This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Comparative Analysis of Outcomes Among Patients Initiating Xeljanz in Combination With Oral MTX Who Withdraw MTX Versus Continue MTX (NCT03975790), a registered phase not stated study. The primary outcome is Number of Participants as Per Type of Insurance Plan.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT03975790 |
| Phase | Phase not stated |
| Condition | Rheumatoid Arthritis |
| Allocation | Not Stated |
| Design model | Not Stated |
| Masking | Not Stated |
| Enrollment | 479 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Pfizer |
| Results posted | Yes |
2. Clinical Question
Population
Patients with rheumatoid arthritis who newly initiated tofacitinib in combination with oral methotrexate, identified through a United States healthcare claims database. The registry reports enrollment of 479. The primary measure is described as covering all participants included in the study; the number contributing to each analysis is not reported in the registry record.
Intervention
There is no assigned intervention in this retrospective cohort. The study examines treatment patterns after patients initiated tofacitinib with oral methotrexate, including the group described as discontinuing methotrexate.
Comparator
The registry describes groups who persist with methotrexate, discontinue it, or experience an interruption. These are treatment-pattern groups rather than randomized treatment arms. The criteria used to define each group are not reported in the registry record.
Primary question
The primary measure is the number of participants by type of insurance plan during the pre-index period. It distinguishes private insurance from employer-provided insurance. It does not directly compare clinical outcomes between the methotrexate treatment-pattern groups.
The study’s broader stated objective is to evaluate patient characteristics, treatment patterns, medication effectiveness, healthcare costs, and utilization among patients newly initiating tofacitinib with oral methotrexate. The registry also lists mean total all-cause monthly healthcare cost and mean total rheumatoid arthritis-related monthly healthcare cost as secondary outcomes. The primary endpoint, however, is a descriptive insurance count. It can characterize the insurance types recorded for the study population, but by itself it cannot answer whether continuing, stopping, or interrupting methotrexate is associated with different clinical outcomes. No insurance-category counts or comparative estimates are reported in the registry record.
3. Trial Design
This record describes a completed, retrospective cohort study using the United States healthcare claims database named in the official title. The brief summary says the study was intended to evaluate patient characteristics, treatment patterns, medication effectiveness, healthcare cost, and utilization among patients with rheumatoid arthritis newly initiating tofacitinib in combination with oral methotrexate. The registry reports enrollment of 479.
The treatment-pattern descriptions distinguish patients who persist with methotrexate, discontinue it, or experience an interruption while taking tofacitinib with methotrexate. These groups are observational categories, not randomized assignments. The registry states that the comparison is not randomized and that allocation is not stated. Consequently, the groups should not be treated as if they were created through random allocation or as necessarily comparable at baseline. The criteria and timing used to classify patients into these categories are not reported in the registry record.
The reported start date is 2018-05-29, and the primary completion date is 2018-10-02. The study status is completed. The primary outcome concerns the number of participants by insurance-plan type during a pre-index period described as 12 months before the index date. The index date is defined as the date of the first tofacitinib claim made to a participant’s insurance provider during an identification period of 3 years.
The registry lists 70 outcome measures and 669 analyses. These counts indicate the breadth of registered outcome and analysis entries, but they do not establish that every analysis used the same population, methods, or data completeness. Specific analytic models, adjustment variables, and definitions for the secondary outcomes are not reported in the registry record.
4. Participant Flow and Accounting
The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.
Before assignment: The study was a retrospective population-based register study. Detailed data on each participant was retrieved from the Truven Health MarketScan Research Database and assessed in this study.
Overall Study
| Stage | Methotrexate (MTX) + Tofacitinib: MTX Persistent | Methotrexate (MTX) + Tofacitinib: MTX Discontinued | Methotrexate (MTX) + Tofacitinib: MTX Interrupted |
|---|---|---|---|
| STARTED | 337 | 94 | 48 |
| COMPLETED | 337 | 94 | 48 |
| NOT COMPLETED | 0 | 0 | 0 |
Analysis population for the primary outcome, as defined in the registry: Analysis was performed on all participants included in the study.
