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 A Two-Part 26-Week Study of Etoricoxib as Treatment for Ankylosing Spondylitis (AS) (MK-0663-108) (NCT01208207), a randomized phase 3 study. The primary outcome is Time-Weighted Average Change From Baseline in the Spinal Pain Intensity in Study Part 1: Etoricoxib 90 mg vs. Naproxen.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT01208207 |
| Phase | Phase 3 |
| Condition | Spondylitis, Ankylosing |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Triple |
| Enrollment | 1015 |
| Arms in the primary outcome | 2 |
| Lead sponsor | Organon and Co |
| Results posted | Yes |
2. Clinical Question
Population
Participants with ankylosing spondylitis were studied. The primary analysis used the Per-Protocol Population, which excluded participants with important protocol deviations that may have had a substantial effect on the primary efficacy endpoint. Further eligibility details are not reported in the registry record.
Intervention
The primary comparison concerns etoricoxib 90 mg in Part I. The study also included etoricoxib 60 mg and treatment sequences that continued or changed assigned treatment in Part II. The registry describes the trial as evaluating efficacy and tolerability across these doses.
Comparator
The primary comparator was naproxen 1000 mg in Part I. The registry also lists placebo-to-naproxen and placebo-to-etoricoxib interventions, alongside treatment sequences involving naproxen 1000 mg in Part II.
Primary question
How did the time-weighted average change from baseline in spinal pain intensity through Week 6 compare between etoricoxib 90 mg and naproxen 1000 mg? The registry’s stated primary hypothesis also describes non-inferiority comparisons of etoricoxib doses with naproxen, but the non-inferiority margin is not reported in the registry record.
The study was designed around a continuous pain measure: spinal pain intensity recorded on a visual analog scale. The registry describes additional objectives, including comparing etoricoxib 90 mg with etoricoxib 60 mg in Part I and assessing the added benefit of increasing the dose from 60 mg to 90 mg in Part II. A secondary outcome also compares the change after Week 6 for the etoricoxib 60/90 mg sequence against continued etoricoxib 60 mg among non-responders from Part I. These questions concern different treatment comparisons and periods; they should not be conflated with the primary Week 6 comparison. The posted primary estimate comes from the Per-Protocol Population, so it does not describe every participant enrolled in the trial.
3. Trial Design
The registry describes a Phase III, randomized, double-blind, active-comparator-controlled, multicenter study with a parallel model and triple masking. The study is titled as a two-part, 26-week study of etoricoxib for ankylosing spondylitis. Its status is completed. The registry lists enrollment as 1015 and identifies Organon and Co as the sponsor.
The trial had distinct treatment questions across Part I and Part II. In Part I, the primary endpoint compared etoricoxib 90 mg with naproxen 1000 mg over the period from baseline through Week 6. The registry’s stated objectives also include etoricoxib 60 mg versus naproxen 1000 mg and etoricoxib 90 mg versus etoricoxib 60 mg. For Part II, the listed treatment sequences include continuing etoricoxib 60 mg, increasing from etoricoxib 60 mg to 90 mg, continuing etoricoxib 90 mg, and continuing naproxen 1000 mg.
Several placebo-to-treatment interventions are listed, but the registry record does not give a complete account of how those interventions were scheduled or how participants moved between study parts. The allocation ratio, concealment details, and operational procedures for maintaining masking are not reported in the registry record. The official title characterizes the trial as randomized, double-blind and active-comparator-controlled; the registry’s masking field says triple.
The recorded start date is 2010-09-27, and the primary completion date is 2014-11-12. These dates describe the registry record and do not, by themselves, explain recruitment duration or the timing of individual participant follow-up. The protocol, reasons for the two-part design, and sample-size justification 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: All Patients Randomized
Only the first 6 of 7 groups are shown.
