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 Celecoxib In Healthy Subjects (NCT00994461), a randomized phase 4 study. The primary outcome is Incidence of Gastroduodenal Ulcers.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT00994461 |
| Phase | Phase 4 |
| Condition | Healthy Volunteers |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Triple |
| Enrollment | 190 |
| Arms in the primary outcome | 3 |
| Lead sponsor | Pfizer's Upjohn has merged with Mylan to form Viatris Inc. |
| Results posted | Yes |
2. Clinical Question
Population
Healthy subjects; the primary endpoint population was the modified-safety analysis set, defined as randomized subjects who received at least one dose and had endoscopy at baseline and at the end or discontinuation of treatment.
Intervention
Celecoxib 100 mg BID.
Comparator
Loxoprofen 60 mg TID was the comparator for the stated superiority question. A placebo arm was also included.
Primary question
After 2 weeks of treatment, how did the incidence of gastroduodenal endoscopic ulcers compare between celecoxib and loxoprofen?
The registry describes a randomized, placebo-controlled study in healthy subjects. Its brief summary states that the study aimed to confirm the superiority of celecoxib over loxoprofen for the incidence of gastroduodenal endoscopic ulcers after 2 weeks. The endpoint definition counts subjects with an ulcer meeting the stated endoscopic criterion: a break in the mucosa at least 3 mm in diameter with unequivocal depth.
The posted arm results also include placebo, which supplies a descriptive context for the treatment groups. The registry lists 74 subjects for the celecoxib arm, 76 for loxoprofen and 37 for placebo in the primary endpoint results. These arm-specific counts should not be treated as a replacement for the population definition: the endpoint is specified for the modified-safety analysis set, while the record does not provide a fuller account of participant flow or exclusions.
The central question is comparative, but the posted estimates are arm-level percentages and the reported statistical test is for celecoxib versus loxoprofen. A result for that contrast does not by itself describe every secondary outcome or establish how the treatments compare for other populations or durations.
3. Trial Design
The registry identifies this as a completed, randomized, parallel-group phase 4 study of celecoxib, loxoprofen and placebo in healthy subjects. The enrollment field gives 190. The record lists the study start as 2009-11 and primary completion as 2010-04. The official title describes a randomized, placebo-controlled, double-blind study; the design fields classify masking as triple. The registry record does not explain the difference between those masking descriptions or specify which roles were masked.
The treatment arms are named celecoxib, loxoprofen and placebo. The official title gives the active-treatment regimens as celecoxib 100 mg BID and loxoprofen 60 mg TID. The record does not give further arm descriptions. It reports arm-specific primary endpoint counts of 74 for celecoxib, 76 for loxoprofen and 37 for placebo. These values are associated with the endpoint results and do not, on their own, document the full enrollment or disposition process.
The primary outcome is incidence of gastroduodenal ulcers, assessed after 2 weeks. The registry defines the endpoint in terms of subjects with ulcers identified by endoscopy and defines an ulcer using a mucosal break at least 3 mm in diameter with unequivocal depth. The stated analysis population is the modified-safety analysis set: randomized subjects who received at least one dose and underwent endoscopy at baseline and at the end or discontinuation of treatment.
The record lists six secondary outcomes, covering ulcer and erosion incidence, post-treatment endoscopic scores, numbers of ulcers and erosions per subject, and treatment-emergent all-causality GI body system adverse events. It records one reported analysis, a Cochran-Mantel-Haenszel comparison of celecoxib with loxoprofen. Details such as the randomization procedure, allocation concealment, masking roles, analysis strata and protocol deviations 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.
Overall Study
| Stage | Celecoxib | Loxoprofen | Placebo |
|---|---|---|---|
| STARTED | 76 | 76 | 37 |
| COMPLETED | 74 | 75 | 37 |
| NOT COMPLETED | 2 | 1 | 0 |
Reasons for not completing this period:
| Reason | Celecoxib | Loxoprofen | Placebo |
|---|---|---|---|
| Adverse Event | 2 | 1 | 0 |
Analysis population for the primary outcome, as defined in the registry: The modified-safety analysis set (m-SAF) consisted of all randomized subjects who received at least one dose of the study drug, and underwent endoscopy at baseline and at the end/discontinuation of treatment, as well.
