This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Study of the Bioequivalence of the 20% and 30% Etoricoxib Tablet Formulations (0663-070) (NCT00945035), a randomized phase 1 study. The primary outcome is Plasma Area Under the Curve (AUC(0 to Infinity)) for Etoricoxib.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT00945035 |
| Phase | Phase 1 |
| Condition | Pain |
| Allocation | Randomized |
| Design model | Crossover |
| Masking | None (open label) |
| Enrollment | 36 |
| Arms in the primary outcome | 2 |
| Lead sponsor | Organon and Co |
| Results posted | Yes |
2. Clinical Question
Population
Healthy adult subjects who completed the study and were included in the statistical analysis. The registry reports enrollment of 36 participants.
Intervention
A single dose of the 20% milled etoricoxib tablet formulation, described in the brief summary as the final market image formulation.
Comparator
A single dose of the 30% unmilled etoricoxib tablet formulation. The registry describes it as unmilled roller compaction in the study summary and labels the arm Etoricoxib URC in the outcome results.
Primary question
How did plasma exposure, measured by area under the concentration-time curve from zero to infinity, compare between the 20% and 30% formulations? The registry’s stated objective was to establish bioequivalence. The record posts a least-squares mean ratio and confidence interval, but does not report the equivalence limits or a conclusion against those limits.
This is a formulation-comparison question, not a comparison of etoricoxib with an inactive treatment. Both study arms received etoricoxib, and the main contrast concerns the tablet formulation. The participants were healthy adults rather than patients receiving treatment for pain, even though pain is listed as the condition. Accordingly, the posted exposure results describe the pharmacokinetic comparison in the study population; they do not directly describe clinical outcomes in people being treated for pain.
The endpoint covers plasma exposure through the concentration-time curve, with the measurement timeframe stated as through 120 hours postdose. The registry also reports a separate safety summary. It does not post secondary outcome measures. The central interpretive issue is therefore what the ratio and its uncertainty say about relative exposure, and what cannot be concluded without the registry’s equivalence criteria and fuller analysis details.
3. Trial Design
The registry identifies this as a phase 1, randomized, open-label, two-period, single-dose, balanced crossover study in healthy subjects. Enrollment was 36, and the study status is completed. Each arm received an etoricoxib tablet formulation: the arm descriptions identify 20% and 30% tablets, while the outcome labels use Etoricoxib FMI and Etoricoxib URC.
In a crossover design, participants receive more than one study condition in different periods, so comparisons can be made within the same participant. This can reduce the influence of stable differences between participants. Randomization is used to assign the treatment sequence, while a balanced design generally aims to distribute sequences across participants. Those are general features of crossover and balanced designs; the registry does not report the sequence details, the interval between periods, or how balance was implemented.
The study was open-label, meaning masking is recorded as none. The record describes a single dose in each period. It does not state whether any washout interval was used, how adherence or period effects were handled, or whether any participants failed to complete a period. The primary outcome population is described as all healthy adult subjects who completed the study and were included in the statistical analysis.
The primary endpoint was plasma area under the curve from zero to infinity, measured through 120 hours postdose. The posted outcome is continuous and is reported as a least-squares mean with standard deviation for each formulation. A sponsor analysis additionally reports a least-squares mean ratio with a 90% confidence interval. Details of the model used to produce that analysis are not reported in the registry record.
The registry lists two outcome measures and two analyses, but no secondary outcomes are posted. It lists Organon and Co as sponsor. The registry record does not report the trial’s original sample-size justification, prespecified analysis plan, or any interim analyses.
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.
