A Study of Baricitinib in Participants With Rheumatoid Arthritis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Baricitinib in Participants With Rheumatoid Arthritis (NCT04086745), a randomized phase 4 study. The primary outcome is Pooled JAJA and JAJD: Time From First Dose of Study Treatment to First Event of Venous Thromboembolism (VTE) [Combined Baricitinib Dose Versus TNF Inhibitor].

2. Clinical Question

The clinical focus is a safety comparison, not a measure of rheumatoid arthritis symptom improvement. The study is described as post-marketing, randomized, controlled, pragmatic, and Phase 3b/4 in its official title; the registered phase is Phase 4. The primary analysis pooled two studies, while the record identifies JAJD as NCT04086745. Therefore, the pooled analysis population is not the same thing as the enrollment count recorded for this individual registry entry.

The question is about the relative rate of experiencing a first event over time, not simply the number of participants who ever had an event. The recorded result estimates that relative rate using a hazard ratio. It does not, by itself, give the event count, absolute risk, or a participant’s individual probability of VTE.

3. Trial Design

The registry describes a randomized, controlled, parallel-group study with no masking. Its official title calls it a pragmatic Phase 3b/4 study, and the registered phase is Phase 4. The condition is rheumatoid arthritis. The listed enrollment is 1050, and the record status is completed. The recorded start date is 2020-02-14 and the primary completion date is 2025-05-19.

Participants were assigned to baricitinib or a TNF inhibitor. The baricitinib arms used 2 mg or 4 mg once daily, while the comparator was a TNF inhibitor, specifically adalimumab or etanercept, given according to approved local labeling. Treatment duration in the arm descriptions was tied to reaching the planned number of adjudicated primary endpoint events or a data monitoring committee recommendation for early termination. The record does not report whether early termination occurred.

The primary result is not based only on the enrolled participants listed for JAJD. It uses pooled data from JAJA and JAJD, with reported analysis group sizes of 2433 for combined baricitinib and 1207 for TNF inhibitor. For the primary analysis, events from both baricitinib dose arms were attributed to the combined group. The record says that no censoring was applied after rescue treatment or non-permitted rheumatoid arthritis treatments for this planned analysis.

The study is described as a safety comparison centered on venous thromboembolic events. Other listed outcomes include arterial thromboembolic events, major adverse cerebro-cardiovascular events, malignancy excluding nonmelanoma skin cancer, opportunistic infection, and serious infection. The registry record does not report the original sample-size justification, the study’s full statistical analysis plan, or details of any interim analyses.

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: All participants from the pooled analysis set. As per the planned analysis, for the Baricitinib 2/4 mg arm, all events from the Baricitinib 2 mg or 4 mg arms were attributed to it, with no censoring applied after initiation of rescue treatment or any non-permitted RA treatments.

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 participants from the safety analysis set.

6. Primary Endpoint

The registry lists the primary outcome as Pooled JAJA and JAJD: Time From First Dose of Study Treatment to First Event of Venous Thromboembolism (VTE) [Combined Baricitinib Dose Versus TNF Inhibitor], measured over From the first dose of study treatment up to 30 days after the last dose or the last study visit, whichever occurs first (approximately up to 6.1 years for JAJA and 5.3 years for JAJD participants).

The primary endpoint measures the time from the first dose of study treatment to a participant’s first venous thromboembolic event. VTE is a time-to-event outcome: the analysis considers both whether an event occurred and when it occurred, while accounting for follow-up that may end before an event is observed. The timeframe runs from first dose until 30 days after the last dose or the last study visit, whichever occurs first. The record describes this as approximately up to 6.1 years for JAJA participants and 5.3 years for JAJD participants.

A first-event endpoint is useful for a safety comparison because it focuses on the first occurrence of the event of interest during follow-up. A Cox proportional hazards model estimates a hazard ratio, comparing the event hazard over time between treatment groups. A hazard ratio is not the same as a ratio of event counts or a difference in absolute event risks. It also does not state how long an individual participant will remain event-free.

