BMS-188667 in Children and Adolescents With Juvenile Rheumatoid Arthritis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for BMS-188667 in Children and Adolescents With Juvenile Rheumatoid Arthritis (NCT00095173), a randomized phase 3 study. The primary outcome is Time to Occurrence of Juvenile Rheumatoid Arthritis/Juvenile Idiopathic Arthritis (JRA/JIA) Disease Flare During Double-Blind Phase (Period B).

2. Clinical Question

The clinical question is therefore about the timing of a defined flare in a particular trial phase, not simply whether a participant ever had worsening disease at any point in the study. The open-label phases and double-blind withdrawal phase have different roles in the study record. Results from one phase cannot automatically be treated as interchangeable with results from another. The record also lists safety outcomes and additional measures of disease activity, but the primary statistical comparison is time to flare during Period B.

3. Trial Design

The registry describes a Phase III, multicenter, multinational, randomized withdrawal study with double masking. Its official title identifies children and adolescents with active polyarticular juvenile rheumatoid arthritis. The study is marked completed, with a reported start of 2003-12 and primary completion of 2006-06. Bristol-Myers Squibb is listed as the sponsor.

The study record distinguishes an open-label lead-in phase, Period A, a double-blind phase, Period B, and an open-label phase, Period C. The primary outcome timeframe is Period B, stated as Day 113 to Day 282. The primary comparison lists an abatacept group and a placebo group for that phase. The registry reports 60 participants for the abatacept Period B group and 62 for the placebo Period B group. Those phase-specific figures are not the same as the overall enrollment figure, and the registry does not report a complete account of who contributed data to every analysis.

The randomized withdrawal label indicates that the study design involves a withdrawal comparison, but the registry record does not report all operational details needed to reconstruct how participants entered or proceeded through each phase. In particular, the entry criteria for Period B, the allocation procedures beyond the randomized designation, and the rules governing transitions between phases are not reported in the registry record. The record identifies double masking but does not state who was masked or how masking was maintained.

The listed interventions include abatacept and placebo. The primary question is restricted to the double-blind comparison, while safety outcomes are listed across the lead-in, double-blind, and open-label phases. This phase structure matters when interpreting results: the population, exposure, and follow-up underlying one phase-specific outcome may differ from those underlying another. The record does not provide enough detail to infer that every outcome uses the same analysis population or follow-up definition.

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: 214 enrolled; in Per A, 190 treated; 24 not treated due to screening failures.170 completed Per A, 123 responders qualified to enter Per B. One subject did not enter Per B; 122 responders were randomized, 60 abatacept and 62 placebo. 36 of 47 Per A non-responders re-entered at Per C. Protocol violation occurred; 5yr old participant.

Open-Label Lead-In Phase (Period A)

Reasons for not completing this period:

Double-Blind Phase (Period B)

Reasons for not completing this period:

Open-Label Extension Phase (Period C)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All treated participants

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: All treated participants in the Lead-In Phase (Period A)

6. Primary Endpoint

The registry lists the primary outcome as Time to Occurrence of Juvenile Rheumatoid Arthritis/Juvenile Idiopathic Arthritis (JRA/JIA) Disease Flare During Double-Blind Phase (Period B), measured over Period B (Day 113 to Day 282).

The primary endpoint is time to occurrence of a confirmed JRA/JIA disease flare during Period B. The registry defines time to flare as the elapsed time between the first dose in Period B and the study day on which flare is confirmed. A flare requires a combination of worsening and limited improvement across the JRA/JIA core response variables. The stated criteria require more than 30% worsening in at least 3 of the 6 core variables and more than 30% improvement in no more than 1 of the 6 variables. Additional requirements apply when the definition relies on a global assessment, joint counts, or a surrogate marker.

This kind of endpoint is used to capture both whether an event occurs and when it occurs. The timing can be informative when some participants have a confirmed flare earlier than others, while others may not have a recorded flare during the observation period. The registry reports a median time for the placebo group and a hazard ratio for the group comparison. It does not report a median value for the abatacept group, so a median comparison between both groups cannot be described from the posted result.