Period 1
| Stage | Etoricoxib FMI Then Etoricoxib URC | Etoricoxib URC Then Etoricoxib FMI |
|---|---|---|
| STARTED | 18 | 18 |
| COMPLETED | 18 | 18 |
| NOT COMPLETED | 0 | 0 |
Period 2
| Stage | Etoricoxib FMI Then Etoricoxib URC | Etoricoxib URC Then Etoricoxib FMI |
|---|---|---|
| STARTED | 18 | 18 |
| COMPLETED | 18 | 18 |
| NOT COMPLETED | 0 | 0 |
Analysis population for the primary outcome, as defined in the registry: All healthy adult subjects completed the study and were included in the statistical 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.
| Characteristic | Unit | All Participants in Study |
|---|---|---|
| Participants analysed | 36 | |
| Age, Continuous | Years | 36.1 (18 to 44) |
| Sex: Female, Male: Female | Participants | 17 |
| Sex: Female, Male: Male | Participants | 19 |
| Height | Centimeters | 168.3 (149.9 to 193.0) |
| Weight | Kilograms | 75.2 (50.5 to 99.1) |
6. Primary Endpoint
The registry lists the primary outcome as Plasma Area Under the Curve (AUC(0 to Infinity)) for Etoricoxib, measured over Through 120 Hours Postdose.
| Element | Registry entry |
|---|---|
| Result type | Least Squares Mean |
| Measure of spread | Standard Deviation |
| Unit | µg times hr/mL |
| Endpoint classification | Continuous Endpoint |
The primary endpoint is plasma area under the concentration-time curve for etoricoxib from zero to infinity, with the timeframe stated as through 120 hours postdose. AUC represents an integrated measure of drug concentration in plasma over time. In general, it is used to characterize overall systemic exposure rather than the concentration at a single sampling time. The registry describes the measure as the area under the plasma concentration versus time curve.
The 20% formulation has a posted least-squares mean of 32.65 µg times hr/mL, with a standard deviation of 16.2. The 30% formulation has a posted least-squares mean of 33.28 µg times hr/mL, with a standard deviation of 13.65. The registry’s sponsor analysis presents the comparison as a least-squares mean ratio, while a separate unadjusted recalculation expresses it as a mean difference. These are different scales for describing the contrast and should not be treated as interchangeable estimates.
AUC is relevant to a bioequivalence question because it compares overall exposure between formulations. However, the endpoint alone does not establish bioequivalence. Interpretation of a bioequivalence analysis depends on the prespecified acceptance limits and the analysis method. Those limits and methodological details are not reported in the registry record. The posted result can therefore be described, but the record does not permit a complete assessment against formal bioequivalence criteria.
7. Results Posted in the Registry
- Etoricoxib URC: mean 33.28 versus 32.65 on control; difference 0.63 (95% CI -6.29 to 7.55), standardized effect d = 0.04; z-test p = 0.858.
Mean difference: Etoricoxib URC
95% CI: -6.29 to 7.55
Recalculated from the posted summary statistics
8. Statistical Interpretation
The sponsor’s posted analysis gives a least-squares mean ratio of 1.02 for the two formulations, with a 90% confidence interval from 0.97 to 1.07. This is a relative-scale comparison: a ratio near 1 indicates similar estimated mean exposure on that scale. The interval describes uncertainty in the estimated ratio under the sponsor analysis. Whether the result meets a formal bioequivalence criterion depends on acceptance limits, which are not reported in the registry record. The ratio and interval alone should not be converted into a categorical conclusion about whether the study established bioequivalence.
The posted formulation means are 32.65 for Etoricoxib FMI and 33.28 for Etoricoxib URC, in the stated unit of µg times hr/mL. The unadjusted recalculation gives a mean difference of 0.63, with an interval from -6.29 to 7.55 and an unadjusted p-value of 0.858. This calculation is based on posted summary statistics and is not the same analysis as the sponsor’s least-squares mean ratio. Its result is not a substitute for the crossover analysis or for the sponsor’s reported comparison.
A large p-value in a test of a difference does not establish equivalence. In general, a test that asks whether there is evidence of a difference answers a different question from an equivalence analysis, which evaluates whether the uncertainty lies within prespecified bounds. The registry does not state the equivalence margins, the model specification, or the sponsor analysis method. The sponsor’s ratio and confidence interval are therefore the direct registry-reported comparative result, while the recalculation is a limited descriptive check.