The outcome is titled as time to first event, with days as the unit and a median parameter designation, but the posted arm-level median values and confidence limits are null. The reported comparative statistic is instead the Cox-model hazard ratio with its 95% confidence interval. The registry record does not provide primary-event counts or an absolute risk estimate in the outcome details summarized for this analysis.

7. Results Posted in the Registry

• Baricitinib 2/4 mg vs TNF Inhibitor: hazard ratio 1.61 (95% CI 0.97–2.66) (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.

Pooled JAJA and JAJD: Time From First Dose of Study Treatment to First Event of Venous Thromboembolism (VTE) [Individual Baricitinib Dose Versus TNF Inhibitor]

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Regression, Cox); it was not recalculated.
• Result:
  • Baricitinib 2mg vs TNF Inhibitor: hazard ratio 1.70 (95% CI 0.94–3.08) (as reported in the registry).
  • Baricitinib 4mg vs TNF Inhibitor: hazard ratio 1.69 (95% CI 0.97–2.94) (as reported in the registry).
• Interpretation:
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Baricitinib 2mg vs TNF Inhibitor and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Baricitinib 4mg vs TNF Inhibitor and the control. The interval gives the range of effects that remain compatible with the data.

Pooled JAJA and JAJD: Time From First Dose of Study Treatment to First Arterial Thromboembolic Event (ATE) [Combined and Individual Baricitinib Dose Versus TNF Inhibitor]

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Regression, Cox); it was not recalculated.
• Result:
  • Baricitinib 2mg vs TNF Inhibitor: hazard ratio 1.10 (95% CI 0.41–2.98), reported p = 0.848 (as reported in the registry).
  • Baricitinib 4mg vs TNF Inhibitor: hazard ratio 1.36 (95% CI 0.57–3.24), reported p = 0.481 (as reported in the registry).
• Interpretation:
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Baricitinib 2mg vs TNF Inhibitor and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Baricitinib 4mg vs TNF Inhibitor and the control. The interval gives the range of effects that remain compatible with the data.

Pooled JAJA and JAJD: Time From First Dose of Study Treatment to First Malignancy (Excluding Nonmelanoma Skin Cancer [NMSC]) [Combined and Individual Baricitinib Dose Versus TNF Inhibitor]

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Regression, Cox); it was not recalculated.
• Result:
  • Baricitinib 2mg vs TNF Inhibitor: hazard ratio 1.53 (95% CI 0.95–2.44), reported p = 0.077 (as reported in the registry).
  • Baricitinib 4mg vs TNF Inhibitor: hazard ratio 1.14 (95% CI 0.72–1.81), reported p = 0.572 (as reported in the registry).
• Interpretation:
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Baricitinib 2mg vs TNF Inhibitor and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Baricitinib 4mg vs TNF Inhibitor and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes extend the safety evaluation beyond VTE to arterial thromboembolic events, major adverse cerebro-cardiovascular events, malignancy excluding nonmelanoma skin cancer, opportunistic infection, and serious infection. The registry also lists separate VTE comparisons for the 2 mg and 4 mg baricitinib groups against TNF inhibitor. These outcomes provide broader context, but they are not substitutes for the primary VTE comparison.

Among the secondary comparisons with numeric estimates reported, the hazard-ratio point estimates are above 1, while their confidence intervals include 1. The record gives p-values for the listed arterial event and malignancy comparisons, but no p-value for the VTE dose-specific comparisons. The estimates should be viewed as uncertain, and differences among endpoints cannot be interpreted as a coherent pattern of treatment effects from these figures alone. Results for the other listed secondary outcomes are not reported in the registry record.

Secondary outcomes are often exploratory or supportive and involve multiple comparisons. When many endpoints or treatment contrasts are examined, some apparently notable findings can arise by chance. The registry record does not report a multiplicity adjustment or a hierarchy for these secondary tests. Accordingly, unadjusted secondary results carry less weight than a prespecified primary analysis and should not be treated as independent confirmatory evidence. They can help frame questions and describe the range of outcomes monitored, but they do not resolve uncertainty in the primary comparison.