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 | Celecoxib | Loxoprofen | Placebo | Total |
|---|---|---|---|---|---|
| Participants analysed | 76 | 76 | 37 | 189 | |
| Age, Continuous | Years | 56.5 (SD 9.65) | 57.8 (SD 9.38) | 59.1 (SD 9.18) | 57.5 (SD 9.45) |
| Sex: Female, Male: Female | Participants | 54 | 55 | 26 | 135 |
| Sex: Female, Male: Male | Participants | 22 | 21 | 11 | 54 |
| Helicobacter pylori (H. pylori) status: Positive | Participants | 32 | 33 | 16 | 81 |
| Helicobacter pylori (H. pylori) status: Negative | Participants | 44 | 43 | 21 | 108 |
6. Primary Endpoint
The registry lists the primary outcome as Incidence of Gastroduodenal Ulcers, measured over 2 weeks.
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | not applicable |
| Unit | Percent |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is the percentage of subjects with gastroduodenal endoscopic ulcers after 2 weeks of treatment. The registry defines an ulcer as a break in the mucosa at least 3 mm in diameter with unequivocal depth. The reported values are 1.4% for celecoxib, 27.6% for loxoprofen and 2.7% for placebo. The corresponding arm counts recorded with these results are 74, 76 and 37, respectively.
This endpoint measures the occurrence of a specified endoscopic finding rather than symptoms, healing, or a broader measure of gastrointestinal health. Endoscopy supplies a direct assessment of the gastroduodenal mucosa, while the threshold definition is intended to distinguish ulcers from less clearly defined findings. The registry also includes secondary outcomes concerning erosions and endoscopic scores, which are related but not identical measures.
Using a percentage expresses the frequency of subjects meeting the endpoint within the relevant analysis population. It does not describe the number or severity of ulcers per subject; those are addressed separately among the listed secondary outcomes. The stated population is the modified-safety analysis set, requiring treatment exposure and endoscopy at baseline and at the end or discontinuation of treatment. The record does not provide additional detail about missing endpoint assessments or how any such cases were handled.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The posted primary endpoint percentages are 1.4% for celecoxib, 27.6% for loxoprofen and 2.7% for placebo. These are arm-level results for the incidence of gastroduodenal endoscopic ulcers after 2 weeks. The registry reports a Cochran-Mantel-Haenszel analysis comparing celecoxib with loxoprofen, with p < 0.001. This is the only reported analysis attached to the primary result.
The p-value indicates that, under the assumptions of the stated test and a null hypothesis of no difference in the comparison, results at least as incompatible with that null as the observed data would be unlikely. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of a difference. The registry does not report a comparison estimate or confidence interval for this test. Consequently, the precision of the between-arm effect cannot be read from the p-value alone.
The arm percentages give descriptive context for the observed outcome. The sponsor-reported inferential test concerns celecoxib and loxoprofen, not a combined three-arm test, and the registry does not report a statistical comparison of either active treatment with placebo. The placebo percentage should therefore be read as an arm-level result, not as evidence of a tested placebo contrast.
The reported analysis uses the Cochran-Mantel-Haenszel method. The registry does not state the strata used in that analysis, so the details of any stratification cannot be assessed. Nor does the posted record describe adjustments for baseline factors, missing observations, or other analysis choices beyond the named method.
The endpoint population is defined by treatment exposure and endoscopic assessments at baseline and at treatment end or discontinuation. That definition matters when interpreting the arm results: they are not explicitly presented as an intention-to-treat analysis of all randomized subjects. The record does not give a participant flow account that would allow readers to assess exclusions or missing assessments in detail.
The displayed result is limited to a particular endoscopic outcome, a healthy-subject population and a 2-week treatment period. It does not establish what would be observed for other outcomes, populations or durations. The study record reports a statistically significant sponsor analysis for its stated comparison, but it does not provide an effect estimate with uncertainty bounds or enough analysis detail to judge all aspects of the comparison's precision.
9. Statistical Methods Explained
The primary outcome is a binary subject-level event: a subject either met the registry's endoscopic ulcer definition after 2 weeks or did not. The reported arm statistic is a percentage, meaning the frequency of subjects with the defined finding relative to the relevant arm population. In general, percentages make group frequencies easier to compare, but they do not alone convey uncertainty around those frequencies or the magnitude and precision of a contrast. The registry reports no confidence interval for the primary arm estimates.
A confidence interval is a range of values compatible with the observed data under a statistical procedure and its assumptions. It helps describe the uncertainty of an estimate; a narrower interval generally indicates greater precision than a wider interval. No confidence interval for the celecoxib-versus-loxoprofen comparison is reported in the registry record. The p-value cannot substitute for that interval because it does not show the range of plausible effect sizes.
