.199 for standard treatment. A standard deviation describes spread among measured values around a mean; it is not the uncertainty of the mean difference. In general, the standard error and confidence interval depend on sample size and, for paired data, on the relationship between measurements within participants. This is why the separate condition means and standard deviations do not fully specify a crossover comparison.

10. Confidence Intervals and Precision

The mean difference reported for sclerotherapy relative to standard treatment is 0.980. Its confidence interval extends from -1.519 to 3.479. The interval includes zero and values in both directions, so the result leaves considerable uncertainty about whether the mean change differs between conditions and, if so, by how much. The interval is not narrow enough to identify a clear direction for the comparison.

Precision is different from the size of the observed estimate. An estimate can be nonzero while remaining imprecise, and an interval that includes zero does not prove that the true difference is exactly zero. Nor does it establish that the conditions are equivalent. Equivalence requires a defined range of differences considered acceptably small and an analysis designed to assess that question; the registry record does not report such a margin or analysis.

Each condition's posted outcome count is 16, and the reported standard deviations are 2.894 for sclerotherapy and 4.199 for standard treatment. In general, more outcome information can improve precision, but in a crossover design, the relation between measurements within the same participant also matters. The registry does not report the paired-data correlation, the number of complete pairs, or the method used to calculate the interval. It also does not state the confidence level.

The unadjusted p-value is 0.442. It is consistent with the interval's broad uncertainty, but it should not replace examination of the estimate and interval. Taken together, the posted summaries indicate that the comparison is not precise enough to support a confident conclusion about the direction or magnitude of a treatment-condition difference.

11. Planning a Similar Trial: Sample Size

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

The planning figures are participants per arm, not total events. They are illustrative calculations based on the observed effect, not the trial's original design rationale. The assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The calculations use the stated fraction of the observed effect as the effect to detect. The original sample-size assumptions for the trial are not reported in the registry record.

For planning around 100% of the observed effect, the figures are 213 participants per arm for 80% power and 285 participants per arm for 90% power. For planning around 75% of the observed effect, they are 379 participants per arm for 80% power and 507 participants per arm for 90% power. For planning around 50% of the observed effect, they are 851 participants per arm for 80% power and 1,140 participants per arm for 90% power.

In general, power is the probability that a study will detect an effect of a specified size using a specified analysis, if that effect is present and the assumptions hold. The 80% and 90% columns describe different planning targets: the higher target requires more participants under the stated assumptions. Smaller effects also require more participants to detect, which is why the figures rise as the planned effect becomes a smaller fraction of the observed effect. The accompanying curve expresses this general relationship visually.

These calculations are planning illustrations, not observed power. They are based on a simplified equal-allocation, two-arm framework. The registry does not report whether that framework reflects the crossover design, nor does it provide the original trial's sample-size justification. The figures should therefore be read as illustrations under the stated assumptions, not as a reanalysis of the trial's design.

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 = 0.980; pooled standard deviation σ = 3.606.
2. n = 2 × (1.96 + 0.842)² × 3.606² / 0.980² = 212.7, rounded up to 213 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 <- 0.9797                      # observed difference in means
sigma <- 3.60632035321323                     # 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(1408732)
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 = 213 (80% power), 285 (90% power)
 75% of observed effect: n per arm = 379 (80% power), 507 (90% power)
 50% of observed effect: n per arm = 851 (80% power), 1,140 (90% power)
power.t.test: 213.7 
Simulated power at n = 213 per arm: 0.801 (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 outcome measure and no secondary outcomes. It also lists no statistical analysis entries, although a primary-outcome comparison is posted. The record does not report a prespecified statistical analysis plan, a testing hierarchy, or whether any interim analyses were planned or conducted. It is therefore not possible to determine from the registry record how the posted comparison relates to the original analysis plan.

The reported p-value is explicitly unadjusted. In general, conducting multiple tests can increase the chance of observing an apparently notable result by chance, and multiplicity methods can address that issue when multiple hypotheses are examined. No adjustment is described for this comparison. Since the record does not report additional outcome tests or an analysis plan, the extent of any multiplicity concern cannot be established from the posted information.

Missing-data handling is not reported in the registry record. The record does not explain whether all randomized participants contributed outcome data, how incomplete ESS assessments were treated, or whether the comparison used complete pairs. The outcome entries report 16 observations for each treatment condition, while enrollment is listed as 18. The registry does not provide a participant flow or explain the difference between the enrollment and posted outcome counts. Because the design is crossover, the number of participants with data under both conditions is particularly relevant, but it is not reported.

Analysis populations are also not reported in the registry record. There is no stated intention-to-treat or per-protocol definition, and no description of exclusions from the analysis. Without those details, it is unclear which participants and measurements underlie the posted comparison. These omissions do not establish that the analysis was improper; they limit the ability to evaluate its methods and reproducibility.

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.

For serious adverse events, the registry reports 0 affected participants among 18 at risk in the sclerotherapy group and 0 among 18 at risk in the standard-treatment group. These are the serious-event figures recorded during the study. The record lists no serious adverse-event terms.

For other adverse events, the registry likewise reports 0 affected participants among 18 at risk in each treatment group. The record does not provide event descriptions in this summary. Death counts are not reported in the registry record.

