Cox-2 Inhibition in Radiation-induced Oral Mucositis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Cox-2 Inhibition in Radiation-induced Oral Mucositis (NCT00698204), a randomized phase 2 study. The primary outcome is Clinical Oral Mucosal Injury Score at Cumulative Radiation Dose of 5000 cGy.

2. Clinical Question

3. Trial Design

The registry records this as a randomized, triple-masked, parallel-group Phase 2 study of celecoxib versus placebo for radiation-induced oral mucositis. The sponsor is UConn Health. The recorded study start is 2003-07, primary completion is 2012-09, and status is completed. Enrollment is recorded as 43. These registry entries identify the trial structure and timeline but do not describe how randomization was generated or implemented.

The primary endpoint was assessed at a cumulative radiation dose of 5000 cGy, with the timeframe described as 5 weeks from the start of radiation therapy. The registry specifies that subjects participating in the study as they reached that dose were included. One subject in the celecoxib group withdrew before reaching the target dose. The primary endpoint results list 19 in the celecoxib group and 20 in the placebo group. These endpoint-specific group counts should not be treated as a complete account of every participant’s progress through the study.

The registry lists celecoxib and placebo as drug interventions. It does not report the dose, treatment schedule, duration of administration, or additional features of the intervention. Arm descriptions are blank, so details beyond the group labels are not reported in the registry record. The record also does not describe eligibility criteria, the method used to maintain masking, or whether masking was assessed during the trial.

The stated primary outcome is a mean clinical oral mucosal injury score at 5000 cGy. Pain severity at that radiation dose is listed as a secondary outcome. The registry includes group means and standard deviations for both outcomes, along with estimated mean differences and confidence intervals. It does not report formal analyses in the secondary outcome record. Further details about the original analysis plan, including adjustment choices and analysis populations, 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: Three participants enrolled in the study but were not randomized. Two did not meet blood test eligibility criteria for randomization (liver function test, renal function test failure) and one withdrew shortly after enrolling, prior to randomization.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Subjects participating in the study as they reached a cumulative dose of 5000 cGy were included. One subject in the celecoxib group withdrew prior to reaching 5000 cGy.

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: Intent-to-treat analysis performed for 40 participants. Per-protocol analysis performed for the period in which individual participants were taking study medication (celecoxib or placebo).

6. Primary Endpoint

The registry lists the primary outcome as Clinical Oral Mucosal Injury Score at Cumulative Radiation Dose of 5000 cGy, measured over 5 weeks from start of radiation therapy (5000 cGy).

The primary endpoint is the mean Oral Mucositis Assessment Scale score at a cumulative radiation dose of 5000 cGy, described as approximately 5 weeks from the start of radiation therapy. The scale assesses oral mucosal injury by scoring ulceration and erythema independently at nine specified sites in the oral cavity. Ulceration is scored from 0-3 according to lesion size, and erythema from 0-2 according to severity. The site scores are summed and divided by 9 to produce the reported measure. The unit is described as units on a scale.

This measure is used to represent the clinical severity of visible oral mucosal injury at a specified point during radiation treatment. A mean summarizes the average score within each group; the standard deviation describes how dispersed individual scores are around that group mean. The endpoint therefore gives a group-level comparison at the target radiation dose, rather than a description of each participant’s course or change across the entire treatment period.

The time point is relevant to the study question because the registry describes oral mucositis as a complication of radiation therapy and evaluates injury as participants reached the stated cumulative dose. The results do not by themselves describe injury at other radiation doses or later follow-up. The registry does not report additional details about how assessments were conducted, such as assessor training or procedures for resolving scoring disagreements.

7. Results Posted in the Registry

• I- Celecoxib: mean 1.42 versus 1.36 on control; difference 0.06 (95% CI -0.43 to 0.55), standardized effect d = 0.08; z-test p = 0.810.

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.