What does the reported comparison test?
The registry names the Cochran-Mantel-Haenszel method for the comparison between celecoxib and loxoprofen. In general, this method can test an association between treatment group and outcome while accounting for defined strata. The record does not identify the strata or explain why this method was selected, so those details should not be inferred. The posted analysis gives p < 0.001, but no comparison estimate, confidence interval, or further test statistic.
The p-value is evidence about compatibility between the data and a null hypothesis under the test's assumptions. A small p-value does not state how large the difference is, how precisely it is estimated, or whether it matters for a clinical decision. Nor does it correct for testing multiple outcomes unless a multiplicity procedure is specified. The registry lists several secondary outcomes, but a multiplicity adjustment is not reported in the registry record.
Why can over-dispersion matter?
General explanation: for a count outcome, a simple probability model may assume that variation is determined by the mean count. Over-dispersion occurs when the observed variation exceeds that assumption. If it is ignored, standard errors may be too small and inferential results can appear more certain than warranted. This primary outcome, however, is reported as a percentage of subjects with an event, not as a count of multiple ulcers per subject. The registry lists counts of ulcers per subject as a secondary outcome, but it does not report a model for that outcome or an over-dispersion assessment.
How should sample-size illustrations be understood?
General explanation: a planning calculation estimates the number of participants that a future study might need under stated assumptions. It is not a measure of the probability that the completed study detected an effect, and it does not change the observed results. The planning assumptions associated with this record specify a two-sided significance level of 0.05, equal allocation between the two arms and a normal approximation. The illustrations are based on fractions of the observed effect; the record characterizes them as planning illustrations rather than as the trial's original design calculations. They should not be described as observed power.
The record does not give a sample-size justification for the original trial, and the sample-size assumptions do not establish that the completed study was designed around those calculations. Any planning illustration is conditional on its assumptions, including the assumed effect to detect and allocation. It should be interpreted separately from the observed arm percentages and the reported test.
What analysis population is represented?
The primary endpoint population is the modified-safety analysis set, defined as randomized subjects who received at least one dose and had endoscopy at baseline and at the end or discontinuation of treatment. General explanation: analysis populations define which participants contribute to a particular result, and differing population definitions can produce differing estimates. The registry gives this definition but does not provide a full account of randomized subjects, exclusions, missing endoscopies or sensitivity analyses. Those omissions limit independent assessment of how the analysis population was formed.
Overall, the available statistical information supports a narrow reading: arm-level endpoint percentages and a sponsor-reported test for the stated active-treatment comparison. The methods and outputs needed to assess uncertainty and analysis choices in greater depth are not reported in the registry record.
10. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| Cochran-Mantel-Haenszel | not reported | < 0.001 |
The registry reports a sponsor analysis comparing celecoxib with loxoprofen using the Cochran-Mantel-Haenszel method. The result is p < 0.001. The arm-level primary endpoint percentages are 1.4% for celecoxib and 27.6% for loxoprofen. No comparison estimate or confidence interval is reported alongside the test.
A recalculated comparison is not available for this count-type endpoint in the posted results. The registry's reported analysis is therefore the only inferential comparison to interpret; no recalculated estimate should be substituted for it. The record does not identify the strata used by the Cochran-Mantel-Haenszel procedure, which limits evaluation of the analysis details.
The p-value addresses the sponsor's stated comparison under the test's assumptions. It does not quantify the effect size or its precision, and the two arm percentages are descriptive results rather than a substitute for a confidence interval. No discrepancy between the sponsor analysis and a recalculation is flagged in the registry record. The sponsor result should be read as the posted analysis, with the absence of an effect estimate and confidence interval kept in view.
11. Multiplicity, Missing Data and Analysis Populations
The registry lists six secondary outcomes in addition to the primary endpoint and records one reported analysis. The secondary outcomes include ulcer incidence by location, combined ulcer and erosion measures, post-treatment endoscopic scores, numbers of ulcers and erosions per subject, and treatment-emergent all-causality GI body system adverse events. These outcomes involve different measures and should not be treated as interchangeable with the primary ulcer-incidence endpoint.