The variability summaries also matter. The reported standard deviations are 16.2 for FMI and 13.65 for URC. These describe spread in the endpoint values within each arm, not the uncertainty of the ratio itself. The sponsor’s confidence interval is the more direct indication of precision for its relative comparison. The registry does not post arm-level confidence intervals or participant-level results.
The safety record lists zero serious adverse events among 36 participants at risk in each formulation group. It also lists 9 participants with other adverse events in the FMI group and 12 in the URC group. These counts describe the posted safety summary and do not establish that an event was caused by a formulation. Death counts are not reported in the registry record.
Overall, the record supports a description of estimated exposure and its uncertainty in a small healthy-subject crossover study. It does not provide enough information to reconstruct the sponsor model, evaluate the equivalence boundaries, or extend the result to clinical outcomes in patients. The appropriate statistical reading is bounded by those omissions.
9. Statistical Methods Explained
What does a crossover design contribute?
In a crossover study, participants receive multiple treatments in separate periods. Because each participant can contribute observations under both formulations, the comparison can account for stable participant characteristics that might otherwise add between-person variation. Randomization of treatment sequence helps guard against systematic differences in the order in which formulations are received. These are general design principles, not a description of undocumented implementation details. The registry identifies the study as randomized, balanced, and crossover, but does not report the sequence allocation, period structure beyond the two-period description, or any washout procedure.
Period and sequence can matter in crossover analyses. For example, an outcome in a later period may be influenced by period conditions or by carryover from an earlier treatment. Statistical methods can account for design features when they are specified and measured. The registry record does not state whether period or sequence terms were included, how carryover was considered, or how the crossover observations were modeled. It is therefore not possible to infer the sponsor’s model from the design label alone.
What is the endpoint scale?
The primary measure is a continuous AUC outcome. A continuous measure can take many numeric values rather than being limited to categories. AUC summarizes concentration over time, and its stated unit is µg times hr/mL. The endpoint is described as extending from zero to infinity, with measurement through 120 hours postdose. In general, an endpoint labeled to infinity may involve estimating exposure beyond the last measured time, but the registry does not describe the method used to obtain that portion of the measure.
The posted arm values are least-squares means, accompanied by standard deviations. A least-squares mean is generally a model-based mean adjusted for terms in a statistical model. The registry does not say which terms were included, so the adjustment cannot be specified. A standard deviation describes how dispersed individual observations are around a group mean; it is not itself a confidence interval and does not show the uncertainty of a between-formulation contrast.
How do the two comparisons differ?
The sponsor analysis reports a least-squares mean ratio of 1.02 and a 90% confidence interval from 0.97 to 1.07. A ratio expresses one estimated mean relative to another. This is a natural scale for many bioequivalence assessments because the question concerns relative exposure. The registry does not specify the direction used to form the ratio in words, the model, or the acceptance limits. The arm labels identify the groups, but the ratio should be interpreted as the sponsor’s reported comparison rather than reconstructed from the displayed arithmetic means.
The unadjusted recalculation reports a mean difference of 0.63, an interval from -6.29 to 7.55, and an unadjusted p-value of 0.858. A difference expresses an absolute contrast in the endpoint’s original units. The recalculation uses posted summary statistics and an approximate standard deviation; it is not a recreation of a crossover model. In particular, a simple comparison based on group summaries does not use the within-participant relationship that can be important in a crossover design. The two results answer related but not identical statistical descriptions of the comparison.
A p-value summarizes how compatible the observed data are with a specified null hypothesis under a statistical model. It does not measure the probability that the null hypothesis is true, and it is not a measure of clinical importance. The recalculated p-value is explicitly unadjusted. The sponsor analysis does not report a p-value. Neither should be treated as a standalone test of bioequivalence, because an equivalence question requires prespecified margins and an appropriate interval-based assessment.