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

• Pooled JAJA and JAJD: Time From First Dose of Study Treatment to First Event of Venous Thromboembolism (VTE) [Individual Baricitinib Dose Versus TNF Inhibitor]
• Pooled JAJA and JAJD: Time From First Dose of Study Treatment to First Arterial Thromboembolic Event (ATE) [Combined and Individual Baricitinib Dose Versus TNF Inhibitor]
• Pooled JAJA and JAJD: Time From First Dose of Study Treatment to First Major Adverse Cerebro-Cardiovascular Event (MACE) [Combined and Individual Baricitinib Dose Versus TNF Inhibitor]
• Pooled JAJA and JAJD: Time From First Dose of Study Treatment to First Malignancy (Excluding Nonmelanoma Skin Cancer [NMSC]) [Combined and Individual Baricitinib Dose Versus TNF Inhibitor]
• Pooled JAJA and JAJD: Time From First Dose of Study Treatment to First Opportunistic Infection [Combined and Individual Baricitinib Dose Versus TNF Inhibitor]
• Pooled JAJA and JAJD: Time From First Dose of Study Treatment to First Serious Infection [Combined and Individual Baricitinib Dose Versus TNF Inhibitor]

9. Statistical Interpretation

10. Statistical Methods Explained

What does a hazard ratio describe?

In a time-to-event analysis, the outcome is the time until an event, such as a first VTE. Some participants may have an event during follow-up, while others may finish observation without an event being recorded. A Cox proportional hazards regression model uses the available event times and follow-up information to estimate a relative hazard between groups. The hazard concerns the instantaneous event rate among participants who have not yet experienced the event, conditional on remaining under observation.

The reported primary hazard ratio is 1.606 for combined baricitinib versus TNF inhibitor. In general, a hazard ratio of 1 represents equal modeled hazards; values above or below 1 represent a higher or lower modeled hazard in the numerator group, respectively. This is a relative measure. It does not directly say how many people experienced VTE, nor does it give the absolute risk in either group.

Why use Cox regression?

Cox regression is a common method for comparing time to an event when participants have different follow-up durations. It can use the ordering and timing of events while incorporating censored observations, meaning observations for which an event was not recorded before follow-up ended. The model is semiparametric: it estimates a relative hazard without requiring a specified baseline hazard shape. The registry identifies Cox regression as the method for the primary and listed time-to-event comparisons.

The model’s hazard ratio relies on a proportional hazards assumption: broadly, the relative hazards are treated as stable over time. Whether this assumption was assessed, and what diagnostics or alternative analyses were used, is not reported in the registry record. The result should therefore be understood as the estimate from the stated Cox model, without assuming that model checks or sensitivity analyses were performed.

What does stratification contribute?

The primary analysis was stratified by study, so the model accounted for the pooled data coming from JAJA and JAJD. It also included treatment group and the randomization stratification factors: VTE history, age/body mass index combination, prior inadequate response or intolerance to a TNF inhibitor, and geographic region. In general, accounting for such factors can align the analysis with the randomization structure and adjust the comparison for those factors. The registry does not provide further detail about the model’s coding or the handling of any factor’s categories.

The primary analysis population was all participants in the pooled analysis set. The record says that events from the 2 mg and 4 mg baricitinib arms were assigned to the combined baricitinib group. It also says no censoring was applied after rescue treatment or non-permitted rheumatoid arthritis treatments. The registry record does not provide further details on missing covariate data, loss to follow-up, or other censoring rules beyond the stated outcome timeframe and this treatment-related rule.

How should the confidence interval be read?

A confidence interval describes the statistical uncertainty around an estimated effect under the model and its assumptions. The primary 95% confidence interval is 0.969 to 2.660 around the hazard ratio of 1.606. In general, a wider interval indicates less precision than a narrower interval. Because this interval spans values below and above 1, it includes the equal-hazard value and also includes estimates indicating a higher hazard for baricitinib. The registry does not report a p-value for this primary comparison.