Part I
| Stage | Etoricoxib 60 mg (Part I) | Etoricoxib 90 mg (Part I) | Naproxen 1000 mg (Part I) | Etoricoxib 60 mg / 60 mg (Part II) | Etoricoxib 60 mg / 90 mg (Part II) | Etoricoxib 90 mg / 90 mg (Part II) |
|---|---|---|---|---|---|---|
| STARTED | 702 | 156 | 157 | 0 | 0 | 0 |
| COMPLETED | 632 | 145 | 142 | 0 | 0 | 0 |
| NOT COMPLETED | 70 | 11 | 15 | 0 | 0 | 0 |
Reasons for not completing this period:
| Reason | Etoricoxib 60 mg (Part I) | Etoricoxib 90 mg (Part I) | Naproxen 1000 mg (Part I) | Etoricoxib 60 mg / 60 mg (Part II) | Etoricoxib 60 mg / 90 mg (Part II) | Etoricoxib 90 mg / 90 mg (Part II) |
|---|---|---|---|---|---|---|
| Adverse Event | 19 | 2 | 6 | 0 | 0 | 0 |
| Lack of Efficacy | 21 | 3 | 2 | 0 | 0 | 0 |
| Lost to Follow-up | 6 | 1 | 0 | 0 | 0 | 0 |
| Non-Compliance With Study Drug | 1 | 0 | 1 | 0 | 0 | 0 |
| Physician Decision | 0 | 1 | 0 | 0 | 0 | 0 |
| Progressive Disease | 1 | 0 | 0 | 0 | 0 | 0 |
| Protocol Violation | 8 | 0 | 2 | 0 | 0 | 0 |
| Withdrawal by Subject | 14 | 4 | 4 | 0 | 0 | 0 |
Part II
| Stage | Etoricoxib 60 mg (Part I) | Etoricoxib 90 mg (Part I) | Naproxen 1000 mg (Part I) | Etoricoxib 60 mg / 60 mg (Part II) | Etoricoxib 60 mg / 90 mg (Part II) | Etoricoxib 90 mg / 90 mg (Part II) |
|---|---|---|---|---|---|---|
| STARTED | 0 | 0 | 0 | 314 | 318 | 145 |
| COMPLETED | 0 | 0 | 0 | 282 | 295 | 129 |
| NOT COMPLETED | 0 | 0 | 0 | 32 | 23 | 16 |
Reasons for not completing this period:
| Reason | Etoricoxib 60 mg (Part I) | Etoricoxib 90 mg (Part I) | Naproxen 1000 mg (Part I) | Etoricoxib 60 mg / 60 mg (Part II) | Etoricoxib 60 mg / 90 mg (Part II) | Etoricoxib 90 mg / 90 mg (Part II) |
|---|---|---|---|---|---|---|
| Withdrawal by Subject | 0 | 0 | 0 | 10 | 9 | 3 |
| Adverse Event | 0 | 0 | 0 | 5 | 9 | 4 |
| Lack of Efficacy | 0 | 0 | 0 | 6 | 4 | 3 |
| Lost to Follow-up | 0 | 0 | 0 | 4 | 0 | 2 |
| Non-Compliance With Study Drug | 0 | 0 | 0 | 4 | 1 | 1 |
| Physician Decision | 0 | 0 | 0 | 1 | 0 | 2 |
| Protocol Violation | 0 | 0 | 0 | 1 | 0 | 0 |
| Technical Problems | 0 | 0 | 0 | 1 | 0 | 1 |
Analysis population for the primary outcome, as defined in the registry: Per-Protocol Population - excluded participants due to important protocol deviations that may have had a substantial effect on the result of the primary efficacy endpoint.