As a general statistical point, time-to-event outcomes can use information about event timing and follow-up, rather than reducing all observations to a simple yes-or-no status. Their interpretation nevertheless depends on how events and incomplete follow-up are handled. Those details are not fully reported in the registry record.

7. Results Posted in the Registry

• Abatacept (Period B) vs Placebo (Period B): hazard ratio 0.31 (95% CI 0.16–0.59), reported p < 0.001 (as reported in the registry).

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 With a Juvenile Rheumatoid Arthritis/Juvenile Idiopathic Arthritis (JRA/JIA) Disease With a Flare During Double-Blind Phase (Period B)

• 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:
  • Abatacept (Period B): 20.0% (12/60) versus 53.2% (33/62) on control; risk difference -33.2 percentage points (95% CI -49.2 to -17.2), risk ratio 0.38 (0.22–0.66); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Abatacept (Period B) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists secondary outcomes covering flare occurrence, JRA/JIA core-set changes, serious adverse events, treatment-related adverse events, deaths, study-drug discontinuation due to adverse events, and events of special interest such as infections, peri-infusional events, autoimmune disorders, and malignancies. These outcomes address disease activity and safety across the double-blind and open-label phases. They provide broader context for the trial, but they do not all answer the same question or use the same analysis population.

The posted flare count during Period B is directionally consistent with the primary time-to-flare result: the registry reports 12 participants with a flare in the abatacept group and 33 in the placebo group, with a chi-squared p-value of <0.001. This secondary outcome is related to the primary endpoint but does not incorporate event timing in the same way. The other listed secondary outcomes do not have results in the record; their estimates and uncertainty are not reported in the registry record.

Secondary outcomes can help characterize different aspects of a trial, but multiple outcomes create multiple opportunities to observe apparently notable results. Unless a multiplicity strategy is specified, an unadjusted result should generally carry less inferential weight than a prespecified primary analysis. The registry record does not report a strategy for controlling multiplicity across the secondary outcomes. A small p-value for one secondary outcome should therefore not be treated as if it were an independently confirmed primary finding.

The outcomes also span Period A, Period B, and Period C. Safety and disease-activity observations across those phases may reflect different durations, treatment exposures, or participant sets. Because the registry does not fully define those analysis populations and follow-up periods, secondary findings should be interpreted within the phase and outcome to which they belong rather than combined into a single overall conclusion.

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

• Number of Participants With a Juvenile Rheumatoid Arthritis/Juvenile Idiopathic Arthritis (JRA/JIA) Disease With a Flare During Double-Blind Phase (Period B)
• Number of Participants With Serious Adverse Events (SAEs), Treatment-Related AEs, Deaths, Discontinuation of Study Drug Due to AEs During Open-Label Lead-In Phase (Period A)
• Number of Participants With Serious Adverse Events (SAEs), Treatment-Related AEs, Deaths, Discontinuation of Study Drug Due to AEs During Double-Blind Phase (Period B)
• Number of Participants With Serious Adverse Events (SAEs), Treatment-Related AEs, Deaths, Discontinuation of Study Drug Due to AEs During Open-Label Phase (Period C)
• Median Percent Change From Baseline in JRA/JIA Core Set Variables During Double-Blind Phase (Period B)
• Median Percent Change From Baseline in JRA/JIA Core Set Variables During Open-Label Phase (Period C)
• Events of Special Interest During Open-Label Lead-In Phase (Period A), Including Infections, Peri-Infusional Adverse Events (AEs), Autoimmune Disorders and Malignancies
• Events of Special Interest During Double-Blind Phase (Period B), Including Infections, Peri-Infusional Adverse Events (AEs), Autoimmune Disorders and Malignancies

9. Statistical Interpretation

10. Statistical Methods Explained

What does a time-to-event analysis measure? A time-to-event analysis studies the time from a defined starting point to a specified event. Here, the starting point is the first dose in Period B, and the event is a confirmed JRA/JIA flare. This differs from analyzing only whether a flare was recorded by the end of follow-up: timing contributes to the analysis, and participants can have different lengths of observed follow-up.