What is the role of non-inferiority?

The posted primary analysis is labeled non-inferiority. A non-inferiority analysis generally evaluates whether an intervention is not worse than a comparator by more than a prespecified margin. That margin is essential to interpreting the result: it defines how large a difference would still meet the criterion. The registry record does not report the margin, the formal decision rule, or whether that criterion was met. The hazard ratio and interval should not be used to infer those missing details.

What do the planning illustrations represent?

The event-count illustrations use the observed effect as a planning reference and show hypothetical totals for 80% and 90% power under stated assumptions. For time-to-event comparisons, the displayed quantities are total events, not participants per arm. General statistical reasoning explains why more information is needed to detect smaller effects: when the target effect is closer to no difference, it takes more events to distinguish it from random variation. These illustrations are not the original trial design justification, which is not reported in the registry record, and they are not observed power.

11. Confidence Intervals and Precision

The primary hazard ratio is 1.606, and its 95% confidence interval extends from 0.969 to 2.660. This interval indicates that the point estimate is not highly precise: a range of materially different relative hazard values remains compatible with the model and observed data. In general, the interval is more informative than the point estimate alone because it makes uncertainty visible.

The interval crosses 1, the value representing equal hazards. Thus the posted interval includes values consistent with a lower hazard for baricitinib as well as values consistent with a higher hazard. The registry record does not report a p-value for the primary estimate. It would be inappropriate to treat the point estimate alone as a conclusive result, or to treat the interval as specifying the probability that the true effect lies within its limits.

Precision is affected by the amount of event information as well as the analysis population and model. The registry lists pooled analysis group sizes, but does not report the primary VTE event count in the outcome details summarized for this comparison. The enrollment count for the individual registry entry should not be substituted for the pooled analysis sizes. Since the outcome is time-to-event, participant counts alone do not fully describe the information available to estimate the hazard ratio.

The confidence interval is model-based and depends on assumptions, including the Cox model’s proportional hazards structure. Assessment of those assumptions is not reported in the registry record. Nor is the non-inferiority margin reported, so the interval cannot be evaluated against that margin from the information recorded. The interval expresses uncertainty about the relative hazard; it does not establish absolute event rates or the likely effect for an individual participant.

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 illustrations are planning calculations based on the observed effect, not the study’s original design justification. Because these are time-to-event comparisons, the figures are total event counts across the comparison, not participants per arm. The 80% power column gives the hypothetical total number of events associated with an 80% chance of detecting the stated effect under the calculation assumptions. The 90% power column gives the corresponding event total for a 90% chance. These are planning concepts, not guarantees about a future study’s result.

The calculations use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect. The rows therefore consider planning for the full observed effect, planning for 75% of the observed effect, and planning for 50% of the observed effect. Smaller effects require more information to detect, which is why the event totals increase as the target effect becomes a smaller fraction of the observed effect.

For these time-to-event illustrations, the requested information is expressed as events rather than participant counts because event occurrence supplies the key information for the comparison. The number of participants needed to accumulate an event total can depend on event frequency and follow-up, so the event-count figures should not be read as enrollment targets. The registry record does not report the original sample-size justification or the assumptions used to set the trial’s planned enrollment. These calculations are not observed power.

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 = 1.606, so ln(HR) = 0.4737.
2. d = 4 × (1.96 + 0.842)² / [ln(1.606)]² = 139.9, rounded up to 140 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 <- 1.606                              # 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(4086745)
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 = 140 (80% power), 188 (90% power)
 75% of observed effect: events = 249 (80% power), 333 (90% power)
 50% of observed effect: events = 560 (80% power), 750 (90% power)
Simulated power with 140 events: 0.798 (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 comparison is reported as a Cox regression hazard ratio of 1.606, with a 95% confidence interval from 0.969 to 2.660. This is the sponsor’s reported model result in the registry record and is the estimate used for interpretation. There is no separate unadjusted recalculated hazard ratio presented for the primary outcome, so there is no competing re-calculation to compare with the sponsor’s analysis.