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 | Etoricoxib 60 mg | Etoricoxib 90 mg | Naproxen 1000 mg | Total |
|---|---|---|---|---|---|
| Participants analysed | 702 | 156 | 157 | 1015 | |
| Age, Continuous | Years | 45.4 (SD 12.4) | 45.2 (SD 11.3) | 44.5 (SD 12.3) | 45.2 (SD 12.2) |
| Sex: Female, Male: Female | Participants | 209 | 45 | 41 | 295 |
| Sex: Female, Male: Male | Participants | 493 | 111 | 116 | 720 |
| Spinal Pain Intensity | mm VAS | 76.7 (SD 14.2) | 76.7 (SD 15.2) | 77.0 (SD 14.0) | 76.8 (SD 14.3) |
6. Primary Endpoint
The registry lists the primary outcome as Time-Weighted Average Change From Baseline in the Spinal Pain Intensity in Study Part 1: Etoricoxib 90 mg vs. Naproxen, measured over Baseline and up to Week 6.
| Element | Registry entry |
|---|---|
| Result type | Least Squares Mean |
| Measure of spread | 95% Confidence Interval |
| Unit | mm VAS |
| Endpoint classification | Continuous Endpoint |
The primary endpoint is the time-weighted average change from baseline in spinal pain intensity from baseline through Week 6 in Study Part I. Pain intensity was measured using a visual analog scale from 0-100 mm; lower values represent a better response. The endpoint is reported as a least-squares mean, with a 95% confidence interval for each treatment group.
The time-weighted average accounts for the timing of observations: the interval between adjacent observations is used as a weight relative to the time from randomization to the last observation in the period of interest. In general, this approach can represent change over a period when measurements are taken at different times, rather than treating every measurement as if it covered an equal duration. The registry does not report the full schedule of endpoint assessments.
Change from baseline expresses the direction and magnitude of change relative to participants’ starting measurements. Since lower pain scores are better, a negative change indicates a decrease in measured pain intensity. The endpoint is a continuous measure, so its summary describes average change rather than the number of participants who crossed a threshold for response. The registry’s primary analysis population is Per-Protocol, excluding participants with important protocol deviations that may have substantially affected the endpoint. That population definition matters when interpreting the estimate: it is not an all-enrolled-participant summary.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
For the Per-Protocol Population, the reported least-squares mean change in spinal pain intensity was -31.23 mm for etoricoxib 90 mg, with a 95% confidence interval from -34.7 to -27.76. For naproxen 1000 mg, the corresponding change was -30.59 mm, with a 95% confidence interval from -34.07 to -27.1. Both estimates are changes from baseline, and the negative values indicate decreases on the pain scale.
The sponsor’s ANCOVA comparison reports a difference in least-squares means of -0.64 mm, with a 95% confidence interval from -5.47 to 4.19. This interval describes a range of values compatible with the model and data under the analysis assumptions; it is not a probability statement that the true difference falls within that range. The reported interval includes values on either side of zero, so the estimate is not precise enough to establish a direction of difference across all values compatible with the interval. The registry does not report a p-value for this comparison.
The trial’s stated primary hypothesis concerns non-inferiority of etoricoxib versus naproxen. A confidence interval alone cannot establish whether a non-inferiority criterion was met without the prespecified margin and analysis rules. The margin is not reported in the registry record. Therefore, the numerical comparison can be described, but a definitive non-inferiority conclusion cannot be reconstructed from the posted endpoint values and interval alone.
The group-level intervals describe uncertainty around each group’s estimated least-squares mean, while the sponsor’s interval describes uncertainty around the estimated between-group difference. The latter is the directly relevant inferential result for the treatment comparison. Comparing the two group estimates in isolation is not a substitute for that contrast, and overlap or separation of group intervals is not the formal test of the comparative hypothesis.
The endpoint summarizes average change, not individual experience. It does not show how responses varied among participants, how many improved by a clinically meaningful amount, or how pain changed after Week 6. Those details are not reported in the registry record. The interpretation is also specific to the Per-Protocol Population and the stated endpoint window; it should not be extended automatically to other populations, treatment sequences, or outcomes.