As a general explanation, time-to-event methods can include participants whose event has not been observed by the end of their available follow-up. Such observations are commonly described as censored. The registry record does not state the censoring rules used for this analysis, how withdrawals were treated, or how missing event information was handled. The estimate should therefore be read as the sponsor's reported result rather than as a fully reproducible analysis from the registry entry.

What does the hazard ratio mean? The reported hazard ratio compares the event hazard over time between the abatacept and placebo groups. A hazard is an instantaneous event rate among participants who have not yet had the event, under the model and analysis conditions. The reported estimate of 0.31 is below 1 and therefore points toward a lower hazard in the abatacept group during Period B. It is not a risk ratio, an absolute risk difference, or a comparison of median times.

As a general explanation, a hazard ratio is often interpreted under a proportional-hazards framework, in which the relative hazard is treated as reasonably stable over the comparison period. The registry identifies a log-rank test and reports a hazard ratio, but it does not report whether proportional hazards were assessed, whether the assumption was considered appropriate, or what model was used to derive the confidence interval. No assessment of that assumption can be inferred from the registry record.

What is the role of the log-rank method? The log-rank method is a general way to compare time-to-event experience between groups while taking event timing and available follow-up into account. The sponsor reports this method for the primary comparison. The registry also reports a p-value, but the full calculation details, the underlying event counts, and the participant-level follow-up data are not reported in the registry record. The analysis cannot be independently reconstructed from the posted summary alone.

Why distinguish the primary and secondary flare outcomes? The primary endpoint is time to confirmed flare. A listed secondary outcome counts participants with a flare during Period B. The count-based outcome can provide a useful view of whether flare was recorded, but it does not encode when the flare occurred in the same way as the primary endpoint. The registry's secondary analysis is identified as chi-squared. As a general explanation, a chi-squared test compares categorical counts; it does not use the detailed timing of events in the manner of a time-to-event analysis.

Why can over-dispersion matter? Over-dispersion is a general issue when a model assumes that outcome variability follows a particular distribution but the observed variability is larger than that assumption allows. It can make standard errors too small if not accommodated. This concept is not identified as a feature of the primary analysis in this record: the outcome is time to flare, and the registry reports a log-rank method. The record does not report use of a count model or an over-dispersion adjustment, so such methods should not be attributed to this trial.

What does the median contribute? A median time summarizes the time by which half of the relevant observed time-to-event distribution has experienced the event, subject to the method used to estimate it. The registry reports the placebo median as 6 months, with a full range from 0.73 to 6.87 months. The abatacept median is not reported in the registry record. A median is only one summary of the event-time distribution and need not capture all between-group differences; the hazard ratio is the sponsor's reported comparative measure.

For interpretation, the methods answer related but distinct questions: the median describes a point in a group's event-time distribution, the hazard ratio gives a relative comparison over time, and the secondary count describes how many participants had a recorded flare. These measures should not be treated as interchangeable or combined into a new estimate when the registry does not report the necessary analysis details.

11. Confidence Intervals and Precision

The sponsor-reported hazard ratio is 0.31, and its 95% confidence interval extends from 0.16 to 0.59. A confidence interval expresses statistical uncertainty around an estimated effect under the assumptions of the analysis. In general, it indicates a range of values compatible with the data and method at the stated confidence level; it is not a guarantee that the true effect lies within the interval.

The interval is entirely below 1, the value corresponding to equal hazards in the two groups. That positioning supports the same direction as the point estimate: a lower hazard in the abatacept group relative to placebo in the primary comparison. The interval is not extremely narrow, so a range of effect sizes remains compatible with the estimate. It does not establish the exact size of the relative hazard difference, and it does not directly quantify an absolute change in the chance of flare.

The placebo median is reported as 6 months, with a full range from 0.73 to 6.87 months. The registry does not report a confidence interval for that median or a median for the abatacept group. The reported range is not a confidence interval; it describes the stated full range of observed values. It therefore should not be read as a precision interval for a group median or as a comparison between treatment groups.