The reported model is described as stratified by study and adjusted for treatment group and the randomization stratification factors. A simple recalculation from summary event counts would not reproduce that model because the record does not provide the primary event counts or participant-level event times needed to fit the Cox regression. The listed planning illustrations are a different calculation: they use the observed effect to illustrate hypothetical event requirements and do not replace the sponsor’s outcome analysis.

The primary comparison has no reported p-value. The registry labels it as non-inferiority, but the margin and formal criterion are not reported in the registry record. As a result, the posted hazard ratio and confidence interval can be described directly, but a determination about meeting the non-inferiority criterion cannot be reconstructed from the available details.

15. Multiplicity, Missing Data and Analysis Populations

The record lists a primary VTE comparison and several secondary safety outcomes, including additional comparisons by baricitinib dose. Testing or estimating effects across multiple endpoints and contrasts creates a multiplicity issue: the more comparisons considered, the greater the opportunity for chance findings to appear noteworthy. This general concern applies even when each individual analysis uses a conventional statistical method.

The registry record does not report a multiplicity adjustment, a testing hierarchy, or a strategy for controlling the family-wise error rate or false discovery rate. It also does not describe how secondary endpoints were designated as confirmatory or exploratory. Without that information, secondary p-values should be interpreted cautiously and in the context of the overall set of comparisons, rather than as stand-alone confirmation.

The primary outcome population is described as all participants in the pooled analysis set. The model pooled JAJA and JAJD, stratified by study, and attributed events from both baricitinib dose groups to the combined group. The record specifies that no censoring was applied after rescue treatment or non-permitted rheumatoid arthritis treatments for the planned primary analysis. Other missing-data procedures, censoring details, and sensitivity analyses are not reported in the registry record.

The arm descriptions refer to treatment continuing until the planned number of adjudicated primary endpoint events occurred or the data monitoring committee recommended early termination. That description does not establish whether an interim analysis took place or whether the study stopped early. Interim-analysis methods and any associated statistical adjustments are not reported. The non-inferiority margin and complete decision rule are also not reported, limiting independent evaluation of the primary non-inferiority claim.

16. 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 serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pneumonia (2.9%), covid-19 (2.7%), prostate cancer (2.7%), acute kidney injury (1.8%), deep vein thrombosis (1.5%).

The registry reports serious adverse-event figures by treatment group: 78 of 340 participants at risk in the baricitinib 2 mg group, 79 of 333 in the baricitinib 4 mg group, and 68 of 327 in the TNF inhibitor group are recorded as affected. The record also lists other adverse-event figures of 30 of 340, 28 of 333, and 35 of 327, respectively. Death counts are recorded as 6, 6, and 5 for those groups. These are descriptive counts; they do not by themselves establish that treatment caused an event.

The listed serious-event terms include pneumonia at 2.9%, Covid-19 at 2.7%, prostate cancer at 2.7%, acute kidney injury at 1.8%, deep vein thrombosis at 1.5%, and prostate cancer stage ii at 1.4%. The registry record does not provide further context in these figures about event timing, duration, relatedness, or how overlapping events were handled. These summaries should be read as reported safety information, not as a direct comparison of causal risk or as an individual risk estimate.

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

The primary result is a model-based comparison of time to a first VTE in pooled data from JAJA and JAJD. The reported hazard ratio is 1.606, with a 95% confidence interval from 0.969 to 2.660. The point estimate is above 1, while the interval includes 1 and values on both sides. The registry does not report a p-value for this primary result.

These data describe the estimate and its uncertainty; they do not establish an absolute probability of VTE for either group. The outcome details do not report the primary event count or arm-specific event risks. The result also does not describe other dimensions of rheumatoid arthritis care, such as symptom response. It addresses the recorded time-to-first-VTE comparison within the stated analysis and follow-up framework.

The analysis is labeled non-inferiority, but the registry record does not report the margin or formal criterion. Therefore, the posted estimate cannot be used on its own to conclude that non-inferiority was established. In general, a non-inferiority conclusion depends on whether the confidence interval meets a prespecified boundary, not merely on whether the point estimate appears close to equal hazards.