9. Statistical Methods Explained
The posted primary outcome is continuous: spinal pain intensity is measured on a millimeter visual analog scale, and the result is expressed as a least-squares mean change. The sponsor’s reported comparison uses ANCOVA and gives a difference in least-squares means with a 95% confidence interval. These summaries answer related but distinct questions: the arm estimates describe average change within each treatment group, while the contrast estimates the difference between groups under the fitted model.
What does a least-squares mean represent?
In general, a least-squares mean is a model-based estimate of an average. Unlike a simple arithmetic mean, it can account for variables included in the analysis model. ANCOVA, or analysis of covariance, models an outcome while adjusting for covariates. The registry identifies ANCOVA as the method for the reported comparison, but the covariates, model specification, and adjustment details are not reported in the registry record. The posted estimates should therefore be read as the sponsor’s model-based results, without assuming which adjustments were made.
The registry labels the arm summaries as least-squares means and describes the reported contrast as a difference in least-squares means. A negative change is a decrease in pain intensity because the scale’s lower values represent a better response. For the contrast, the sign depends on the comparison’s ordering; the registry reports the estimate as -0.64 and presents its confidence interval. The confidence interval is more informative about uncertainty than the point estimate alone, because it shows a range of plausible contrast values under the analysis model.
How should the confidence interval be read?
In general, a confidence interval reflects sampling uncertainty in an estimated quantity under the statistical model and its assumptions. A narrower interval indicates greater precision than a wider interval on the same scale. It does not capture every source of uncertainty: for example, it does not correct for an unsuitable model, unmeasured bias, or selective reporting. The reported 95% confidence interval for the treatment difference extends from -5.47 to 4.19 mm. Its span is compatible with estimates in either direction relative to zero. The registry does not report a p-value for the comparison.
For non-inferiority questions, interpretation also depends on a prespecified margin: the largest difference considered consistent with non-inferiority under the trial’s rules. The registry’s brief summary states a non-inferiority hypothesis, but the margin and decision procedure are not reported in the registry record. Without them, the reported interval cannot be matched to a documented decision boundary.
Why use a time-weighted average?
The primary endpoint combines change measurements across the period from baseline through Week 6. In general, a time-weighted average gives observations weights according to the amount of time they represent. That can be useful when the intervals between assessments differ, because a measurement’s contribution reflects its place in the observed period. The registry describes weighting by the time between adjacent observations relative to the time from randomization to the last observation in the period of interest. The full visit schedule and details of handling missing measurements are not reported in the registry record.
This endpoint remains an average over a defined treatment period. It does not convey the trajectory for each participant, the distribution of changes, or the count meeting a response threshold. Those summaries are not reported in the registry record. The posted estimate also belongs to the Per-Protocol Population, which excludes participants with important protocol deviations that may have substantially affected the endpoint. In general, a per-protocol analysis focuses on participants meeting specified protocol conditions, while an intention-to-treat analysis aims to retain participants according to randomized assignment. The registry does not report an intention-to-treat estimate for this primary outcome.
What can planning illustrations tell a reader?
In general, sample-size planning uses assumptions about the effect to detect, variability, allocation, significance level, and statistical method. The registry’s planning illustration is based on the observed effect, two-sided significance level 0.05, equal allocation, and a normal approximation; the target effect is expressed as a stated fraction of the observed effect. These are planning illustrations, not observed power and not evidence that the study had a particular probability of detecting its result. The trial’s own sample-size justification and planning assumptions are not reported in the registry record. The posted planning values should therefore be understood as hypothetical design calculations rather than a retrospective assessment of the trial’s performance.
In general, the assumptions behind such illustrations are consequential. Planning from an observed effect can be optimistic because the observed estimate is uncertain, and equal allocation or a normal approximation may not represent every practical design. The registry record does not provide enough detail to assess how the illustration relates to the trial’s original design. It also does not report interim analyses, a statistical analysis plan, or further multiplicity procedures for the outcomes.
10. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| ANCOVA | Difference in Least Squares Mean -0.64 | -5.47 to 4.19 | not reported |
The registry reports the sponsor’s primary comparison as an ANCOVA difference in least-squares means of -0.64 mm, with a 95% confidence interval from -5.47 to 4.19. The arm-level estimates are also reported as least-squares means, but the inferential comparison is the sponsor’s model-based contrast. It should be distinguished from any simple recalculation based directly on posted summary statistics, which would not reproduce the sponsor’s adjustment unless the model and its inputs were available.
A separate recalculated estimate is not reported in the registry record. Accordingly, there is no posted recalculated value to compare numerically with the sponsor’s analysis. The registry identifies the sponsor’s method as ANCOVA but does not report its covariates or full model specification. A summary-statistic recalculation, if available elsewhere, would be unadjusted and would answer a narrower question than the sponsor’s adjusted analysis. The sponsor’s reported result remains the documented inferential comparison in this record.
The registry describes a non-inferiority hypothesis, but does not report the margin or complete decision rule. The confidence interval can be read as the uncertainty around the reported contrast; it cannot by itself establish the trial’s non-inferiority conclusion without that criterion. No p-value for this comparison is reported in the registry record.
11. Multiplicity, Missing Data and Analysis Populations
The record lists multiple efficacy questions, including comparisons of etoricoxib doses with naproxen, a dose comparison in Part I, and an assessment of increasing etoricoxib dose in Part II among non-responders from Part I. It also reports five outcome measures and four analyses. These multiple questions raise a general multiplicity issue: when many hypotheses are assessed, the chance of at least one apparently notable result can increase unless the analysis plan accounts for the number and relationships of the tests. The registry record does not report a multiplicity adjustment or a hierarchy for interpreting the outcomes.
The primary comparison is identified, and the sponsor’s ANCOVA result is reported for that comparison. However, the registry record does not provide a complete statistical analysis plan, the prespecified testing sequence, or details of how secondary outcomes were handled. It is therefore not possible to determine from the record whether the family of analyses was controlled for multiplicity, or how confirmatory versus exploratory analyses were distinguished.
Missing-data handling is also not reported in the registry record. The endpoint definition refers to observations through Week 6 and describes time weighting, but this does not specify how the analysis handled missed visits, withdrawal, or incomplete outcome measurements. The record does not describe imputation, assumptions about missingness, sensitivity analyses, or how participants with incomplete data entered the model. These details matter because estimates can depend on which observations are available and on assumptions used for missing values.
The primary analysis population is the Per-Protocol Population, defined as excluding participants with important protocol deviations that may have had a substantial effect on the primary efficacy endpoint. The registry does not report the full criteria for deviations, the number excluded, or an accompanying intention-to-treat analysis for this endpoint. The implications of exclusions therefore cannot be assessed fully from the registry record. The sample-size justification, interim analyses, and further statistical-plan details are not reported in the registry record.
