Exploratory Study of PENNSAID Gel to Treat Symptoms of Knee Osteoarthritis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Exploratory Study of PENNSAID Gel to Treat Symptoms of Knee Osteoarthritis (NCT01119898), a randomized phase 2 study. The primary outcome is Pain When Walking On A Flat Surface.

2. Clinical Question

This is a phase 2 randomized study with a parallel-group design and quadruple masking recorded in the registry. Its stated aim was to evaluate safety and efficacy of PENNSAID Gel in symptomatic knee osteoarthritis. Random allocation can support a fair treatment comparison when outcomes are measured and analyzed as planned, while a vehicle comparator helps distinguish the investigational product from the effects of its carrier. Those design features do not, by themselves, establish an efficacy result.

The primary outcome description is a distribution of baseline pain categories, rather than a reported change in pain or a comparison of treatment groups after treatment. The category counts are not reported in the registry record. Consequently, the record does not answer whether the groups differed at baseline, nor does it provide evidence about a treatment effect on this outcome. The registry also does not report a separate primary outcome result that could answer the stated treatment question.

3. Trial Design

This was a phase 2, randomized, double-blind, vehicle-controlled, parallel-group clinical study. The registry lists quadruple masking. The study was marked completed, with enrollment of 260. Its recorded start date is 2010-08-02 and its primary completion date is 2011-02-16. Mallinckrodt is named as the sponsor.

The two interventions were PENNSAID Gel and Vehicle. PENNSAID Gel is described as diclofenac sodium 2.0% w/w. Vehicle is described as the complete carrier, containing ingredients at the same concentrations as the experimental arm but without diclofenac sodium. A vehicle comparison is useful in principle because it separates the investigational ingredient from the carrier formulation, although the registry record does not describe the practical details of treatment administration or adherence.

Randomization assigns participants to groups by chance, reducing systematic differences in expectation when assignment is implemented as intended. A parallel-group design means the assigned groups are compared rather than each participant receiving both interventions in sequence. Masking is intended to limit effects of treatment knowledge on participant reports, care, or outcome assessment. The registry does not identify which people were included in its quadruple-masking description, so their roles should not be inferred.

The primary outcome is described as pain when walking on a flat surface at baseline, with responses classified into five ordered categories. The record also lists 10 outcome measures. It does not report comparative primary-outcome results, and the outcome entry is not reduced to one comparison. The planned analysis methods, allocation ratio, eligibility criteria, treatment duration, and details of follow-up are not reported in the registry record.

The study’s phase, randomization, comparator, masking, and enrollment describe the design, not the findings. In particular, a completed trial is not necessarily a trial with interpretable posted results for every outcome. The available record does not state a treatment effect estimate or its uncertainty for the primary outcome.

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: One person who failed screening was randomized in error, but did not receive study drug.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: mITT population

5. Baseline Characteristics

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

Population: Modified Intent-to-Treat (mITT) - all participants who were randomized and received study drug

6. Primary Endpoint

The registry lists the primary outcome as Pain When Walking On A Flat Surface, measured over at Baseline.

The primary outcome is titled “Pain When Walking On A Flat Surface.” It measures participants’ reported pain intensity while walking on a flat surface, using the categories none, mild, moderate, severe or extreme. The registry specifies the time as baseline and identifies the mITT population. It describes presenting the number of participants in each category.

Pain during a defined activity can be a clinically relevant symptom measure in a study of symptomatic knee osteoarthritis. Categorizing responses preserves more information about the distribution than reporting only a single threshold, because it distinguishes several levels of reported intensity. The ordered categories also make clear that the measurement is not a direct physical measurement: it is a participant-reported rating, interpreted within the category options in the outcome description.

Timing matters for interpretation. A baseline measurement describes the participants before any treatment effect could be assessed from a later outcome. Baseline values can characterize the study groups and may be useful context for interpreting follow-up results, but they are not themselves evidence that one intervention changed pain more than another. A treatment-effect question would ordinarily require a defined post-baseline assessment or change measure and a comparison between groups; such a primary result is not reported in the registry record.

The record gives no category counts for this endpoint. It also does not report a treatment contrast, confidence interval, or statistical test for it. Therefore, the outcome’s definition can be described, but its observed distribution and any treatment-related conclusion cannot be assessed from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does randomization contribute?

In general, randomization assigns participants to study groups by chance. If implemented as planned, it helps make treatment groups comparable on average with respect to measured and unmeasured characteristics at assignment. It does not ensure that every group characteristic will be identical in a particular trial, and it cannot substitute for reporting the outcome data. This record describes the study as randomized but does not report the allocation sequence, concealment procedures, or allocation ratio.

What does masking mean?

Masking is intended to reduce the influence of knowledge about assigned treatment on participants’ reports, clinical decisions, or outcome assessment. The registry records quadruple masking and separately describes the study title as double-blind. The record does not identify the roles covered by either description or explain how masking was maintained. It is therefore possible to report the registry’s terminology, but not to infer exactly who was unaware of assignment.

How should the endpoint scale be understood?

