An Oral p38 Inhibitor Investigating Safety, Efficacy, And PK In Subjects With Active Rheumatoid Arthritis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for An Oral p38 Inhibitor Investigating Safety, Efficacy, And PK In Subjects With Active Rheumatoid Arthritis (NCT00383188), a randomized phase 2 study. The primary outcome is Percentage of Participants Achieving American College of Rheumatology 20 Percent (%) (ACR 20) Response at Week 12.

2. Clinical Question

3. Trial Design

The registry record describes a phase 2a, randomized, double-blind, placebo-controlled, parallel-group study of oral PH-797804 in participants with active rheumatoid arthritis. Its official title specifies a 12-week study and once-daily administration. The record lists four PH-797804 dose groups—0.5 mg, 3 mg, 6 mg, and 10 mg—and a placebo group. It reports quadruple masking. The arm descriptions are not reported in the registry record.

The record lists enrollment as 305 and status as completed. The recorded study start is 2006-12-15, and primary completion is 2008-07-01. These dates describe the registry record; they do not establish when every participant began or completed treatment. The outcome results include assessments extending to Week 16, including the primary endpoint at Week 12.

The primary endpoint analysis population was the FAS, defined in the record as participants randomized to treatment who took at least 1 dose of study medication. For this endpoint, the overall number analyzed refers to participants evaluable for the outcome, and missing values were handled using last observation carried forward. The outcome denominators therefore need to be read as the evaluable counts for that measure, not as a statement about the entire enrolled group.

The original design’s allocation proportions, sample-size justification, and statistical analysis plan are not reported in the registry record. The planning illustrations accompanying the posted results are based on observed effects, not on documented assumptions used to design the trial.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: FAS included all participants randomized to treatment and who had taken at least 1 dose of study medication. Here 'overall number of participants analyzed' signifies participants who were evaluable for this outcome measure. Missing values were imputed using last observation carried forward (LOCF).

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: The full analysis set (FAS) included all participants randomized to treatment and who had taken at least 1 dose of study medication.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants Achieving American College of Rheumatology 20 Percent (%) (ACR 20) Response at Week 12, measured over Week 12.

The primary endpoint is the percentage of participants achieving an ACR20 response at Week 12. Under the registry’s definition, a responder had at least 20% improvement from baseline in tender and swollen 28-joint counts, as well as at least 20% improvement in at least 3 of 5 additional measures: participant global assessment of arthritis, physician global assessment, participant-reported pain on a visual analogue scale, health assessment questionnaire-disability index, and C-reactive protein.

ACR20 is a composite measure: it combines changes in joint counts with changes in other measures of disease activity and participant experience. A responder percentage therefore reflects the proportion meeting the complete response rule, rather than the average change in any single measure. It is used to summarize whether participants reach a prespecified threshold of improvement across several aspects of rheumatoid arthritis.

The component measures do not all have the same scale. For example, the registry describes global assessments and pain using visual analogue scales, and the HAQ-DI as an assessment of functional disability. The response definition requires improvement across multiple components, so the endpoint does not by itself show how each component changed or how participants who did not meet the responder rule fared. The posted result is also limited to the stated assessment time; it does not establish the persistence of response beyond the recorded follow-up.

7. Results Posted in the Registry

• PH-797804, 0.5 mg: 39.1% (27/69) versus 31.1% (23/74) on control; risk difference 8.0 percentage points (95% CI -7.6 to 23.7), risk ratio 1.26 (0.80–1.97); unadjusted two-proportion z-test p = 0.313.
• PH-797804, 3 mg: 41.9% (31/74) versus 31.1% (23/74) on control; risk difference 10.8 percentage points (95% CI -4.6 to 26.2), risk ratio 1.35 (0.87–2.08); unadjusted two-proportion z-test p = 0.172.
• PH-797804, 6 mg: 40.9% (18/44) versus 31.1% (23/74) on control; risk difference 9.8 percentage points (95% CI -8.1 to 27.8), risk ratio 1.32 (0.81–2.15); unadjusted two-proportion z-test p = 0.278.
• PH-797804, 10 mg: 40.0% (16/40) versus 31.1% (23/74) on control; risk difference 8.9 percentage points (95% CI -9.6 to 27.4), risk ratio 1.29 (0.77–2.14); unadjusted two-proportion z-test p = 0.338.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Number of Participants Who Withdrew From Study Due to Lack of Efficacy