Why do confidence intervals matter?
A confidence interval gives a range of values compatible with the data and the assumptions of the analysis, at the stated confidence level. In general, a narrower interval indicates greater precision, while a wider interval indicates more uncertainty about the estimate. The sponsor’s interval is explicitly a 90% confidence interval for the ratio. The confidence level for the interval attached to the recalculated difference is not stated in the registry record, so it should not be assigned a level by inference.
In an equivalence framework, the interval is considered in relation to boundaries chosen before the analysis. If the full interval falls within those boundaries, that can support an equivalence conclusion under the specified method. If the boundaries are absent, readers cannot determine whether the posted interval satisfies the criterion. No such limits are reported in this record. The closeness of an estimated ratio to 1 is not enough on its own to determine equivalence.
What remains undocumented?
The registry does not report the sponsor’s statistical method, the original sample-size assumptions, the randomization sequence details, missing-data handling, or multiplicity procedures. It also does not provide participant-level data or a detailed analysis population beyond stating that all healthy adult subjects who completed the study were included in the statistical analysis. Those omissions limit independent reconstruction and assessment of the reported results.
The planning calculations are illustrations based on the observed effect, with assumptions stated separately in the sample-size interpretation. They are not the study’s original design calculation. General power calculations concern the probability that a specified analysis would detect a chosen effect under its assumptions; they do not tell whether a completed study had a particular probability of success after observing its results. The planning values should therefore be read as hypothetical design scenarios, not as observed power or as evidence about the validity of the sponsor analysis.
10. Confidence Intervals and Precision
The sponsor’s estimate is a least-squares mean ratio of 1.02, with a 90% confidence interval from 0.97 to 1.07. The point estimate is close to 1, but the interval is the more informative description of statistical precision: it indicates a range of relative effects compatible with the sponsor analysis and its assumptions. Precision should not be confused with whether the interval satisfies bioequivalence criteria. The acceptance limits needed for that judgment are not reported in the registry record.
The unadjusted recalculation gives a mean difference of 0.63, with an interval from -6.29 to 7.55. The interval includes values on either side of zero, so the recalculated summary comparison does not isolate a single direction of difference with high precision. The registry does not state the confidence level for this interval. It is also based on an approximate standard deviation and does not reproduce the sponsor’s least-squares ratio analysis or fully use the crossover structure.
For each formulation, the posted mean is accompanied by a standard deviation: 16.2 for FMI and 13.65 for URC. These values describe variability among outcome observations, whereas a confidence interval describes uncertainty about an estimate. The registry does not report confidence intervals for the individual arm means. No participant-level values are posted, so the distributions, outliers, and within-participant correlation cannot be examined.
The study included 36 enrolled participants, and the primary analysis population is described as subjects who completed the study and were included in the statistical analysis. The record does not separately report the number analyzed by period, missing measurements, or exclusions. Those details would help readers understand how much information contributed to the endpoint estimate. The available precision statements should consequently be read as summaries of the posted analyses, not as a complete account of all uncertainty in design, measurement, and analysis.
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.
| Effect to detect | Participants per arm, 80% power | Participants per arm, 90% power |
|---|---|---|
| 100% of the observed effect | 8,875 | 11,881 |
| 75% of the observed effect | 15,777 | 21,121 |
| 50% of the observed effect | 35,498 | 47,522 |
The sample-size entries are planning illustrations expressed as participants per arm, not total event counts. Each scenario takes a stated fraction of the observed effect as the effect to detect. The entries show the participants per arm for 80% power and for 90% power under each scenario. The 80% column means a hypothetical design would have an 80% chance of detecting the specified effect if the planning assumptions and statistical model held; the 90% column uses a higher target probability of detection.
The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The scenarios use 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. These assumptions describe the planning illustration, not the original trial protocol. The registry’s original sample-size assumptions and justification are not reported.