Evaluation of Pain Severity at 5000 cGy Radiation

• 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:
  • I- Celecoxib: mean 4.47 versus 3.70 on control; difference 0.77 (95% CI -1.15 to 2.69), standardized effect d = 0.26; z-test p = 0.431.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between I- Celecoxib and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists one secondary endpoint: pain severity at a cumulative radiation dose of 5000 cGy. Its estimated mean difference is positive, as is the primary estimate, so both posted point estimates place the celecoxib group mean numerically above the placebo group mean. This is a directional description of the estimates, not a conclusion about efficacy or clinical importance. The endpoints measure different aspects of the study question: oral mucosal injury is primary, while pain severity is secondary.

The secondary confidence interval includes zero and extends to values on either side of it. Its unadjusted p-value also does not provide a clear distinction between the groups. This broadly matches the primary result in the limited sense that neither comparison precisely separates the group means. It does not amount to replication across multiple secondary outcomes: only one secondary endpoint is posted in the registry record.

Secondary results generally carry less weight than a prespecified primary result because they may be less central to the main question and can be more vulnerable to selective interpretation when several outcomes are considered. The registry does not report whether the secondary endpoint was prespecified in the analysis plan, how many outcomes were considered during the study, or whether any multiplicity adjustment was planned. It reports no formal analysis for the secondary outcome. These omissions make it difficult to assess the inferential status of the secondary comparison.

The secondary result can still provide descriptive context, but it should not be treated as a separate confirmation of the primary endpoint. Nor should the agreement in the signs of the point estimates be taken as evidence that the effects are consistent across the broader course of treatment. The record reports results at the specified dose, not a sequence of comparisons across time. The secondary finding is best read as an uncertain, additional comparison that does not materially resolve the primary result.

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

• Evaluation of Pain Severity at 5000 cGy Radiation

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous and is reported as a mean with a standard deviation. A mean is the arithmetic average of the recorded scores in a group. It gives a compact description of the group’s central score, while the standard deviation describes how much individual observations vary around that average. Neither statistic alone describes the full distribution of scores. The registry does not report individual participant data or a distributional plot for the endpoint.

What does the mean difference represent?

The comparison is expressed as the celecoxib group mean minus the placebo group mean. The primary estimate of 0.06 therefore represents the observed difference between the two posted means, 1.42 and 1.36. A value of zero would correspond to equal group means on this scale. A positive estimate means the celecoxib group mean is numerically higher; a negative estimate would mean it is lower. This sign describes direction only. Its clinical meaning depends on the outcome’s scoring and context, and the registry does not report a threshold for a clinically important difference.

The same general interpretation applies to the secondary mean difference. A difference in means is a comparison of group averages, not a count of participants who improved or worsened. It does not show whether the group distributions overlap, whether a small number of observations influence the averages, or how scores changed within individuals. Those questions require information not reported in the registry record.

What does a confidence interval add?

A confidence interval communicates uncertainty around an estimated group difference. It gives a range of values reasonably compatible with the data under the method used to calculate the interval. For the primary result, the interval includes zero and extends to both negative and positive values. This means the posted estimate does not tightly distinguish a difference in one direction from a difference in the other direction. A confidence interval is not a guarantee that the true value lies within its limits, and it is not a range containing a stated percentage of individual scores.

The interval is especially important when an estimate is interpreted alongside a p-value. A p-value describes how unusual the observed data, or data more extreme, would be under a specified null model. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of a difference. Here, the primary unadjusted p-value is 0.81. The interval and p-value both indicate that these results do not sharply resolve the comparison, but neither proves equality.

Why report standard deviations?

Standard deviations provide information about within-group variability. The primary outcome standard deviation was 0.67 for celecoxib and 0.88 for placebo. Greater dispersion can make a group mean less precise when the number of observations is limited, because the observed average is then less stable from sample to sample. In general, standard deviations describe variation among scores, whereas a confidence interval for a mean difference describes uncertainty about the contrast between group means. A standard deviation is not a standard error and should not be read as the uncertainty interval for the comparison.

For the secondary pain-severity outcome, the registry also reports means and standard deviations. The same distinction applies: the standard deviations describe spread within each group, and the estimated mean difference compares the group averages. The pain scale’s scoring details and a threshold for meaningful change are not reported in the registry record.