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • PH-797804, 0.5 mg: 13.0% (9/69) versus 8.0% (6/75) on control; risk difference 5.0 percentage points (95% CI -5.0 to 15.1), risk ratio 1.63 (0.61–4.34); unadjusted two-proportion z-test p = 0.322.
  • PH-797804, 3 mg: 4.1% (3/74) versus 8.0% (6/75) on control; risk difference -3.9 percentage points (95% CI -11.6 to 3.7), risk ratio 0.51 (0.13–1.95); unadjusted two-proportion z-test p = 0.312.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between PH-797804, 0.5 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between PH-797804, 3 mg and the control. The interval gives the range of effects that remain compatible with the data.

Minimum Observed Plasma Pre-dose Concentration (Ctrough Min) of Steady State

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • PH-797804, 3 mg: mean 30.80 versus 5.40 on control; difference 25.40 (95% CI 23.33 to 27.47), standardized effect d = 3.89; z-test p = <0.001.
  • PH-797804, 6 mg: mean 51.30 versus 5.40 on control; difference 45.90 (95% CI 41.18 to 50.62), standardized effect d = 4.59; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for PH-797804, 3 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for PH-797804, 6 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Number of Participants With Concomitant Medications

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • PH-797804, 0.5 mg: 100.0% (69/69) versus 98.7% (74/75) on control; risk difference 1.3 percentage points (95% CI -1.3 to 3.9), risk ratio 1.01 (0.99–1.04); unadjusted two-proportion z-test p = 0.336.
  • PH-797804, 3 mg: 98.6% (73/74) versus 98.7% (74/75) on control; risk difference -0.0 percentage points (95% CI -3.7 to 3.7), risk ratio 1.00 (0.96–1.04); unadjusted two-proportion z-test p = 0.992.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between PH-797804, 0.5 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between PH-797804, 3 mg and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes span response thresholds, joint counts, participant and physician assessments, pain, withdrawal for lack of efficacy, pharmacokinetic measurements, and concomitant medication use. This range adds descriptive context to the primary ACR20 endpoint, but it does not produce a single unified measure: the outcomes differ in what they measure, their scales, and assessment times. The registry record lists 22 posted secondary outcomes, while detailed results for outcomes not included in the record’s available result summaries are not reported in the registry record.

The available summaries include withdrawals due to lack of efficacy, minimum observed plasma pre-dose concentration, and participants with concomitant medications. These measures address different questions from whether a participant met the ACR20 response rule at Week 12. They should not be interpreted as replications of the primary endpoint or combined into an overall efficacy conclusion. The pharmacokinetic measure is descriptive of drug exposure, while the withdrawal and concomitant medication counts are separate participant-level outcomes.

Secondary outcomes can help describe a trial, but their results generally carry less weight when many outcomes are examined without a reported multiplicity adjustment. With multiple measures and time points, chance findings become more plausible, and estimates may be less precise. The registry record does not report a multiplicity strategy for secondary outcomes. Accordingly, any unadjusted secondary comparisons are best treated as exploratory context rather than independent confirmation of the primary finding. The displayed secondary results should be read according to each outcome’s stated population and timeframe, without assuming a common denominator or analysis method.