12. 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 60 mg / Etoricoxib 60 mg | 1 | 313 | 0.3% |
| Etoricoxib 60 mg / Etoricoxib 90 mg | 7 | 319 | 2.2% |
| Etoricoxib 90 mg / Etoricoxib 90 mg | 6 | 145 | 4.1% |
| Naproxen 1000 mg / Naproxen 1000 mg | 2 | 142 | 1.4% |
| Etoricoxib 60 mg | 5 | 70 | 7.1% |
| Etoricoxib 90 mg | 0 | 10 | 0.0% |
| Naproxen 1000 mg | 0 | 14 | 0.0% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Etoricoxib 60 mg / Etoricoxib 60 mg | 55 of 313 |
| Etoricoxib 60 mg / Etoricoxib 90 mg | 50 of 319 |
| Etoricoxib 90 mg / Etoricoxib 90 mg | 24 of 145 |
| Naproxen 1000 mg / Naproxen 1000 mg | 25 of 142 |
| Etoricoxib 60 mg | 18 of 70 |
| Etoricoxib 90 mg | 6 of 10 |
Most frequent serious adverse event terms
| Event term | Organ system | Etoricoxib 60 mg / Etoricoxib 60 mg | Etoricoxib 60 mg / Etoricoxib 90 mg | Etoricoxib 90 mg / Etoricoxib 90 mg | Naproxen 1000 mg / Naproxen 1000 mg | Etoricoxib 60 mg | Etoricoxib 90 mg |
|---|---|---|---|---|---|---|---|
| Ankylosing spondylitis | Musculoskeletal and connective tissue disorders | 0 of 313 | 1 of 319 | 0 of 145 | 0 of 142 | 1 of 70 | 0 of 10 |
| Pulmonary embolism | Respiratory, thoracic and mediastinal disorders | 0 of 313 | 0 of 319 | 2 of 145 | 0 of 142 | 0 of 70 | 0 of 10 |
| Angina pectoris | Cardiac disorders | 0 of 313 | 1 of 319 | 0 of 145 | 0 of 142 | 0 of 70 | 0 of 10 |
| Left ventricular hypertrophy | Cardiac disorders | 0 of 313 | 1 of 319 | 0 of 145 | 0 of 142 | 0 of 70 | 0 of 10 |
| Glaucoma | Eye disorders | 1 of 313 | 0 of 319 | 0 of 145 | 0 of 142 | 0 of 70 | 0 of 10 |
| Gastric ulcer | Gastrointestinal disorders | 0 of 313 | 1 of 319 | 0 of 145 | 0 of 142 | 0 of 70 | 0 of 10 |
| Gastric ulcer haemorrhage | Gastrointestinal disorders | 0 of 313 | 1 of 319 | 0 of 145 | 0 of 142 | 0 of 70 | 0 of 10 |
| Death | General disorders | 0 of 313 | 1 of 319 | 0 of 145 | 0 of 142 | 0 of 70 | 0 of 10 |
| Non-cardiac chest pain | General disorders | 0 of 313 | 0 of 319 | 0 of 145 | 1 of 142 | 0 of 70 | 0 of 10 |
| Abscess | Infections and infestations | 0 of 313 | 0 of 319 | 0 of 145 | 1 of 142 | 0 of 70 | 0 of 10 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Etoricoxib 60 mg / Etoricoxib 60 mg | Etoricoxib 60 mg / Etoricoxib 90 mg | Etoricoxib 90 mg / Etoricoxib 90 mg | Naproxen 1000 mg / Naproxen 1000 mg | Etoricoxib 60 mg | Etoricoxib 90 mg |
|---|---|---|---|---|---|---|---|
| Hypertension | Vascular disorders | 18 of 313 | 15 of 319 | 8 of 145 | 8 of 142 | 4 of 70 | 0 of 10 |
| Nasopharyngitis | Infections and infestations | 17 of 313 | 21 of 319 | 7 of 145 | 4 of 142 | 0 of 70 | 0 of 10 |
| Headache | Nervous system disorders | 10 of 313 | 13 of 319 | 9 of 145 | 5 of 142 | 4 of 70 | 1 of 10 |
| Abdominal pain upper | Gastrointestinal disorders | 15 of 313 | 8 of 319 | 2 of 145 | 10 of 142 | 3 of 70 | 0 of 10 |
| Diarrhoea | Gastrointestinal disorders | 7 of 313 | 8 of 319 | 2 of 145 | 4 of 142 | 4 of 70 | 1 of 10 |
| Blood pressure increased | Investigations | 9 of 313 | 7 of 319 | 0 of 145 | 0 of 142 | 1 of 70 | 1 of 10 |
| Nausea | Gastrointestinal disorders | 5 of 313 | 3 of 319 | 3 of 145 | 1 of 142 | 4 of 70 | 0 of 10 |
| Abdominal discomfort | Gastrointestinal disorders | 4 of 313 | 7 of 319 | 0 of 145 | 1 of 142 | 0 of 70 | 0 of 10 |
| Abdominal pain | Gastrointestinal disorders | 3 of 313 | 3 of 319 | 0 of 145 | 1 of 142 | 1 of 70 | 1 of 10 |
| Cough | Respiratory, thoracic and mediastinal disorders | 1 of 313 | 6 of 319 | 1 of 145 | 0 of 142 | 0 of 70 | 1 of 10 |
Serious events with the highest arm-level rate: ankylosing spondylitis (1.4%), hip fracture (1.4%), cerebral infarction (1.4%), cerebrovascular accident (1.4%), headache (1.4%).