What does an unadjusted result mean?

The posted p-values are identified as unadjusted. An unadjusted p-value is calculated for a particular comparison without an adjustment for the possibility that several outcomes or comparisons are being considered together. When many tests are performed, the chance of encountering a small p-value by chance can increase. Whether and how to adjust depends on the prespecified analysis plan and the role of each endpoint. The registry does not report a multiplicity strategy or formal analyses for the secondary outcome, so the secondary result should be treated as exploratory rather than as independent confirmation of the primary result.

These concepts describe general statistical principles. The registry record does not give a full account of the model, assumptions, handling of missing observations, or details of the interval calculation. The displayed estimates can be interpreted as group-level comparisons, but the unreported analysis details limit assessment of how sensitive the results might be to alternative methods.

11. Confidence Intervals and Precision

The primary estimated mean difference is 0.06, with an unadjusted p-value of 0.81. Its confidence interval extends from below zero to above zero. That span matters: the data are compatible with a negative difference, no difference, and a positive difference within the reported interval. The point estimate alone is therefore not a precise account of the range of differences that remain plausible given the results.

The group means are based on 19 in the celecoxib arm and 20 in the placebo arm for the primary endpoint. The standard deviations are 0.67 and 0.88, respectively. These values give context for the observed spread of scores, while the interval for the mean difference expresses uncertainty in the comparison. The registry’s result does not establish that any one value inside the interval is more likely than another.

For the secondary pain-severity outcome, the estimated mean difference is 0.77, and its confidence interval runs from -1.148 to 2.688. The interval crosses zero and includes values in both directions. Its width indicates that the estimated contrast is not sharply determined by these data. The associated unadjusted p-value is 0.431. Since the endpoint is secondary and the registry does not report a formal analysis or multiplicity plan, its interval should be read as descriptive evidence with that context in mind.

Precision is distinct from statistical significance and clinical importance. A narrow interval can still include differences that matter, and a point estimate near zero does not prove equivalence. The registry does not state a clinically important difference for either outcome. Accordingly, these results describe the degree of uncertainty in the posted comparisons but do not establish a threshold-based conclusion.

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 participants per arm, not total events. They are illustrations computed from the observed effect, not the trial’s original design assumptions. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. The original trial’s own sample-size assumptions are not reported in the registry record.

For the primary outcome, the planning figures use the observed mean difference of 0.06. The row for planning for 100% of the observed effect gives 2,687 participants per arm for 80% power and 3,597 participants per arm for 90% power. Planning for 75% of the observed effect gives 4,776 participants per arm for 80% power and 6,394 participants per arm for 90% power. Planning for 50% of the observed effect gives 10,746 participants per arm for 80% power and 14,385 participants per arm for 90% power.

The 80% power column represents a design that, under the stated assumptions and if the target effect were the true effect, would detect that effect in 80% of repeated studies using the specified testing approach. The 90% column describes the corresponding design target at a higher probability of detection. These are design concepts, not probabilities that a result is true or that a particular trial succeeded.

Smaller effects require more participants to distinguish them from random variation under otherwise unchanged assumptions. That pattern explains why the planning figures rise as the target fraction of the observed effect becomes smaller. The figures are planning illustrations, not observed power. They should not be read as evidence that the trial had a particular probability of detecting its result.

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 = 0.060; pooled standard deviation σ = 0.785.
2. n = 2 × (1.96 + 0.842)² × 0.785² / 0.060² = 2,686.3, rounded up to 2,687 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 <- 0.0599999999999998                      # observed difference in means
sigma <- 0.784887666261006                     # 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(698204)
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 = 2,687 (80% power), 3,597 (90% power)
 75% of observed effect: n per arm = 4,776 (80% power), 6,394 (90% power)
 50% of observed effect: n per arm = 10,746 (80% power), 14,385 (90% power)
power.t.test: 2687.2 
Simulated power at n = 2,687 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.