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

• Percentage of Participants Achieving American College of Rheumatology 20 Percent (%) (ACR 20) Response at Weeks 1, 2, 4, 8 and 16
• Percentage of Participants Achieving American College of Rheumatology 50 Percent (%) (ACR 50) Response at Weeks 1, 2, 4, 8, 12 and 16
• Percentage of Participants Achieving American College of Rheumatology 70 Percent (%) (ACR 70) Response at Weeks 1, 2, 4, 8, 12 and 16
• Change From Baseline in Tender/Painful Joint Count at Weeks 1, 2, 4, 8, 12 and 16
• Change From Baseline in Swollen Joint Count at Weeks 1, 2, 4, 8, 12 and 16
• Change From Baseline in Participant Assessment of Arthritis Pain at Weeks 1, 2, 4, 8, 12 and 16
• Change From Baseline in Participant Global Assessment (PGA) of Arthritis at Weeks 1, 2, 4, 8 12 and 16
• Change From Baseline in Physician Global Assessment of Arthritis at Weeks 1, 2, 4, 8, 12 and 16

9. Statistical Interpretation

10. Statistical Methods Explained

How should a responder percentage be understood?

For a binary outcome such as achieving or not achieving ACR20, the group result is the proportion of evaluable participants who met the response definition. A percentage gives this proportion on a familiar scale. It does not describe the size of improvement among responders or the pattern of component outcomes among nonresponders. In this trial, the primary result is the percentage meeting the complete ACR20 rule at Week 12.

What does a risk difference express?

A risk difference subtracts the comparator group’s event proportion from the intervention group’s event proportion. For this endpoint, it is presented as a difference in percentage points. A positive value means the observed responder percentage was higher in the named PH-797804 group than in placebo; a negative value would mean it was lower. This is an absolute contrast, rather than a relative measure such as a risk ratio. It is useful for conveying the size of the observed group separation on the same scale as the outcome percentages.

What does a confidence interval add?

A confidence interval describes uncertainty around an estimated contrast under the analysis assumptions. A narrower interval indicates greater precision; a wider interval indicates that a broader range of underlying contrasts remains compatible with the data. The intervals for the primary contrasts include the no-difference value and extend to both negative and positive values. They therefore do not rule out no difference, and they leave substantial uncertainty about the direction and magnitude of the underlying contrast. A confidence interval is not a guarantee that the true value lies within its limits.

How should the p-values be read?

The sponsor reports chi-squared analyses for the primary superiority comparisons, with percentage-point differences, 95% confidence intervals, and p-values. A p-value summarizes how unusual the observed data, or data more extreme, would be under a specified no-difference model and the test assumptions. It does not measure the probability that the treatment has no effect, and it does not show whether a difference is clinically important. The unadjusted recalculations use posted summary statistics and are not a replacement for the sponsor’s analysis.

Why does missing-data handling matter?

The registry states that missing values for the primary endpoint were imputed using last observation carried forward. This carries a participant’s last recorded observation forward to a later assessment when a value is missing. As a general statistical consideration, this method can affect estimated response proportions because it assigns a value in place of an unobserved later measurement. Its implications depend on why observations are missing and how participants’ outcomes would have changed after their last observation. Those details and sensitivity analyses are not reported in the registry record.

Why do the analysis population and denominators matter?

The stated FAS includes randomized participants who took at least 1 dose, while the endpoint denominator refers to those evaluable for the measure. The reported denominator can therefore be smaller than the number enrolled or the number assigned to an arm. Different outcomes may also have different evaluable populations. The registry’s safety group counts and primary endpoint denominators should not be assumed to refer to precisely the same analysis set.

Why do multiple comparisons require care?

The primary results compare four dose groups with placebo. When several contrasts are examined, the chance of at least one apparently unusual result can increase if each is assessed without accounting for the others. This is the general reason multiplicity procedures are considered. The registry does not report a multiplicity adjustment for these primary comparisons, so the recalculated p-values are unadjusted and the confidence intervals should be interpreted as reported, not as family-wise adjusted intervals. The statistical plan’s intended handling of multiplicity is not reported in the registry record.