The pooled nature of the analysis is also important. The primary groups include participants from JAJA and JAJD, with study included in the stratified model. The listed enrollment of 1050 pertains to the individual registry entry, whereas the primary pooled analysis groups are reported separately. The result should not be read as if it came only from the enrolled JAJD population.

More broadly, a registry result is not a substitute for a complete statistical report. Details not reported in the registry record include the original sample-size justification, the full multiplicity strategy, missing-data procedures beyond the stated censoring rule, interim-analysis methods, and proportional-hazards diagnostics. The interpretation is consequently limited to the analysis and details that are recorded, without extending to medical advice or an individual treatment decision.

18. Why This Trial Matters Statistically

This study is statistically notable because it addresses a safety outcome that is naturally analyzed over time. Time-to-event methods use the timing of a first event and the length of observation, rather than treating every participant as if they had identical follow-up. That makes the hazard ratio a relevant summary for the registry’s VTE question, while also requiring careful explanation of what a relative hazard does and does not measure.

The analysis pools two studies and stratifies by study, alongside the listed randomization factors. This structure is central to interpreting the result: the primary estimate is not a raw comparison of totals from a single registry cohort. The posted group sizes likewise belong to the pooled analysis, not only to the individual JAJD registry entry.

The primary confidence interval communicates substantial uncertainty and spans the equal-hazard value. The registry’s non-inferiority label makes the missing margin and decision rule especially consequential. Without those details, the result can be described statistically but the formal non-inferiority conclusion cannot be independently assessed from the record alone.

The secondary safety outcomes broaden the scope of observation, but multiple comparisons make it important to distinguish supportive results from confirmatory evidence. The absence of a reported multiplicity strategy and full analysis plan limits how strongly isolated secondary findings can be interpreted. These features make transparent reporting of event counts, censoring, model assumptions, and prespecified criteria important for understanding the evidence.

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.

The primary pooled analysis includes JAJA and JAJD, but the registry entry is for JAJD. The record reports a JAJD enrollment of 1050 and separate pooled analysis group sizes; these populations should not be conflated. The primary outcome details do not include the VTE event counts or arm-level median times, limiting interpretation of absolute event occurrence alongside the relative hazard estimate.

The primary hazard ratio has a 95% confidence interval from 0.969 to 2.660, and no primary p-value is reported. The record labels the analysis non-inferiority but does not report the margin or formal decision rule. These omissions prevent independent assessment of whether the prespecified non-inferiority criterion was met.

The registry does not report the original sample-size justification, the full analysis plan, multiplicity handling, interim-analysis methods, or diagnostics for the Cox proportional hazards assumption. Details of missing-data handling and censoring beyond the stated rule concerning rescue or non-permitted rheumatoid arthritis treatments are not reported. Results for some listed secondary outcomes are also not reported in the registry record.

Finally, the posted adverse-event summaries are descriptive. They do not supply the detail needed to assess event timing, relatedness, or individual risk. The registry result should therefore be interpreted within its reported model and population, rather than treated as a complete account of all safety evidence.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of Baricitinib in Participants With Rheumatoid Arthritis — NCT04086745.
• Registry record last updated 2026-07-10; status last verified 2026-06.
• Results first posted 2026-07-10.
• 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

This completed, randomized, controlled, unmasked Phase 4 study concerns participants with rheumatoid arthritis and compares baricitinib with TNF inhibitors. The primary result is a pooled JAJA and JAJD Cox regression comparison of time to first VTE: hazard ratio 1.606, with a 95% confidence interval from 0.969 to 2.660. The primary p-value is not reported in the registry record.

The estimate is above 1, but its confidence interval includes 1 and indicates uncertainty. The registry labels the analysis non-inferiority but does not report the margin or formal criterion. The pooled analysis group sizes differ from the enrollment listed for the individual registry entry. Secondary outcomes and adverse-event summaries add context, but missing analysis details and the absence of a reported multiplicity strategy limit further inference.