14. Multiplicity, Missing Data and Analysis Populations

The registry reports a primary endpoint and one secondary endpoint, but does not report a statistical analysis plan or a strategy for handling multiplicity. Multiplicity arises when several outcomes, comparisons, or analyses are examined: the more opportunities there are to find an apparently unusual result, the greater the risk that one may arise by chance. An adjustment can address that risk, but the appropriate approach depends on which outcomes were prespecified and how they were intended to support conclusions. Those details are not reported in the registry record.

The posted p-values are unadjusted. The registry lists no reported secondary analyses, and the secondary outcome record contains no formal analysis. This does not mean that no analyses were conducted; it means their details are not reported in the registry record. It is therefore not possible to determine whether the posted comparisons were the only analyses considered or how they relate to any planned family of tests.

The primary endpoint population is defined by reaching a cumulative radiation dose of 5000 cGy. The record notes that one subject in the celecoxib group withdrew before reaching that dose. It does not describe an imputation method, a missing-data sensitivity analysis, or how other incomplete observations were handled. It also does not report whether the primary comparison followed an intention-to-treat principle or another analysis population definition. The endpoint-specific group counts are reported, but those counts alone do not establish the full analysis strategy.

The secondary endpoint counts are 19 in each group. The registry does not explain why the endpoint-specific counts differ from the primary endpoint counts, or describe any missing-data procedures for pain severity. Details about the randomization sequence, concealment, covariate adjustment, model diagnostics, and prespecified handling of protocol deviations are also not reported in the registry record.

These gaps limit assessment of the robustness and confirmatory status of the comparisons. They do not invalidate the posted descriptive statistics, but they restrict conclusions about how the analyses were planned and whether alternative reasonable choices would change the interpretation.

15. 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: fever (15.0%), nausea/vomiting (15.0%), electrolyte abnormality (5.0%), febrile neutropenia (5.0%), myocardial infarction (5.0%).

The registry reports serious adverse events for 6 of 20 at risk in the celecoxib group and 5 of 20 at risk in the placebo group. Other adverse events are reported for 1 of 20 at risk in the celecoxib group and 0 of 20 at risk in the placebo group. These are group-specific counts and denominators as recorded; they should be read as safety summaries, not as a complete account of each event’s timing, severity, or relationship to study treatment.

The serious-event terms listed most often in the registry include Fever and Nausea/vomiting, each at 15.0%, and Electrolyte abnormality, Febrile neutropenia, Myocardial infarction, and Pulmonary embolism, each at 5.0%. The registry does not specify in this listing how the terms were distributed between treatment groups. The terms are descriptive and do not establish causality.

Deaths are not reported in the registry record. The available adverse-event counts alone do not support a full comparative safety assessment. The record does not provide event-level narratives, exposure-adjusted rates, timing, or detailed attribution beyond the listed group summaries.

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

The posted primary result describes mean oral mucosal injury scores at a particular cumulative radiation dose. The celecoxib group mean was 1.42 and the placebo group mean was 1.36, with an estimated mean difference of 0.06. The confidence interval includes values on both sides of zero, and the unadjusted p-value is 0.81. The result therefore does not precisely distinguish the two group means in this study. It does not prove that the groups are equivalent, and it does not establish that one treatment changes oral mucositis.

The confidence interval also means that the point estimate should not be treated as the only plausible description of the comparison. The data allow uncertainty around the direction and size of the group difference. The registry does not define a clinically important difference for the OMAS score, so statistical estimates cannot be translated into a threshold-based judgment about clinical importance.

The trial record concerns people receiving radiation therapy for head and neck cancer and an assessment at 5000 cGy. It does not report the effect of treatment at other radiation doses, the pattern of symptoms over time, or outcomes beyond the listed measures. The brief summary identifies pain, nutrition, quality of life, and potential interruptions to radiation therapy as relevant concerns, but the posted outcome results do not quantify those concerns.

The secondary pain-severity comparison is also uncertain and is unadjusted. It should not be treated as confirmation of the primary result, especially since the registry does not report a multiplicity plan or formal secondary analysis. Similarly, adverse-event counts are descriptive and do not establish whether an event was caused by either intervention.