What is and is not recalculated?

The recalculated binary comparisons use the posted response proportions and denominators to form simple unadjusted contrasts. They do not account for covariates, the trial’s randomization procedures beyond the recorded design, or any adjustment that might be part of a full participant-level model. No discrepancy flag is recorded for the primary comparison results. The sponsor’s own reported analyses remain the authoritative trial analyses.

11. Confidence Intervals and Precision

The sponsor-reported 95% confidence intervals show the uncertainty in each Week 12 percentage-point difference from placebo. For PH-797804, 0.5 mg, the estimate is 8.05 percentage points, with an interval from -7.565 to 23.664. For 3 mg, the estimate is 10.81 points, with an interval from -4.602 to 26.224. For 6 mg, the estimate is 9.83 points, with an interval from -8.123 to 27.799. For 10 mg, the estimate is 8.92 points, with an interval from -9.566 to 27.404.

Each interval includes the no-difference value. Consequently, the posted data do not distinguish a positive contrast from no contrast at the confidence level reported, and each interval also allows a negative contrast. The intervals are broad enough to encompass meaningfully different possible effect sizes. This is why the observed positive point estimates should not be read as precise estimates of the underlying treatment-placebo differences.

Confidence interval width is affected by the amount of information available and by the variability of the outcome. Here the outcome is a responder percentage, and the primary endpoint denominators differ across the PH-797804 groups. The smaller denominators in some groups are one relevant feature when interpreting precision, although the record does not provide a formal decomposition of the interval widths.

The intervals are estimates under the analysis method used for the reported comparisons. They do not account for unreported analysis choices, and the registry does not state whether multiplicity-adjusted intervals were planned. The p-values and confidence intervals are complementary summaries: the interval conveys a range of plausible contrasts, while the p-value summarizes evidence against the specified no-difference model. Neither establishes clinical importance or resolves uncertainty about missing outcomes.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size planning illustrations express the number of participants per arm, not total events. They use the observed effect as a starting point and show planning for 100%, 75%, and 50% of that observed effect. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The original trial’s own sample-size assumptions and justification are not reported in the registry record.

The 80% power column is the estimated participants per arm for a future study to have an 80% chance of detecting the specified effect under the planning assumptions, if that effect and assumptions were correct. The 90% power column uses a higher target chance of detection, and therefore generally calls for more participants per arm. These are planning calculations, not observed power for this completed trial.

Smaller target effects require more participants because a smaller separation between groups is harder to distinguish from sampling variability. The effect fractions are planning scenarios, not alternative results observed in this trial. The calculations also depend on the stated assumptions, including equal allocation and the normal approximation; departures from those assumptions could change the requirements. They should not be treated as the study’s actual enrollment target or evidence that the original trial was adequately or inadequately sized.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.311; treatment proportion p₁ = 0.391; average p̄ = 0.351.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6750 = 1.3230.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6726 = 0.5661.
4. n = (1.3230 + 0.5661)² / (0.0805)² = 550.8, rounded up to 551 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
p0 <- 0.310810810810811                          # control proportion
p1 <- 0.391304347826087                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(383188)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 551 (80% power), 737 (90% power)
 75% of observed effect: n per arm = 967 (80% power), 1,294 (90% power)
 50% of observed effect: n per arm = 2,145 (80% power), 2,871 (90% power)
power.prop.test: 550.8 
Simulated power at n = 551 per arm: 0.802 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The sponsor’s primary analyses are reported as chi-squared tests of the difference in percentage, with 95% confidence intervals and p-values. The reported contrasts are positive for all four PH-797804 doses, and the confidence intervals include the no-difference value. The p-values reported by the sponsor are 0.313, 0.172, 0.278, and 0.338 in ascending dose order.