Smaller effects are harder to distinguish from random variation and therefore require more participants under the same planning framework. That is why the scenarios using a smaller fraction of the observed effect call for larger per-arm sample sizes. The sample-size curve should be read as depicting this relationship between the effect size selected for planning and the number of participants needed under the stated assumptions; it is not a record of enrollment or of outcomes.
These values are planning scenarios, not observed power. They cannot establish whether the completed study was adequately designed for its original objective, because the original design assumptions are not reported, and they do not replace evaluation of the sponsor’s confidence interval against prespecified equivalence limits.
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 z1−α/2 = 1.960) and the desired power (z1−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
- Observed difference Δ = 0.630; pooled standard deviation σ = 14.979.
- n = 2 × (1.96 + 0.842)2 × 14.9792 / 0.6302 = 8,874.5, rounded up to 8,875 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 <- 0.630000000000003 # observed difference in means
sigma <- 14.9793608007819 # 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(945035)
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 = 8,875 (80% power), 11,881 (90% power)
75% of observed effect: n per arm = 15,777 (80% power), 21,121 (90% power)
50% of observed effect: n per arm = 35,498 (80% power), 47,522 (90% power)
power.t.test: 8875.4
Simulated power at n = 8,875 per arm: 0.798 (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. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| not stated | Least-Squares Mean Ratio 1.02 | 0.97 to 1.07 | not reported |
The registry’s sponsor-reported comparison is a least-squares mean ratio of 1.02, with a 90% confidence interval from 0.97 to 1.07. This is the result to use when describing the sponsor’s own analysis. The registry does not identify the analysis method or provide the equivalence margins, so the ratio cannot be fully evaluated against the study’s stated bioequivalence objective from this record alone.
The separate recalculation is an unadjusted mean difference of 0.63, with an interval from -6.29 to 7.55 and an unadjusted p-value of 0.858. It is computed from posted summary statistics and uses an approximate standard deviation. Because it is on a difference scale and does not reproduce the sponsor’s least-squares ratio model, it is a supplementary summary rather than a replacement for the sponsor result. The record reports no discrepancy between the posted results and the recalculation.
These analyses should not be combined into one estimate or treated as competing versions of the same exact model. The sponsor ratio addresses relative exposure, while the recalculation expresses an absolute difference. A formal conclusion about bioequivalence would require the sponsor’s method and prespecified acceptance limits, neither of which is reported in the registry record.
14. Multiplicity, Missing Data and Analysis Populations
The registry lists two outcome measures and two analyses, but reports no secondary outcome measures. The record does not describe whether multiple hypotheses were tested, whether any multiplicity adjustment was planned, or how the two analyses relate to a prespecified testing sequence. Multiplicity is a general concern when several outcomes or comparisons are evaluated: repeated testing can increase the chance of at least one nominally notable result. Without the analysis plan, it is not possible to determine whether that issue applied or how it was addressed.
The registry states that the primary analysis population included all healthy adult subjects who completed the study and were included in the statistical analysis. It does not report the number of participants who completed each period, whether all randomized participants completed the study, or whether any endpoint data were missing. The number enrolled was 36, but enrollment alone does not establish the exact number contributing usable measurements to every analysis.
Missing data can matter if the availability of measurements is related to treatment, period, or participant characteristics. Statistical methods may address missing observations in different ways, but no missing-data procedure is reported. The registry also does not explain how incomplete crossover sequences, if any, were handled. No such incompleteness should be inferred; the point is that the record does not describe it.
The statistical method for the sponsor’s least-squares mean ratio is not reported, nor are model terms, assumptions, or diagnostic checks. The unadjusted recalculation is not a substitute for those details. The registry also does not report an interim analysis plan, and no interim analyses should be presumed. These gaps limit evaluation of the prespecification and reproducibility of the analysis, rather than demonstrating that any particular procedure was or was not used.
15. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Etoricoxib FMI | 0 | 36 | 0.0% |
| Etoricoxib URC | 0 | 36 | 0.0% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Etoricoxib FMI | 9 of 36 |
| Etoricoxib URC | 12 of 36 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Etoricoxib FMI | Etoricoxib URC |
|---|---|---|---|
| Dysgeusia | Nervous system disorders | 6 of 36 | 7 of 36 |
| Somnolence | Nervous system disorders | 3 of 36 | 4 of 36 |
| Headache | Nervous system disorders | 1 of 36 | 3 of 36 |
| Dry Mouth | Gastrointestinal disorders | 1 of 36 | 2 of 36 |
| Loose Stools | Gastrointestinal disorders | 1 of 36 | 0 of 36 |
| Nausea | Gastrointestinal disorders | 0 of 36 | 1 of 36 |
| Back Pain | Musculoskeletal and connective tissue disorders | 0 of 36 | 1 of 36 |
| Pharyngolaryngeal Pain | Respiratory, thoracic and mediastinal disorders | 0 of 36 | 1 of 36 |
The serious adverse event summary lists 0 serious adverse events among 36 participants at risk in the Etoricoxib FMI group and 0 among 36 at risk in the Etoricoxib URC group. These are the registry’s reported counts for the study groups. They describe what is posted for this study and should not be interpreted as proof that serious events could not occur.
The record also lists 9 participants with other adverse events in the FMI group and 12 in the URC group, each with 36 participants at risk. The registry categorizes these separately from serious adverse events. It does not provide event terms in the posted top-serious-terms field, and it does not report the counts of deaths. Death counts are not reported in the registry record.
These summaries do not establish that the study drug or a particular formulation caused an event. The registry record does not provide event-level timing, severity details for the other adverse events, or a causal assessment. Because this was a single-dose study in healthy subjects, the safety counts should be understood in that study context and not treated as a comprehensive assessment of safety in other populations or use conditions.
16. What the Result Does, and Does Not, Mean
The posted data describe a comparison of plasma exposure between two etoricoxib tablet formulations in healthy subjects. The sponsor reports a least-squares mean ratio of 1.02 and a 90% confidence interval from 0.97 to 1.07. The arm-level least-squares means are 32.65 for FMI and 33.28 for URC, in µg times hr/mL. These are estimates of exposure, not clinical outcomes such as pain relief.
The registry states that the study was intended to establish bioequivalence. A result close to a ratio of 1 may be consistent with similar estimated exposure, but a formal conclusion depends on predefined equivalence boundaries and the analysis method. Neither is reported in the registry record. Thus, the posted estimate and interval do not by themselves establish whether the study met the bioequivalence criterion.
The unadjusted recalculation gives a mean difference of 0.63, with an interval from -6.29 to 7.55 and an unadjusted p-value of 0.858. This is a simple comparison of posted summaries, not a recreation of the sponsor’s crossover analysis. A non-small p-value for a difference does not prove equivalence; absence of clear evidence for a difference and affirmative evidence that values lie within equivalence limits are different statistical claims.
The findings are limited to the study population and design described in the registry. Participants were healthy adults, and the measure was pharmacokinetic exposure through 120 hours postdose. The record does not report clinical effectiveness outcomes, and it does not support conclusions about treatment benefit for people with pain. Nor does the safety summary establish the absence of risk: it reports zero serious adverse events in each group and counts of other adverse events, with limited further detail.
The result is therefore best read as a registry-reported exposure comparison with an estimated ratio and stated uncertainty, accompanied by important missing information about criteria and methods. It should not be expanded into a broader claim about clinical outcomes, the original design’s adequacy, or conclusions not documented in the registry record.