Overall, the registry results are a record of group summaries and estimates from this trial. They can inform statistical understanding of the measured outcomes, but they do not justify a broad conclusion about treatment effects, safety, or individual care. The registry does not report enough methodological detail to assess all possible sources of bias or the sensitivity of findings to alternative analyses.

17. Why This Trial Matters Statistically

This trial illustrates why an estimated difference should be interpreted together with its uncertainty. The primary group means are close, and the estimated mean difference is 0.06, but the confidence interval extends in both directions around no difference. The point estimate alone does not capture that uncertainty. A reader who focuses only on the observed means could overstate how clearly the comparison has been resolved.

It also demonstrates the difference between within-group variability and uncertainty in a between-group contrast. The standard deviations describe the spread of scores within the celecoxib and placebo groups. The confidence interval describes uncertainty about the difference between their means. Keeping those concepts separate helps prevent misinterpretation of descriptive variation as evidence about the precision of the comparison.

The record also makes clear that a posted p-value is not a complete statistical conclusion. The primary p-value is unadjusted, and the registry does not report a multiplicity plan, formal analysis details, or the original sample-size assumptions. The secondary result therefore needs cautious interpretation, rather than being counted as an independent confirmation. Transparent reporting of those methods would help readers understand how much inferential weight to give the findings.

Finally, planning illustrations based on the observed effect can show how target effect size and desired power relate to participant requirements. They are not a substitute for a trial’s prespecified design rationale and do not describe the power actually achieved. This distinction is important when interpreting a trial with endpoint-specific group counts and uncertain estimates.

18. 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 endpoint-specific comparisons are based on 19 in the celecoxib group and 20 in the placebo group for the primary outcome. The registry records enrollment as 43, but does not provide a complete participant-flow account linking enrollment, withdrawals, endpoint availability, and analysis populations. It notes that one celecoxib participant withdrew before reaching 5000 cGy, while further details of missing observations are not reported in the registry record.

The principal outcome is measured at one specified cumulative radiation dose. The registry does not provide results across multiple radiation doses or a longitudinal description of scores. This limits what can be learned about the timing and course of oral mucosal injury from the posted comparison. The secondary pain-severity outcome is also reported at the target dose, and the record does not describe its scale in detail.

Several design and analysis details are not reported in the registry record, including the randomization method, masking roles, intervention dose and schedule, original sample-size rationale, analysis plan, handling of missing data, and multiplicity strategy. Arm descriptions are blank. Without those details, readers cannot fully assess the planned comparison or the robustness of the results to analytic choices.

The primary confidence interval includes zero and values on either side, so the estimate remains uncertain. The registry does not state a clinically important threshold for the injury score or pain outcome. The adverse-event summaries provide counts and selected terms but do not include timing, event-level narratives, or detailed group allocation for the listed serious-event terms. These are limitations of interpretation of the registry record, not proof that the trial lacked additional information outside it.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Cox-2 Inhibition in Radiation-induced Oral Mucositis — NCT00698204.
• Registry record last updated 2014-06-05; status last verified 2014-05.
• Results first posted 2014-06-05.
• 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.

22. Record Summary

NCT00698204 is a completed randomized, triple-masked, parallel-group Phase 2 study sponsored by UConn Health. It compared celecoxib with placebo for oral mucositis associated with radiation therapy. The registry records enrollment of 43, a start date of 2003-07, and primary completion in 2012-09.

The primary outcome was mean OMAS clinical oral mucosal injury score at a cumulative radiation dose of 5000 cGy. The posted means were 1.42 for celecoxib and 1.36 for placebo; the estimated difference was 0.06, with an unadjusted p-value of 0.81 and a confidence interval spanning zero. Pain severity at the same dose is listed as a secondary outcome, with an uncertain unadjusted comparison. The registry reports adverse-event counts by group but limited analysis details. Original design assumptions, multiplicity handling, missing-data procedures, and formal secondary analyses are not reported in the registry record.