The unadjusted recalculations use the posted group percentages and denominators to estimate simple risk differences and corresponding uncertainty. Their direction and broad inferential reading are consistent with the sponsor’s reported results. Small differences in displayed values can arise from rounding and from the fact that the recalculations are simple summary-statistic analyses rather than a reproduction of the sponsor’s full analysis process.

The sponsor’s analysis is the primary record of the trial’s own statistical comparison. The recalculations are supplementary checks of the posted summaries, not a substitute for the sponsor’s analysis or its underlying participant-level data. The registry does not report an adjusted analysis that would change the interpretation of these unadjusted comparisons.

15. Multiplicity, Missing Data and Analysis Populations

The primary endpoint includes four dose-versus-placebo comparisons. Evaluating several contrasts creates a multiplicity issue: without an adjustment, the chance of at least one false-positive finding across a family of tests can be greater than the error rate for a single test. This is a general statistical concern, not a claim that any particular finding in this trial is false. The registry reports chi-squared tests and confidence intervals but does not state whether a multiplicity procedure was planned or used. The recalculated p-values are unadjusted.

The registry also lists many secondary outcomes, several assessed at multiple time points. Testing numerous endpoints and time points can make isolated nominal results difficult to interpret. A prespecified hierarchy, adjustment procedure, or clear exploratory designation can guide interpretation, but these details are not reported in the registry record. The absent information means the multiplicity context cannot be reconstructed from the posted results alone.

For the primary endpoint, the registry identifies the FAS as participants randomized and taking at least 1 dose, and says that outcome denominators represent participants evaluable for the measure. It also states that missing values were imputed using last observation carried forward. The record does not explain the reasons for missingness, give sensitivity analyses for alternative missing-data assumptions, or describe how the method’s impact was assessed.

The original statistical analysis plan, covariate adjustment details, sample-size justification, interim analyses, and prespecified multiplicity handling are not reported in the registry record. These omissions limit the ability to judge whether the displayed analyses fully reflect the trial’s original analysis strategy. They do not change the arithmetic of the posted summaries, but they constrain interpretation of the results as confirmatory evidence.

16. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: myocardial infarction (2.5%), gastrointestinal haemorrhage (2.5%), colitis (2.5%), gastritis (2.5%), urinary tract infection (2.5%).

The registry reports serious adverse events in 2 of 69 participants in the PH-797804, 0.5 mg group; 3 of 74 in the 3 mg group; 4 of 44 in the 6 mg group; 5 of 40 in the 10 mg group; and 1 of 75 in the placebo group. The record also lists serious-event terms with percentages of 2.5%: myocardial infarction, gastrointestinal haemorrhage, colitis, gastritis, urinary tract infection, and gastroenteritis.

These counts describe the safety results as posted. They do not establish that a particular event was caused by study treatment, and the term list alone does not show how events were distributed across groups. The registry’s death values are not reported in the registry record. The specific adverse-event observation window and further details needed to interpret event timing are not reported in the registry record.

Safety counts should be considered separately from the ACR20 response outcome. Differences in group counts are descriptive and may reflect the different group sizes; they are not, by themselves, evidence of a comparative safety effect. The posted figures do not replace a full assessment of event severity, timing, relatedness, and participant-level clinical context.

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

The primary endpoint results show higher observed ACR20 response percentages in each PH-797804 dose group than in placebo at Week 12. The sponsor-reported percentage-point estimates are positive, but all four 95% confidence intervals include the no-difference value and extend to negative values. The data therefore leave substantial uncertainty about the size and direction of the underlying group differences. The posted results do not establish that any dose differs from placebo on this endpoint.

These results answer a specific question: the percentage meeting the registry’s ACR20 response definition at Week 12 in the stated evaluable population, with missing values handled using LOCF. They do not directly describe average changes in each component measure, response durability beyond the recorded follow-up, or outcomes outside the endpoint definition. The registry lists other outcomes, but each addresses a distinct question and may have its own analysis population and uncertainty.