The pain outcome uses five named categories, from none to extreme. These are ordered categories: their labels convey a ranking of intensity, but the intervals between adjacent categories are not established as equal. A general analysis can preserve the category distribution or define a justified contrast, but those approaches answer different questions. Treating categories as though they were equally spaced requires assumptions that are not stated in the registry record. No specific statistical method for this endpoint is reported.

The registry also characterizes the outcome as a count of participants and its analysis is not reduced to one comparison because several categories or time points are involved. That is an important interpretive constraint. A set of category counts is not automatically a single treatment effect. To form a one-number contrast, an analyst would need a prespecified definition, such as a particular category threshold or an ordinal model, and the choice can affect the question being answered. No such contrast or model is reported.

Why do uncertainty measures matter?

In general, an estimate describes the observed contrast under a chosen analysis, while a confidence interval conveys its sampling uncertainty under the model assumptions. A wide interval indicates that the data are compatible with a broad range of effects; a narrow interval indicates greater precision, not necessarily clinical importance. A p-value addresses compatibility with a specified null model and is not a measure of effect size. The registry record reports no primary-outcome contrast, confidence interval, or p-value, so none of these aspects of inference can be evaluated for pain.

What is over-dispersion?

In general, over-dispersion occurs when the variability in data is greater than a statistical model expects. If a model assumes less variability than the data contain, its uncertainty estimates can be too small, making evidence appear more precise than warranted. Appropriate model choice depends on the outcome structure and observed data. The primary measure is a categorical distribution, and the record does not report a model or any over-dispersion assessment. It would not be appropriate to assume that a count model was used simply because the outcome description refers to numbers of participants in categories.

How do planning illustrations differ from trial findings?

A sample-size calculation is a planning exercise based on assumptions about the effect to detect, significance level, allocation, and analysis. It is not an observed result and does not show that a completed study had a particular probability of success. The planning note for this record describes illustrations based on a fraction of an observed effect, a two-sided significance level of 0.05, equal allocation, and a normal approximation. It specifies that non-time-to-event illustrations are expressed as participants per arm. The actual illustration values are not reported in the registry record, and the registry does not report a trial-specific design justification or sample-size assumptions.

Even if planning values are displayed alongside results, they should not be confused with the study’s enrollment or observed outcome counts. They are conditional on assumptions and are not “observed power.” Since the primary comparison and its effect estimate are not reported, the precision of the trial’s evidence on that endpoint cannot be reconstructed from the planning description.

Why distinguish descriptive counts from comparative inference?

Counts describe what was recorded in each group. A treatment comparison requires a defined estimand, a method that accounts for the outcome and study design, and uncertainty around the contrast. The adverse-event counts are descriptive in the registry record; no comparative analysis is reported for them. Likewise, baseline pain category counts, even if present, would not by themselves show a treatment effect. Keeping description, estimation, and inference separate prevents an unsupported conclusion from being drawn from group counts alone.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 10 outcome measures. The primary outcome has several categories or time points and is not reduced to one comparison. The record does not report how outcomes were prioritized beyond identifying the primary outcome, whether a multiplicity adjustment was planned, or how any adjustment would be applied. With multiple outcomes, categories, or assessment times, testing many contrasts can increase the chance of at least one apparently unusual result. In general, prespecifying a primary contrast and an approach to multiplicity helps readers understand how confirmatory claims are controlled. The registry record does not describe such an approach.

The primary outcome names the mITT population, but the registry record does not define that population or give its size. It does not report the number of participants with observed primary-outcome data, reasons for missing observations, or methods for handling missing data. Missingness can affect interpretation when the availability of measurements is related to participant characteristics or outcomes. General approaches may include assumptions about the missing-data process, but no particular approach can be attributed to this study from the record.

The statistical analysis plan, model specifications, covariate adjustments, and handling of protocol deviations are not reported in the registry record. The record also does not describe interim analyses or stopping rules. These omissions do not show that such procedures were absent; they mean the record does not allow them to be assessed. Without outcome counts and analysis details, readers cannot check whether the reported population and method match a prespecified plan or evaluate the sensitivity of conclusions to missing data and multiplicity.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent non-serious adverse event terms

The registry reports 0 serious adverse events among 130 at risk in the PENNSAID Gel group and 0 among 129 at risk in the Vehicle group. It reports 45 other adverse events among 130 at risk in the PENNSAID Gel group and 59 among 129 at risk in the Vehicle group. Deaths are reported as 0 in each group.

These are counts, not a complete comparison of safety. The registry record does not report serious-event terms, and it does not give a statistical comparison or detailed information about event collection and follow-up in the registry record for this analysis. The label “other” distinguishes those events from the serious-event count in the record; it does not establish their severity or clinical importance.

Zero recorded serious adverse events or deaths does not establish that no such risk exists. Nor do the other-event counts, by themselves, establish that one group had greater or lesser risk. Interpretation depends on how events were defined, collected, timed, and analyzed, as well as the number of participants with relevant follow-up. Those details and comparative estimates are not reported in the registry record. The figures should therefore be read as descriptive safety information, not as proof of safety or equivalence.