When multiple outcomes or comparisons are examined, the chance of at least one apparently notable result can increase. General explanation: multiplicity procedures can control that chance or otherwise define how results across tests should be interpreted. A plan for handling multiplicity, including any adjustment or hierarchy among outcomes, is not reported in the registry record. The primary comparison's p-value should therefore be understood as the reported result for that analysis, not as evidence that all listed outcomes were tested under a single specified multiplicity strategy.
The record identifies the modified-safety analysis set for the primary outcome: randomized subjects who received at least one dose and underwent endoscopy at baseline and at treatment end or discontinuation. It does not report an intention-to-treat analysis population, a participant flow description, the number of missing endpoint assessments, or the methods used to address missing data. Those details are important because the available endpoint percentages represent the defined analysis set, not necessarily every randomized subject.
The record also does not report sensitivity analyses examining alternative missing-data assumptions or analysis populations. It does not specify whether analyses were planned in advance or whether there were interim analyses. These omissions do not invalidate the posted result, but they constrain what can be concluded about robustness, selective reporting and the handling of incomplete observations.
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 |
|---|---|---|---|
| Celecoxib | 0 | 76 | 0.0% |
| Loxoprofen | 0 | 76 | 0.0% |
| Placebo | 0 | 37 | 0.0% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 2% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Celecoxib | 19 of 76 |
| Loxoprofen | 37 of 76 |
| Placebo | 8 of 37 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Celecoxib | Loxoprofen | Placebo |
|---|---|---|---|---|
| Gastritis erosive | Gastrointestinal disorders | 8 of 76 | 28 of 76 | 3 of 37 |
| Abdominal discomfort | Gastrointestinal disorders | 2 of 76 | 4 of 76 | 0 of 37 |
| Erosive duodenitis | Gastrointestinal disorders | 1 of 76 | 5 of 76 | 0 of 37 |
| Diarrhoea | Gastrointestinal disorders | 1 of 76 | 2 of 76 | 2 of 37 |
| Somnolence | Nervous system disorders | 1 of 76 | 4 of 76 | 0 of 37 |
| Gastric ulcer | Gastrointestinal disorders | 0 of 76 | 4 of 76 | 0 of 37 |
| Abdominal distension | Gastrointestinal disorders | 0 of 76 | 3 of 76 | 0 of 37 |
| Abdominal pain upper | Gastrointestinal disorders | 2 of 76 | 1 of 76 | 0 of 37 |
| Epigastric discomfort | Gastrointestinal disorders | 1 of 76 | 2 of 76 | 0 of 37 |
| Gastritis | Gastrointestinal disorders | 2 of 76 | 1 of 76 | 0 of 37 |
The serious adverse-event figures list 0 for celecoxib among 76 at risk, 0 for loxoprofen among 76 at risk and 0 for placebo among 37 at risk. These are the posted arm-specific serious-event counts. The death fields are null, so deaths are not reported in the registry record; a null value should not be read as a documented count of zero.
Zero recorded serious adverse events in these groups describes what appears in the safety results, but it does not establish that the underlying risk is absent. Event counts from a study describe observations during the recorded follow-up and do not, by themselves, precisely estimate uncommon risks. The registry record does not provide confidence intervals for these safety counts or additional detail about safety follow-up.
The record separately lists other adverse events affecting 19 for celecoxib, 37 for loxoprofen and 8 for placebo, with the corresponding at-risk counts. Those figures concern the other-adverse-event category and are not the serious adverse-event counts. The distinction between categories matters when reading the safety results.
13. What the Result Does, and Does Not, Mean
The registry's primary endpoint concerns the percentage of healthy subjects with gastroduodenal endoscopic ulcers after 2 weeks of treatment. The posted results are 1.4% for celecoxib, 27.6% for loxoprofen and 2.7% for placebo. The sponsor-reported comparison of celecoxib with loxoprofen has p < 0.001 using the Cochran-Mantel-Haenszel method.
These results describe an endoscopic finding in the specified analysis population and study setting. They do not measure every aspect of gastrointestinal health, symptoms, longer-term outcomes or all possible adverse effects. The listed ulcer definition is a mucosal break at least 3 mm in diameter with unequivocal depth, so the result is specific to that criterion. Erosions and other endoscopic measures are listed separately as secondary outcomes.
The comparison result does not provide a confidence interval or an effect estimate for the difference between the active-treatment arms. The p-value indicates statistical evidence under the stated test, but it does not show the magnitude or precision of the comparative effect. The placebo result is descriptive in the posted endpoint values; no placebo comparison is included in the reported analysis.