The results also do not establish a dose-response pattern. The observed response percentages are not progressively higher across the four doses, and the uncertainty around each placebo contrast is substantial. A pattern in observed percentages alone is not enough to establish a biological or clinical dose relationship.

These findings should not be interpreted as medical advice or as a general conclusion about rheumatoid arthritis treatment. The trial is a phase 2a investigation with the design and analysis details recorded in the registry. The available record does not report the complete statistical plan, multiplicity handling, or sensitivity analyses for missing data. Those limitations matter when judging how broadly the posted results can be applied.

18. Why This Trial Matters Statistically

This trial illustrates why a point estimate and its precision need to be read together. The primary response percentages are higher in the PH-797804 groups than in placebo, but their confidence intervals are wide and include the no-difference value. Looking only at the direction of the point estimates would omit the central statistical uncertainty in the comparison.

It also illustrates the importance of distinguishing an endpoint-specific analysis population from overall enrollment. The record reports enrollment as 305, while the primary outcome uses evaluable denominators that vary across arms. The stated FAS and LOCF rule help explain how the endpoint was analyzed, but they do not resolve uncertainty about missing outcomes or the assumptions behind imputation.

The study includes multiple dose comparisons and a broad set of secondary outcomes. Without a reported multiplicity strategy, individual unadjusted findings require caution. Likewise, observed-effect sample-size illustrations can inform future planning, but they are not the trial’s original design rationale and do not measure the power achieved by this trial.

Finally, the registry record separates efficacy, pharmacokinetic, withdrawal, medication-use, and safety outcomes. Keeping these questions distinct supports a more disciplined reading: each result should be interpreted according to its endpoint, analysis population, timeframe, and uncertainty rather than treated as a single overall verdict.

19. 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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record does not report the full protocol or statistical analysis plan. Details about allocation procedures, original sample-size justification, interim analyses, covariate adjustment, and multiplicity handling are therefore unavailable. The record describes randomization, masking, and a parallel design, but does not provide enough detail to reconstruct all implementation or analysis choices.

The primary endpoint uses LOCF for missing values. The reasons for missing data, the amount missing by group, and sensitivity analyses under other assumptions are not reported in the registry record. Since LOCF relies on carrying a prior observation forward, the result may depend on how well that assumption represents participants’ subsequent outcomes.

The primary endpoint denominators differ among dose groups, and the placebo denominator for the endpoint differs from its safety denominator. The registry defines the endpoint denominator as the evaluable participants for that measure, but does not explain the specific reasons for these differences. This makes it important not to assume the same analysis population across outcomes.

The confidence intervals for the primary contrasts include no difference and extend in both directions, limiting precision. The registry’s detailed descriptions of the arm assignments are blank, and full results for all listed secondary outcomes are not available in the record. The safety summary does not report deaths, and the event observation window is not reported in the registry record. These gaps restrict assessment of longer-term outcomes, event timing, and broader applicability.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: An Oral p38 Inhibitor Investigating Safety, Efficacy, And PK In Subjects With Active Rheumatoid Arthritis — NCT00383188.
• Registry record last updated 2021-08-20; status last verified 2021-07.
• Results first posted 2021-08-20.
• 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.

23. Record Summary

The record describes a completed phase 2a randomized, double-blind, placebo-controlled study of once-daily oral PH-797804 in active rheumatoid arthritis. It lists four dose groups and placebo, enrollment of 305, and primary completion on 2008-07-01. The primary endpoint was ACR20 response at Week 12 in the stated FAS, with missing values imputed using LOCF.

Observed ACR20 percentages were higher in each dose group than placebo, but each reported 95% confidence interval for the percentage-point difference includes no difference. The registry does not report the complete statistical plan, multiplicity handling, or sensitivity analyses for missing data. The endpoint estimates should therefore be read as uncertain trial results, not as a definitive conclusion about treatment effects.