The registry reports a p-value for the primary log-rank analysis, but the exact number is not needed to understand the interval's direction and breadth. A p-value and confidence interval convey different aspects of the analysis: the former measures compatibility with a specified null hypothesis under the test, while the latter communicates uncertainty about the effect estimate. Neither one by itself resolves limitations in the analysis population, missing-data handling, or assumptions that are not reported in the registry record.

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 planning figures are illustrations based on the observed effect, not the study's original design calculations. The time-to-event planning entries are total event counts, not participants per arm. The other planning entries are participants per arm. The distinction matters: an event target is not a group-size target, while a per-arm participant target refers to each of two equally allocated groups in the illustration.

The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The assumed effect is a stated fraction of the observed effect. The registry's original sample-size assumptions and justification are not reported in the registry record. These planning figures therefore should not be read as a reconstruction of the trial's original design or as evidence that the trial was planned around the observed result.

The 80% power column gives the planning number of events or participants, as applicable, for an illustrative design with an 80% chance of detecting the stated effect under the assumptions. The 90% power column gives the corresponding planning quantity for a design targeting 90% power. Power is a design property under specified assumptions, not a probability that a particular result is true.

Rows planning for 100%, 75%, and 50% of the observed effect show how the required planning quantity changes as the assumed effect becomes smaller. Smaller effects need more participants or, for the time-to-event illustration, more total events. This is not observed power: the figures are forward-looking planning illustrations, not a measure of the probability that this completed trial detected its effect.

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

d is the total number of events needed (Schoenfeld's formula) with 1:1 allocation and proportional hazards.

Worked calculation for this trial

1. Hazard ratio HR = 0.310, so ln(HR) = -1.1712.
2. d = 4 × (1.96 + 0.842)² / [ln(0.310)]² = 22.9, rounded up to 23 events in total 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
hr <- 0.31                              # observed hazard ratio
events_needed <- function(power, frac = 1) ceiling(round(4 * (z_a + qnorm(power))^2 / log(hr^frac)^2, 6))   # Schoenfeld, 1:1 allocation
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: events = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(events_needed(0.80, f), big.mark = ","), format(events_needed(0.90, f), big.mark = ",")))

# Simulation: given d total events under proportional hazards, events in the treatment arm are Binomial(d, hr / (1 + hr)).
# The score (log-rank type) statistic is (D1 - d / 2) / sqrt(d / 4).
set.seed(95173)
d    <- events_needed(0.80)
reps <- 20000
D1 <- rbinom(reps, d, hr / (1 + hr))
z  <- (D1 - d / 2) / sqrt(d / 4)
cat(sprintf("Simulated power with %s events: %.3f (%d replicates)\n", format(d, big.mark = ","), mean(abs(z) > z_a), reps))

Output from running the code

100% of observed effect: events = 23 (80% power), 31 (90% power)
 75% of observed effect: events = 41 (80% power), 55 (90% power)
 50% of observed effect: events = 92 (80% power), 123 (90% power)
Simulated power with 23 events: 0.704 (20000 replicates)

The simulation uses the fact that, with d events in total and proportional hazards, the number of events in the treatment arm is binomial. It is an approximation that conditions on the total number of events, and it should land close to the target power.

14. Analyses Reported by the Sponsor

The primary analysis is reported by the sponsor as a log-rank comparison, with a hazard ratio of 0.31 and a 95% confidence interval from 0.16 to 0.59. The registry also reports the sponsor's p-value for that analysis. The primary result should be described using these sponsor-reported statistics; the record does not provide participant-level event times or sufficient detail to recreate the analysis independently.

The registry marks no discrepancy between the posted primary result and the sponsor analysis. There is therefore no reported conflict requiring a recalculated estimate to replace the sponsor's result. The posted summary does not include a recalculated primary hazard ratio, and none should be inferred from the group medians or the secondary flare counts. A time-to-event estimate cannot be reconstructed from those summaries alone.

For the secondary binary flare outcome, the registry reports the group counts and a chi-squared p-value of <0.001. Its reported effect measures include a risk difference based on the posted summary counts, but this is a different analysis from the primary log-rank comparison. The sponsor's primary hazard ratio remains the relevant statistic for the primary time-to-event question; the secondary count result is complementary rather than a substitute.