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

The record establishes that this was a completed, randomized, phase 2, vehicle-controlled, parallel-group study of PENNSAID Gel for symptomatic knee osteoarthritis. It identifies a primary outcome concerning pain while walking on a flat surface and records adverse-event counts in both study groups. Those details help describe the study’s question and some of the information captured.

What the record does not establish is whether PENNSAID Gel changed pain, whether any difference favored either group, or how precise a treatment estimate might be. The primary endpoint is a baseline distribution across pain categories, and its category counts are not reported. No post-baseline primary-outcome contrast, effect estimate, confidence interval, or p-value is reported in the registry record. The design alone cannot supply those results.

The adverse-event counts are also limited in what they can show. They are descriptive and lack a reported comparative analysis and detailed event terms. They do not establish that the interventions have equivalent safety profiles or that a risk is absent. The findings should not be extended beyond the outcomes and details recorded.

In general, a randomized trial can provide evidence about a treatment comparison when the relevant outcomes, analysis population, statistical methods, and uncertainty are reported. Here, the registry record gives design information and selected counts but not the primary comparative result. Readers can understand what was intended to be measured and what event counts were recorded, but cannot draw a conclusion about the magnitude or direction of an efficacy effect from this record alone.

The trial’s completion status and enrollment are details about conduct, not evidence of benefit or lack of benefit. A missing result in the registry record is not a null result. It means the treatment effect cannot be determined from the information reported there.

13. Why This Trial Matters Statistically

This study illustrates why trial design information and trial results must be read separately. Randomization, a vehicle comparator, parallel groups, and masking describe safeguards intended to support a fair comparison. They do not guarantee that a result is reported, precise, or free from bias. A reader still needs the outcome data, analysis definition, and uncertainty measures to evaluate the treatment question.

The endpoint also highlights the importance of timing and scale. A baseline distribution of ordered symptom categories can characterize participants, but it is not the same as a follow-up comparison or change from baseline. Reporting the five categories separately may preserve useful detail, while choosing a single contrast would require a clear definition and analysis plan. Without category counts or a defined treatment contrast, neither the pattern nor its statistical uncertainty can be assessed.

The record lists 10 outcome measures and does not reduce the primary outcome to one comparison. This makes transparent outcome selection and multiplicity methods important to interpretation. If several outcomes, categories, or time points are explored, readers need to know which tests were planned and how their results were interpreted. The registry record does not report those methods.

The safety information provides a further statistical lesson: counts are not the same as comparative risk estimates. Serious adverse events and deaths are reported as zero in both groups, while other adverse-event counts are reported for each group. Without event detail and an appropriate comparative analysis, those observations remain descriptive.

Overall, the record is useful for understanding the study’s design and the limits of its posted findings. Its central statistical lesson is that a trial’s question cannot be answered from design labels or baseline outcomes alone. An interpretable treatment conclusion requires reported comparative results and enough methodological detail to judge their precision and robustness.

14. 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 primary-outcome entry is a baseline measurement and gives no category counts. It does not provide a post-baseline treatment contrast, effect estimate, confidence interval, or p-value. As a result, the main efficacy question cannot be evaluated from the registry record.

The record describes the primary outcome as several categories or time points and does not reduce it to one comparison. It does not report the statistical model, contrast definition, covariate adjustment, or analysis plan. Although the mITT population is named, its definition and size are not reported. Missing-data counts, reasons for missingness, and handling methods are also not reported in the registry record.

There are 10 listed outcome measures, but the approach to multiplicity and the status of other outcomes are not described in the record. The registry does not report whether there were interim analyses or provide a trial-specific sample-size justification. Its planning note describes general assumptions for illustrations, but the actual planning values are not reported in the record summarized for this analysis.

The safety information is limited to grouped counts. Serious-event terms are not listed, and the record does not provide detail on event definitions, ascertainment, duration of exposure, or a comparative safety analysis. The registry lists quadruple masking but does not identify the roles covered by that term. These gaps constrain evaluation of the results; they should not be taken as proof that unreported procedures or information did not exist.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Exploratory Study of PENNSAID Gel to Treat Symptoms of Knee Osteoarthritis — NCT01119898.
• Registry record last updated 2019-10-22; status last verified 2019-10.
• Results first posted 2019-10-22.
• 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.

18. Record Summary

This completed phase 2 study evaluated PENNSAID Gel, described as diclofenac sodium 2.0% w/w, against a vehicle without diclofenac sodium in knee osteoarthritis. The registry records randomized, parallel-group allocation and quadruple masking, enrollment of 260, and Mallinckrodt as sponsor. The recorded start date is 2010-08-02 and primary completion date is 2011-02-16.

The primary outcome is pain when walking on a flat surface at baseline, classified as none, mild, moderate, severe or extreme in the mITT population. The registry record does not report category counts or a comparative primary-outcome result. It lists 10 outcome measures, with no single comparison specified for this multi-category or multi-time-point outcome.

Serious adverse events are reported as 0 in each group; other adverse events are reported as 45 for PENNSAID Gel and 59 for Vehicle, and deaths as 0 in each group. These are descriptive counts. The registry record does not report a comparative efficacy estimate, its uncertainty, or a detailed statistical analysis plan.