These figures describe what was recorded; they do not establish that no adverse events of any kind occurred or that either treatment is risk-free. The registry does not provide further information about event definitions, monitoring, timing, exposure, or the method used to collect adverse-event data. The at-risk figures should be read as reported for each treatment group, not added together: the study used a crossover design, and the same participants may have received both conditions. The safety entries are descriptive and do not constitute a comparative safety analysis.

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

The registry reports mean ESS changes of -0.458 under sclerotherapy and -1.438 under standard treatment. The reported mean difference is 0.980, with an interval from -1.519 to 3.479 and an unadjusted p-value of 0.442. Those summaries show a difference between the posted means, but the interval includes zero and extends in both directions. The result therefore does not precisely establish the direction or size of a difference between treatment conditions.

These results do not establish that sclerotherapy works or does not work. They do not show that the approaches are equivalent, nor do they determine which approach is clinically preferable. The registry's outcome description identifies higher ESS scores as more severe, but does not explain the sign convention for the change score. Claims about which condition reduced severity more should not be inferred from the signs alone.

The study's randomized crossover design is relevant to interpretation. Participants could contribute information under both conditions, and treatment order and time may matter. The registry does not report how the posted analysis handled within-participant dependence, treatment sequence, period effects, possible carryover, or missing measurements. It also does not report a complete statistical analysis plan. Those limitations make it difficult to determine how fully the posted comparison represents the intended design.

The record is useful as a report of a small study's registered objective, design, and outcome summaries. Its figures can inform discussion of uncertainty and planning, but they should not be treated as a definitive estimate of treatment benefit or harm. The posted adverse-event entries show no affected participants recorded in the listed categories, while additional safety-collection details are not reported. The registry record alone leaves important questions about outcome analysis and completeness unresolved.

16. Why This Trial Matters Statistically

This trial is statistically interesting because the crossover design and the outcome summary raise different interpretive questions. In a crossover, participants receive treatment conditions in sequence, so the comparison may be informed by within-participant data. That can be useful in general, but it also makes sequence, time, and the statistical handling of repeated measurements important. The registry gives treatment sequences but does not report the model used for the comparison or whether those features were addressed.

The posted mean difference is accompanied by a confidence interval that includes zero and spans values in both directions. This makes precision central to interpretation: a point estimate alone can obscure how much uncertainty surrounds it. The unadjusted p-value is also not a substitute for considering the estimated difference and its interval, and it cannot establish equivalence or clinical importance.

The record provides a useful example of the difference between trial enrollment and the counts attached to an outcome. It lists enrollment of 18 and reports 16 observations for each treatment condition, but gives no participant flow or explanation for the outcome counts. In crossover research, knowing whether outcome data are paired and complete is important to understanding the analysis. Those details are not reported.

Finally, the planning illustrations demonstrate how sample-size requirements depend on the effect a future study is designed to detect and on the desired power. They are not a justification for this trial's enrollment, because its original assumptions are not reported, and they do not show the power achieved by the completed study. The distinction between planning calculations and observed evidence is a core part of responsible statistical interpretation.

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 limitations unresolved. Although the study is described as randomized and crossover, it does not report the randomization procedure, sequence allocation details, or whether treatment order was balanced. It gives two six-week treatment periods but does not describe a washout period. The record also does not state whether treatment effects could persist across periods or how such carryover was considered in the analysis.

The primary endpoint is described as change from baseline to 14 weeks, but the registry does not explain the timing of assessments relative to the treatment periods. It does not provide sequence-specific results, period-specific results, or a description of how the crossover design was incorporated into the reported mean difference. The posted group summaries alone do not resolve these issues.

The outcome count is 16 for each condition, whereas enrollment is listed as 18. The record does not explain this difference, identify reasons for missing outcome data, or report how missing values were handled. It also does not say whether the comparison was based on complete pairs. The participant characteristics and eligibility criteria are not reported in the registry entries, limiting assessment of how broadly the results may apply.

The statistical analysis plan, analysis population, confidence level, and original sample-size justification are not reported in the registry record. The posted p-value is unadjusted, and no secondary outcomes are listed. For safety, the record provides affected-participant counts but not detailed event definitions, collection methods, or death counts. These omissions limit assessment of both the outcome analysis and the completeness of the registry summary; they should not be mistaken for evidence that the underlying trial lacked these procedures.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Office-sclerotherapy for Epistaxis Due to Hereditary Hemorrhagic Telangiectasia — NCT01408732.
• Registry record last updated 2019-11-01; status last verified 2019-10.
• Results first posted 2018-04-20.
• 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

NCT01408732, titled “Office-sclerotherapy for Epistaxis Due to Hereditary Hemorrhagic Telangiectasia,” is listed as a completed phase 1/phase 2 randomized, unmasked crossover study sponsored by the University of Minnesota. The record gives a study start of 2011-02, primary completion of 2014-09, and enrollment of 18.

The study compared office-based sclerotherapy with standard treatment for epistaxis associated with HHT. Its primary outcome was change in Epistaxis Severity Score from baseline to 14 weeks. The posted mean difference is 0.980, with a confidence interval from -1.519 to 3.479 and an unadjusted p-value of 0.442. The interval includes zero, and the registry does not report key details of the crossover analysis, missing-data handling, or original sample-size rationale.