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: Analysis was performed on all participants included in the study.
| Characteristic | Unit | Methotrexate (MTX) + Tofacitinib: MTX Persistent | Methotrexate (MTX) + Tofacitinib: MTX Discontinued | Methotrexate (MTX) + Tofacitinib: MTX Interrupted | Total |
|---|---|---|---|---|---|
| Participants analysed | 337 | 94 | 48 | 479 | |
| Age, Continuous | years | 56.10 (SD 10.37) | 55.02 (SD 9.81) | 58.08 (SD 9.43) | 56.09 (SD 10.18) |
| Sex: Female, Male: Female | Participants | 269 | 82 | 35 | 386 |
| Sex: Female, Male: Male | Participants | 68 | 12 | 13 | 93 |
| Race and Ethnicity Not Collected | Participants | — | — | — | 0 |
6. Primary Endpoint
The registry lists the primary outcome as Number of Participants as Per Type of Insurance Plan, measured over During pre-index period (12 months duration before the index date [Index date: date of first claim for tofacitinib made by participants to their insurance provider during identification period of 3 years]).
| Element | Registry entry |
|---|---|
| Result type | Count Of Participants |
| Measure of spread | not applicable |
| Unit | Participants |
| Endpoint classification | Binary Endpoint |
The primary endpoint counts participants according to insurance-plan type during the pre-index period. The registry describes two types: private insurance purchased by participants and commercially available, and employer-provided insurance posted through a participant’s employer. The time window is described as 12 months before the index date, with the index date defined by the first tofacitinib claim during the identification period of 3 years.
This is a descriptive endpoint. Counts by insurance type can show how the study population was distributed across the two recorded plan categories. Such information can help characterize the population represented in a claims-based study, because insurance coverage determines which healthcare claims may be available in the database. It is not, on its own, a measure of rheumatoid arthritis activity, treatment response, medication effectiveness, cost, or utilization. Nor does it compare outcomes among the methotrexate persistence, discontinuation, and interruption groups.
The registry identifies the endpoint population as all participants included in the study. The actual counts for either insurance category are not reported in the registry record. The record also does not report how participants with missing or changing insurance information were handled. Accordingly, the endpoint’s definition and intended time window can be described, but its observed distribution and completeness cannot be assessed from the registry record.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The central interpretive point is that the registered primary outcome is an insurance-plan count, not a treatment-effect estimate. It classifies participants by private or employer-provided insurance during the defined pre-index period. The registry describes enrollment of 479, but the category counts and any uncertainty estimates for this outcome are not reported in the registry record. Without those results, the insurance distribution cannot be described or compared across treatment-pattern groups.
The study’s stated aim is broader than its primary endpoint. Its brief summary mentions treatment patterns, medication effectiveness, healthcare costs, and utilization. The registry lists mean total all-cause monthly healthcare cost and mean total rheumatoid arthritis-related monthly healthcare cost as secondary outcomes. Their observed values, comparisons, and analytic details are not reported in the registry record. The presence of registered outcomes does not itself establish what the data showed for them.
The study is retrospective and nonrandomized. Patients were grouped by observed methotrexate use patterns: persistence, discontinuation, or interruption. In general, observational comparisons may reflect differences between groups that existed before the treatment pattern occurred, as well as differences associated with treatment. Randomization can help balance measured and unmeasured characteristics on average; this record does not describe randomized allocation. The registry does not report the adjustment strategy, covariates, or methods used to address confounding, so the degree to which group differences were controlled cannot be evaluated.
The registry’s adverse-event entries list zero participants at risk and zero affected for serious events, other events, and deaths in each of the three groups. These entries should be read as registry entries, not as evidence that no events occurred in the underlying population. The record does not explain whether the zeros reflect collected safety data, unavailable reporting, or another convention. No serious-event terms are listed.
Overall, the registry record establishes the study’s observational setting, stated aims, treatment-pattern categories, and primary endpoint definition. It does not report the primary category counts, secondary-outcome results, or enough analytic detail to judge comparative effectiveness. No conclusion about the direction or magnitude of treatment differences can be drawn from the endpoint definition or from the presence of registered analyses alone.
9. Statistical Methods Explained
Statistical interpretation begins with the question each outcome can answer. Here, the primary outcome is a count of participants by insurance-plan type. A count describes how many participants fall into a stated category; it is not automatically a rate, risk, or treatment effect. The registry’s endpoint definition distinguishes private insurance from employer-provided insurance and places the measurement in the pre-index period. The category-specific results are not reported in the registry record.
What does a count measure?
In general, a count is the number of observations assigned to a category under a defined rule. To interpret it, readers need to know the population included, the category definitions, and the time window. The registry specifies the intended population as all participants included in the study and defines the pre-index period relative to the first tofacitinib claim. It does not report the observed counts, any missing-data handling, or whether insurance status could change during the window. Therefore, the endpoint’s design is clear in broad terms, but its observed distribution and data completeness cannot be evaluated.
Counts differ from proportions. A proportion relates a category count to a denominator, and a percentage expresses that proportion on a percentage scale. No category-specific counts or denominators are reported for this endpoint, so no proportions or percentages can be stated. No new values should be inferred from the enrollment figure, because the registry does not establish that the enrollment count is the denominator for every analysis.
What does a retrospective cohort comparison imply?