The findings should not be generalized beyond the study's stated population and treatment period without additional evidence. The participants were healthy subjects, and the primary endpoint was assessed after 2 weeks. The registry record does not establish results for people with a medical condition, different treatment durations, or outcomes not measured in this study. It also does not provide enough information to assess all potential effects of missing assessments or alternative analysis populations.
Accordingly, the result is best understood as an observed difference in arm-level ulcer percentages accompanied by a sponsor-reported statistical test for the celecoxib-versus-loxoprofen comparison. It does not by itself establish broader clinical consequences, answer every secondary question, or quantify the uncertainty around the between-arm effect.
14. Why This Trial Matters Statistically
This study illustrates why statistical interpretation requires more than reading a p-value. The registry gives distinct arm-level percentages for a defined endoscopic outcome and reports a test for a specified treatment comparison. Yet it does not report a comparison estimate or confidence interval. Readers can identify the direction of the arm-level pattern from the posted percentages, while the size and precision of the tested contrast remain incompletely described.
The endpoint also makes the analysis population and outcome definition consequential. The primary result applies to subjects meeting the modified-safety criteria and to ulcers meeting the registry's endoscopic threshold. The registry separately lists erosions, endoscopic scores and counts per subject, underscoring that related measurements can answer different questions. A percentage with an event/no-event definition is not interchangeable with a count of lesions or a broader safety outcome.
The comparison method raises a further interpretive issue: the Cochran-Mantel-Haenszel analysis is named, but its strata are not. Method labels are useful, but without relevant implementation details they do not fully explain how an analysis was conducted. The absence of multiplicity and missing-data details likewise limits assessment of the wider set of reported outcomes.
Finally, planning illustrations must be kept separate from the completed trial's evidence. They are conditional design calculations based on stated assumptions, not observed power and not a substitute for effect estimates or uncertainty intervals. Together, these features make the record a useful example of distinguishing descriptive outcomes, a hypothesis test, analysis-population definitions and the limits of what a registry result can establish.
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 does not report a confidence interval or a comparison estimate for the primary celecoxib-versus-loxoprofen analysis. Although arm-level percentages and a p-value are available, the missing interval limits assessment of the precision of the comparative effect. The record names the Cochran-Mantel-Haenszel method but does not describe its strata or other implementation details.
The stated primary analysis population is the modified-safety analysis set, requiring treatment exposure and endoscopy at baseline and at the end or discontinuation of treatment. A participant flow account, counts of missing endpoint assessments and details of exclusions are not reported in the registry record. The record also does not describe missing-data methods or sensitivity analyses for the primary endpoint.
The registry lists six secondary outcomes, but a multiplicity strategy is not reported. The record does not provide a sample-size justification for the original trial, and the planning assumptions should not be taken as that justification. Randomization procedures, allocation concealment, masking roles and protocol deviations are not described. The official title calls the study double-blind, while the design field classifies masking as triple; the record does not explain this difference.
The study population was healthy subjects and the primary assessment followed 2 weeks of treatment. The results therefore address a specific population, endpoint and period. The record does not establish what would be observed in other populations, over different durations, or for outcomes beyond those measured. Death fields are null, which means deaths are not reported in the registry record rather than that a zero count is documented.
16. Related Tutorials
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: Study Of Celecoxib In Healthy Subjects — NCT00994461.
- Registry record last updated 2021-02-02; status last verified 2021-01.
- Results first posted 2011-05-17.
- 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.
19. Record Summary
NCT00994461 is a completed randomized, parallel-group phase 4 study in healthy subjects, with celecoxib, loxoprofen and placebo arms. The registry gives enrollment as 190 and describes the primary outcome as the percentage of subjects with gastroduodenal endoscopic ulcers after 2 weeks. The primary endpoint population is the modified-safety analysis set.
Posted primary endpoint percentages are 1.4% for celecoxib, 27.6% for loxoprofen and 2.7% for placebo. The registry reports one analysis: a Cochran-Mantel-Haenszel comparison of celecoxib with loxoprofen, with p < 0.001. A comparison estimate and confidence interval are not reported.
The record lists six secondary outcomes and arm-specific serious adverse-event counts of 0 in each group. It does not report a multiplicity strategy, missing-data methods, detailed analysis strata or a full participant flow account. The findings describe the stated endoscopic endpoint in healthy subjects over the recorded treatment period and should be interpreted within those limits.