17. Why This Trial Matters Statistically
This trial illustrates why a formulation comparison is not interpreted in quite the same way as a conventional test for a difference. The stated objective was bioequivalence, so the central statistical question concerns whether the relative exposure is sufficiently close under predefined criteria. A point estimate near 1 is informative, but it is not enough: the confidence interval and the boundaries chosen for the decision both matter. When the boundaries are not reported, readers cannot verify the formal decision from the registry result alone.
The crossover design also highlights the importance of matching the analysis to the design. Comparisons within participants can use information about how each person responds under both formulations. A simple recalculation from group-level summaries may not capture that paired structure. The difference-based recalculation is useful as a descriptive check, but it is not a substitute for the sponsor’s model-based ratio.
The record further shows why endpoint labels and analysis populations need careful reading. AUC is a measure of plasma exposure, not a measure of symptom improvement. The study enrolled healthy subjects, while pain is listed as a condition. That context limits how readily the pharmacokinetic results can be connected to clinical outcomes.
Finally, the numerical estimate is only one part of interpretation. The design’s original sample-size assumptions, analysis details, equivalence limits, and missing-data procedures are not reported in the registry record. Their absence does not prove that the study lacked those procedures, but it prevents readers from independently checking them. Clear reporting of these features is important for evaluating precision, reproducibility, and whether the analysis answers the stated objective.
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 registry record leaves several trial-specific details unavailable. It does not report the sponsor’s statistical model, the equivalence acceptance limits, or the original sample-size justification. Those omissions prevent a full independent assessment of whether the sponsor’s ratio and interval satisfy the intended bioequivalence criterion or whether the study was planned around that criterion.
Although the design is identified as a balanced, two-period crossover, the record does not describe treatment sequences, the interval between periods, or any washout procedure. It also does not explain how potential period or carryover effects were considered. These are design and analysis details relevant to interpreting crossover results, but their absence from the registry record does not establish that they were omitted from the study conduct or analysis.
The endpoint population is described as participants who completed the study and were included in statistical analysis. The record does not give a separate completion flow, period-specific analysis counts, or details of missing measurements and exclusions. It also provides no participant-level observations, so the distribution of AUC values and within-participant association cannot be examined.
The sponsor’s method is not stated, and the unadjusted recalculation uses summary statistics rather than the full crossover data. The two analyses therefore have different roles and cannot be used to independently verify one another. The confidence level for the recalculated difference interval is not reported. The sponsor does state a 90% confidence level for the ratio interval, but the criteria needed to interpret that interval are absent.
Finally, the study involved healthy subjects and a single-dose comparison, and the primary endpoint is pharmacokinetic exposure. The record does not post clinical effectiveness outcomes or detailed adverse-event terms. These features limit the scope of conclusions to the reported exposure comparison and the available safety counts.
19. Related Tutorials
20. Related Statistical Calculators
21. Sources
- ClinicalTrials.gov: Study of the Bioequivalence of the 20% and 30% Etoricoxib Tablet Formulations (0663-070) — NCT00945035.
- Registry record last updated 2022-02-09; status last verified 2022-02.
- Results first posted 2009-11-17.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
22. Record Summary
The registry describes a completed phase 1, randomized, open-label, two-period, single-dose, balanced crossover study in healthy subjects, with enrollment of 36. It compared 20% and 30% etoricoxib tablet formulations. The primary endpoint was plasma AUC from zero to infinity, measured through 120 hours postdose.
The sponsor-reported least-squares mean ratio is 1.02, with a 90% confidence interval from 0.97 to 1.07. The registry does not report the equivalence limits or the analysis method. An unadjusted recalculation reports a mean difference of 0.63, with an interval from -6.29 to 7.55 and an unadjusted p-value of 0.858; it is not a reconstruction of the sponsor’s crossover analysis.
The safety summary reports zero serious adverse events in each group, with 36 participants at risk per group, and 9 and 12 participants with other adverse events in FMI and URC, respectively. Death counts are not reported in the registry record. The record does not post secondary outcomes, the original sample-size justification, or full analysis details.