The registry reports serious adverse events by treatment group and gives the number at risk and the number affected. For the continued etoricoxib 60 mg sequence, it reports 1 serious event among 313 at risk; for the etoricoxib 60 mg/90 mg sequence, 7 among 319; for the continued etoricoxib 90 mg sequence, 6 among 145; and for continued naproxen 1000 mg, 2 among 142. Other listed groups are etoricoxib 60 mg, with 5 among 70; etoricoxib 90 mg, with 0 among 10; and naproxen 1000 mg, with 0 among 14. These are group-specific counts as recorded, not comparative estimates adjusted for follow-up or other factors.
The registry lists serious-event terms including ankylosing spondylitis, hip fracture, cerebral infarction, cerebrovascular accident, headache, and prostatitis, each at 1.4%. The record does not provide further detail about event timing, attribution, recurrence, or how these terms map to the treatment groups. Death counts are not reported in the registry record.
In general, adverse-event counts should be interpreted in light of group size, follow-up duration, event definitions, and how events were collected. The listed counts do not by themselves establish whether a treatment caused an event or whether groups differ reliably. No rates, adjusted comparisons, or uncertainty intervals for safety outcomes are reported in the registry record, so the figures are descriptive rather than a complete comparative safety analysis.
13. What the Result Does, and Does Not, Mean
The posted primary outcome shows decreases in average spinal pain intensity from baseline in both treatment groups over the period through Week 6. The reported least-squares mean change was -31.23 mm for etoricoxib 90 mg and -30.59 mm for naproxen 1000 mg. The sponsor’s estimated difference was -0.64 mm, with a 95% confidence interval from -5.47 to 4.19. These values describe the primary endpoint in the Per-Protocol Population.
The result does not show that every participant experienced a decrease, nor does it describe the distribution of individual changes. It also does not tell how many participants achieved a particular degree of improvement, because a responder threshold and such counts are not reported in the registry record. The average change does not establish what happened after the primary Week 6 period or whether the pattern persisted across the full study period.
The result does not, on its own, establish non-inferiority. The registry states a non-inferiority hypothesis, but the margin and full decision rules are not reported in the registry record. The confidence interval shows uncertainty around the reported contrast, but the appropriate non-inferiority interpretation depends on the prespecified margin and analysis approach. The absence of a reported p-value does not prevent reading the estimate and interval, but it does mean that no p-value should be inferred or substituted for the missing decision details.
The registry’s analysis is specific to the population and model described. It should not be treated as an estimate for all enrolled participants, because the primary analysis used a Per-Protocol Population. The ANCOVA covariates, missing-data procedures, and complete analysis plan are not reported in the registry record. The findings also do not answer every listed secondary question or provide a complete assessment of tolerability. The adverse-event figures are descriptive counts and do not establish causation or comparative risk.
In general, a trial result is best understood by combining the estimated effect, its precision, the population analyzed, the outcome definition, and the prespecified decision criteria. Here, the primary estimate and confidence interval are available, while important details needed to reconstruct the non-inferiority decision and assess robustness are not reported in the registry record.
14. Why This Trial Matters Statistically
This trial is statistically notable for pairing a continuous, time-weighted pain endpoint with an active-comparator question framed around non-inferiority. Such a question is not answered simply by observing that two group estimates appear similar. In general, non-inferiority requires a prespecified margin and a comparison whose uncertainty is assessed against that margin. The registry provides the sponsor’s contrast and confidence interval, but the margin is not reported in the registry record.