In general, a registry's reported analysis may be the most complete available summary when individual-level data and analysis details are absent. That does not remove the need to consider what the registry leaves unreported, including analysis-population definitions, censoring rules, and model checks.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study evaluates several outcomes, comparisons, or analyses. In general, the more statistical tests that are performed, the greater the opportunity for at least one result to appear unusual by chance. A prespecified testing sequence or another adjustment strategy can help clarify how confirmatory conclusions should be drawn across multiple analyses.

The registry lists 14 secondary outcomes in addition to the primary endpoint. They include flare counts, core-set measures, adverse events, deaths, discontinuations, and events of special interest. A secondary chi-squared analysis of flare occurrence is reported with p <0.001. The registry record does not report whether this result, or the other secondary analyses, was adjusted for multiplicity. It also does not report a hierarchical testing procedure or another family-wise error strategy. The secondary p-value should therefore be read as reported, without assuming that it has a confirmatory interpretation across all outcomes.

The record does not report the full statistical analysis plan, details of any interim analyses, or the original sample-size justification. These omissions do not show that such plans or procedures were absent; they mean they are not reported in the registry record. They also prevent a reader from assessing whether all analyses were prespecified or whether the reported analysis choices followed the original plan.

Analysis populations and missing data are also important. The primary population is described as all treated participants, but complete operational definitions and analysis denominators are not reported for every result. The record does not report the censoring rules for time to flare, how participants without a confirmed event were handled, how incomplete outcome data were addressed, or whether any sensitivity analyses were performed. These details can affect both the estimate and its uncertainty.

For a time-to-event endpoint, missing follow-up and censoring are especially relevant because the amount and timing of observed follow-up contribute to the analysis. For secondary outcomes, the denominator and the period of observation determine what a count represents. Without those details, the reported figures can be described, but the full implications of missingness, outcome selection, or multiplicity cannot be assessed from the registry record alone.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: arthritis (8.3%), rheumatoid arthritis (6.2%), appendicitis (3.3%), ovarian cyst (3.1%), acute lymphocytic leukaemia (3.1%).

The registry reports serious adverse event counts by phase-specific treatment group. The listed figures are 4 of 32 for abatacept in Period A only, 10 of 36 for abatacept in Period A/Period C, 9 of 60 for abatacept in Period A/Period B, and 14 of 62 for abatacept in Period A followed by placebo in Period B. These are group-specific counts with their listed denominators; they should not be added together, because the record does not establish that the groups represent nonoverlapping populations or comparable exposure periods.

The registry also lists other adverse-event counts of 22 of 32, 32 of 36, 57 of 60, and 54 of 62 for those respective groups. Death counts are not populated, so deaths are not reported in the registry record for these group summaries. The most frequent listed serious-event terms include arthritis, rheumatoid arthritis, appendicitis, ovarian cyst, acute lymphocytic leukaemia, and abscess limb. The record gives percentages for these terms but does not make their phase-specific denominators clear in the posted summary.

These figures are descriptive safety counts. Phase duration, treatment exposure, group composition, and analysis rules may differ, so raw counts should not be interpreted as direct comparative risks without the relevant denominators and follow-up context. They do not establish whether an event was caused by treatment.

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

The primary result describes a statistically estimated difference in time-to-flare experience during the double-blind Period B comparison. The reported hazard ratio of 0.31, with a 95% confidence interval from 0.16 to 0.59, points toward a lower hazard of confirmed flare in the abatacept group than in the placebo group under the sponsor's analysis. The secondary Period B flare counts also show fewer recorded flare participants in the abatacept group than in the placebo group. Both findings concern this trial's participants and the stated comparison period.

The result does not mean that every participant in one group had a longer time without flare than every participant in the other group. It does not provide an individual prediction, an absolute benefit, or a direct estimate of how long a particular participant might remain without flare. The placebo median is reported, but the abatacept median is not reported in the registry record. The hazard ratio also should not be interpreted as a percentage change in the probability of flare.