In general, a retrospective cohort study classifies people using information already recorded and compares outcomes across groups defined by exposures or treatment patterns. This design can describe associations in the recorded population. It does not create randomized groups. The registry’s categories concern methotrexate persistence, discontinuation, and interruption among patients initiating tofacitinib with methotrexate. The rules used to define these categories are not reported in the registry record.
When groups arise through observed treatment decisions, their outcomes may also reflect differences in health status, treatment history, access to care, or other characteristics. This is confounding: a difference attributed to one exposure may partly reflect other factors related to both the exposure and outcome. Statistical adjustment can address measured factors when appropriate information and methods are available, but it cannot guarantee control of unmeasured factors. The registry record does not report adjustment variables or the methods used to account for confounding. The listed groups therefore should not be interpreted as randomized treatment arms or as a like-for-like control comparison.
How should uncertainty be considered?
General statistical principle: an estimate is more informative when accompanied by a measure of uncertainty, such as a confidence interval. A confidence interval expresses the range of values reasonably compatible with the data and the specified method. It is not a guarantee that the true value lies within the interval. Neither an interval nor a primary-outcome estimate is reported in the registry record, so precision cannot be assessed for the insurance counts. The record also does not report model specifications or standard errors for the secondary outcomes.
For cost outcomes, general statistical considerations include skewness and the possibility that a small number of high-cost observations may influence a mean. Claims-based cost analyses may also depend on the period of observation, cost definitions, and handling of incomplete follow-up. The registry names mean monthly cost outcomes but does not report their definitions beyond their titles, observed values, distributions, or analytic procedures. These general considerations cannot establish how the study’s cost data behaved.
What do the planning illustrations represent?
General power calculations estimate how much information a proposed study would need to have a specified chance of detecting an effect of a chosen size under stated assumptions. The planning illustrations associated with the analyses use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the stated effect is a fraction of the observed effect. For time-to-event outcomes, the illustration concerns total events using the Schoenfeld method, while other illustrations refer to participants per arm. Such calculations are planning illustrations, not observed results or observed power. They do not show that the study had a particular probability of detecting an effect after the data were collected. The registry does not report a sample-size justification as part of its design, and the calculated planning values are not reported in the registry record.
The registry lists 70 outcome measures and 669 analyses. Multiple analyses may raise questions about which results were prespecified and how multiplicity was handled. The number of entries alone does not reveal the statistical plan, whether tests were adjusted, or how results were selected for interpretation. Those details are not reported in the registry record. The record also does not state whether interim analyses were planned or conducted.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists 70 outcome measures and 669 analyses. With many outcomes or comparisons, the chance of finding at least one apparently notable result can increase if each is tested separately. General approaches to multiplicity include prespecifying a primary comparison, adjusting significance thresholds, or interpreting secondary analyses as exploratory. The registry record does not report a multiplicity strategy, identify a hierarchy among the analyses beyond naming a primary outcome, or describe how results were selected for emphasis. The count of registered analyses alone does not establish whether or how multiplicity was addressed.
Missing-data handling is not reported in the registry record. For the primary insurance endpoint, the record identifies the period and the intended population but does not explain how missing coverage information, changes in coverage, or incomplete observation were treated. For monthly cost outcomes, the record does not state how incomplete follow-up or absent cost records were handled. These omissions matter because the available claims may not represent every participant equally over every relevant period.
The primary endpoint population is described as all participants included in the study, but analysis-specific population counts are not reported. The registry does not describe exclusions, data availability requirements, or whether the analysis populations differed across the 70 outcomes and 669 analyses. Nor does it state whether sensitivity analyses were used to assess alternative assumptions about missing data or population definitions.
Because the study is observational, details about adjustment and the analysis population are also relevant to interpretation. The registry does not report the covariates or analytic methods used for comparative outcomes. Thus, the record does not allow assessment of how missingness, multiplicity, or differences between the treatment-pattern groups were handled.
11. What the Result Does, and Does Not, Mean
The registry record supports a limited set of conclusions about study scope. It describes a completed retrospective cohort study of patients with rheumatoid arthritis initiating tofacitinib with oral methotrexate, with treatment-pattern categories based on methotrexate persistence, discontinuation, or interruption. It identifies insurance type as the primary outcome and names two monthly healthcare-cost outcomes as secondary measures. Enrollment is reported as 479.
The primary outcome can describe insurance-plan categories only if the corresponding counts are available. Those category counts are not reported in the registry record. The insurance measure is not a clinical endpoint and cannot establish whether one methotrexate pattern was associated with better or worse rheumatoid arthritis outcomes. Although medication effectiveness is part of the stated study objective, the results and definitions needed to interpret effectiveness are not reported in the registry record.