The endpoint also illustrates why a treatment contrast and its uncertainty should be read together. The group means describe average change from baseline, while the between-group contrast addresses the comparative question. The sponsor reports an ANCOVA contrast and confidence interval, but does not give the model covariates in the registry record. Readers can interpret the published estimate as the documented model-based comparison while recognizing that its adjustment details cannot be reconstructed from the record.
The two-part structure creates distinct questions across treatment periods: dose comparisons in Part I and a dose-increase question in Part II among a specified subgroup. Those questions are not interchangeable with the primary Week 6 comparison. The record’s multiple outcomes and analyses also make multiplicity relevant, yet it does not report a testing hierarchy or adjustment strategy.
Finally, the primary analysis uses a Per-Protocol Population. That choice is especially consequential for a non-inferiority question in general, because excluding participants can affect how treatment differences are estimated. The registry provides a definition at a high level but not the full deviation rules or a corresponding intention-to-treat result. These reporting gaps limit independent reconstruction and emphasize the importance of complete analysis plans, population definitions, and prespecified decision criteria when interpreting clinical trial results.
15. 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 important details unavailable for independent assessment. It does not report the non-inferiority margin, full statistical analysis plan, ANCOVA covariates, or complete decision procedure for the primary hypothesis. Without these, the reported contrast and confidence interval can be described, but the non-inferiority decision cannot be reconstructed from the record alone.
The primary result is reported for the Per-Protocol Population. The record defines this population in general terms but does not give the number excluded, the specific protocol deviations, or a parallel intention-to-treat estimate for the primary endpoint. The potential effect of exclusions therefore cannot be assessed from the posted information. Missing-data methods, sensitivity analyses, and the handling of incomplete Week 6 measurements are also not reported in the registry record.
The registry lists multiple outcomes and analyses but does not describe the testing hierarchy or multiplicity strategy. It also does not report interim analyses or the trial’s own sample-size justification. Planning illustrations based on the observed effect should not be mistaken for the original design rationale or for observed power.
For safety, the record presents serious-event counts and selected terms, but does not report comparative uncertainty estimates, event attribution, timing, or death counts. The listed treatment groups include different sequences and group sizes; the counts should not be read as adjusted comparisons or as directly comparable risks without additional context. Further details about participant eligibility, allocation procedures, assessment schedules, and adherence are not reported in the registry record. These limitations concern what can be established from the record, not whether the trial was conducted according to a protocol.
16. Related Tutorials
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: A Two-Part 26-Week Study of Etoricoxib as Treatment for Ankylosing Spondylitis (AS) (MK-0663-108) — NCT01208207.
- Registry record last updated 2024-06-18; status last verified 2022-02.
- Results first posted 2015-07-02.
- Data retrieved from ClinicalTrials.gov on 2026-11-09.
- Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.
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.
19. Record Summary
This completed Phase III, randomized, active-comparator-controlled, multicenter trial studied etoricoxib and naproxen for ankylosing spondylitis. The registry describes a two-part study, lists enrollment as 1015, and identifies Organon and Co as sponsor. Its primary endpoint was time-weighted average change from baseline in spinal pain intensity through Week 6, measured on a 0-100 mm visual analog scale.
In the Per-Protocol Population, the least-squares mean change was -31.23 mm for etoricoxib 90 mg and -30.59 mm for naproxen 1000 mg. The sponsor’s ANCOVA contrast was -0.64 mm, with a 95% confidence interval from -5.47 to 4.19. The registry states a non-inferiority hypothesis, but the margin and full decision rules are not reported in the registry record.
The registry also reports secondary outcomes and serious adverse-event counts. Missing-data methods, multiplicity procedures, the full analysis plan, and further details needed to reconstruct the primary decision are not reported in the registry record.