The registry result does not by itself establish the effect in other populations, disease settings, treatment durations, or clinical contexts. The study has a specific pediatric population and a randomized withdrawal design with defined phases. Results from the double-blind phase should not be blended with observations from open-label phases, where masking and treatment comparisons differ.

The record also does not provide all information needed to assess the complete analysis process. The handling of censoring and missing data, checks of statistical assumptions, original design rationale, and multiplicity approach are not reported in the registry record. These limitations affect how completely the posted result can be audited and how broadly it can be interpreted. The statistical finding is evidence about the comparison as reported; it is not, by itself, medical advice or a guarantee of an outcome for any individual.

18. Why This Trial Matters Statistically

This trial illustrates why endpoint definition and statistical method need to be read together. A flare is not defined as any worsening alone: the registry describes a rule based on several JRA/JIA core response variables, with additional requirements for certain assessment components. That definition determines what event the time-to-event analysis is estimating. Without attending to the rule, a reader could mistake the primary result for a comparison of a broader or simpler measure of worsening.

The study also demonstrates the distinction between timing and event counts. The primary endpoint compares time to confirmed flare, whereas the secondary flare outcome counts participants with a flare during Period B. The count-based result can reinforce the general pattern in the time-to-event analysis, but it is not a replacement for timing information. Treating the two measures as identical would overlook what each analysis contributes.

The randomized withdrawal structure makes phase boundaries important. Period A, Period B, and Period C are not interchangeable settings, and the primary comparison is specifically the double-blind phase. Likewise, safety counts from different phase-defined groups require attention to group composition and exposure rather than simple aggregation.

Finally, the registry provides a reported hazard ratio and confidence interval while leaving important analysis details unreported. This is a useful example of the balance between interpreting a clear summary estimate and recognizing the limits of a registry record. Statistical interpretation includes both the direction and precision of the estimate and the design, endpoint, and analysis information needed to understand its scope.

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.

Several limitations are specific to what is reported for this study. The primary outcome population is described as all treated participants, but the registry does not provide a full analysis-population definition or participant flow for the Period B comparison. The overall enrollment figure and the Period B group sizes refer to different parts of the record and do not establish the denominators used for every outcome.

The placebo median time to flare is reported, but the abatacept median is not reported in the registry record. The registry also does not give a confidence interval for the median, event counts for the primary analysis, or participant-level event times. As a result, the primary analysis cannot be reconstructed from the posted summary, and the median time-to-event distributions cannot be compared directly from the reported medians.

Key analytic details are not reported in the registry record, including censoring rules, missing-data methods, checks of the hazard-ratio assumptions, and sensitivity analyses. The original sample-size rationale, the full statistical analysis plan, interim analyses, and multiplicity procedures are also not reported. These gaps limit assessment of prespecification and robustness; they do not prove that the procedures were not conducted.

Safety results are organized by phase-specific groups, and the posted summary does not fully clarify exposure duration, overlap, or denominators for every listed serious-event term. Death counts are not reported in the group summaries. The serious adverse event and other adverse event counts are therefore descriptive and should not be treated as adjusted comparative risks.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: BMS-188667 in Children and Adolescents With Juvenile Rheumatoid Arthritis — NCT00095173.
• Registry record last updated 2017-01-18; status last verified 2016-11.
• Results first posted 2017-01-18.
• 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 registry identifies NCT00095173 as a completed Phase III randomized, double-blind withdrawal study sponsored by Bristol-Myers Squibb. It concerns abatacept in children and adolescents with active polyarticular juvenile rheumatoid arthritis. The primary endpoint is time to confirmed JRA/JIA flare during the double-blind Period B, stated as Day 113 to Day 282.

The sponsor reports a log-rank hazard ratio of 0.31 for abatacept versus placebo, with a 95% confidence interval from 0.16 to 0.59. The placebo median time to flare is 6 months; an abatacept median is not reported in the registry record. A secondary Period B flare count is also reported, while results for many other listed secondary outcomes are not reported in the registry record.

The result should be interpreted as a phase-specific time-to-event comparison. The registry does not report several details needed for a full assessment of analysis populations, missing data, multiplicity, and the original design assumptions.