The secondary cost titles identify areas of interest, not results. Without reported values, comparisons, uncertainty estimates, or analytic details, the record does not show whether monthly costs differed among groups or how precisely any difference was estimated. The registry’s adverse-event entries list zero counts in all three groups, but the record does not explain the meaning of those entries. They should not be used alone to infer that no events occurred.
More generally, observational group comparisons can describe associations but may be affected by differences between people who follow different treatment patterns. The record says allocation was not randomized and does not report methods for controlling confounding. It therefore does not support treating the groups as equivalent at baseline or attributing any unreported outcome difference to methotrexate use alone.
In short, the record describes the study question and data structure more clearly than it reports the findings. It does not establish a treatment benefit, lack of benefit, or comparative safety result.
12. Why This Trial Matters Statistically
This study is statistically useful to examine because it combines an observational treatment-pattern question with a primary endpoint that is descriptive rather than clinical. That distinction prevents a common interpretive error: treating every registered primary outcome as a direct measure of treatment effectiveness. Here, the primary endpoint counts insurance-plan types during a defined pre-index period. It characterizes the study population rather than measuring response to rheumatoid arthritis treatment.
The registry also illustrates why the design matters. Patients were grouped according to observed methotrexate patterns rather than randomized allocation. In general, treatment decisions and treatment persistence can be related to patient characteristics and healthcare circumstances. A comparison across such groups needs clear definitions and an appropriate strategy for confounding if it is to support causal interpretation. The methods needed to assess those issues are not reported in the registry record.
The record’s breadth is notable: it lists 70 outcome measures and 669 analyses, alongside secondary outcomes concerning monthly healthcare costs. This creates an important statistical question about how outcomes and analyses were prioritized and interpreted. The registry record does not report a multiplicity strategy or detailed analysis plan, so the number of entries cannot be translated into a judgment about statistical rigor or the reliability of any particular result.
Finally, the case highlights the difference between a registered question and a fully interpretable result. Endpoint definitions, population descriptions, estimates, uncertainty, missing-data handling, and analytic methods each contribute to interpretation. Some of those elements are present, while key outcome values and methods are not reported in the registry record. Readers can identify what the study was designed to examine, but should not infer clinical findings from the design description alone.
13. 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 endpoint is a count of participants by insurance type, not a direct clinical outcome. The registry does not report the counts for private or employer-provided insurance, so the endpoint’s observed distribution cannot be interpreted. Although the endpoint is assigned to all participants included in the study, analysis-specific population sizes are not reported.
The study is retrospective and nonrandomized. The registry identifies methotrexate persistence, discontinuation, and interruption as treatment-pattern categories, but does not report their operational definitions or the timing rules used to assign participants. Allocation is not stated, and the registry indicates that the comparison is not randomized. The record does not report adjustment variables or methods for addressing confounding, limiting assessment of whether group differences could be explained by characteristics other than treatment pattern.
Important outcome information is unavailable in the registry record. The cost outcomes are named, but their results and detailed definitions are not reported. Medication-effectiveness results are also not reported. The adverse-event entries list zero counts in each group, but the record does not clarify how to interpret those entries or describe safety-data collection. These gaps prevent a substantive assessment of effectiveness, costs, or safety from the registry information alone.
The record lists many outcomes and analyses but does not report a multiplicity plan, missing-data procedures, interim-analysis details, or a full statistical analysis plan. It also does not report the sample-size justification. These omissions make it difficult to evaluate prespecification, analysis-population consistency, and the precision of comparative results. The registry reports enrollment of 479, but that figure should not be assumed to be the denominator for every analysis.
14. Related Tutorials
15. Related Statistical Calculators
16. Sources
- ClinicalTrials.gov: Comparative Analysis of Outcomes Among Patients Initiating Xeljanz in Combination With Oral MTX Who Withdraw MTX Versus Continue MTX — NCT03975790.
- Registry record last updated 2023-06-09; status last verified 2023-06.
- Results first posted 2019-10-22.
- Data retrieved from ClinicalTrials.gov on 2026-11-09.
- Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
17. Record Summary
The registry describes a completed retrospective cohort study of patients with rheumatoid arthritis who initiated tofacitinib with oral methotrexate, using a United States healthcare claims database. Reported enrollment is 479. Treatment-pattern groups are described as methotrexate persistent, discontinued, or interrupted. Allocation is not randomized.
The primary outcome counts participants by insurance-plan type during the 12-month pre-index period, distinguishing private from employer-provided insurance. The registry also lists mean total all-cause monthly healthcare cost and mean total rheumatoid arthritis-related monthly healthcare cost as secondary outcomes. It reports 70 outcome measures and 669 analyses.
Observed insurance counts, cost results, comparative effectiveness results, analysis methods, missing-data procedures, and multiplicity handling are not reported in the registry record. The adverse-event entries list zero counts in each treatment-pattern group, without an explanation of those entries.