Aflibercept for Retinopathy of Prematurity - Intravitreal Injection Versus Laser Therapy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Aflibercept for Retinopathy of Prematurity - Intravitreal Injection Versus Laser Therapy (NCT04004208), a randomized phase 3 study. The primary outcome is Proportion of Participants With Absence of Active ROP and Unfavorable Structural Outcomes.

2. Clinical Question

The primary outcome combines two kinds of findings. Active ROP was defined as ROP requiring treatment. Unfavorable structural outcomes included retinal detachment, macular dragging, macular fold, or retrolental opacity. A participant had the primary outcome if the required absence of both active disease and those structural outcomes was recorded.

The registry describes an efficacy, safety, and tolerability comparison, with additional interest in how aflibercept moves into, through, and out of the body. The primary comparison is therefore narrower than the study's full set of objectives: it concerns a specified ROP outcome at a specified follow-up time, rather than every safety, recurrence, or pharmacokinetic question. Results for the primary measure do not by themselves answer the other objectives.

The registry record does not report further detail about the participants' clinical characteristics or eligibility beyond the condition and study population description. The treatment descriptions allow for additional treatment in either group if needed, so the comparison is between the assigned treatment strategies as described, not necessarily between one injection and one laser procedure in every participant.

3. Trial Design

This was a phase 3, randomized, open-label, parallel-group, two-arm controlled study. The registry record lists enrollment of 113 participants and identifies Bayer as the sponsor. The study status is completed. The recorded start date is 2019-09-25, and the primary completion date is 2021-02-12.

Participants were assigned to aflibercept or laser photocoagulation. The registry reports 75.0 participants in the aflibercept arm and 38.0 in the laser arm for the primary outcome. Masking was listed as none, so the study was open-label rather than masked. The registry record does not report further details about the randomization procedure, such as how the allocation sequence was generated or concealed.

The aflibercept strategy was an intravitreal injection of Dose A at baseline in each eligible eye, with additional injections possible if needed. The laser strategy was treatment in each eligible eye at baseline, with retreatments possible if needed. The primary outcome was assessed at 24 weeks after starting study treatment.

The study also listed outcomes concerning second treatment modalities, recurrence, an ROP activity scale, ocular and systemic treatment-emergent adverse events, serious adverse events, and free aflibercept concentrations in plasma. Its stated purpose included assessing efficacy, safety, tolerability, and the movement of the drug through the body. The primary endpoint is only one element of that broader set of questions.

The registry record does not report the planned sample-size justification, details of any interim analyses, or a full statistical analysis plan. It also does not provide enough information to establish how all participants were handled in every outcome analysis. Those omissions limit how much can be inferred about the design choices and analysis population beyond the details explicitly recorded.

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: 121 participants were screened. 1 participant was a screen fail and 2 participants were withdrawn by parent/guardian. 118 participants were randomized, 75 participants were randomized to the aflibercept arm and 43 to the laser arm. 113 participants were treated, 5 participants randomized to the laser photocoagulation arm were withdrawn before receiving any study intervention.

Overall Study

Reasons for not completing this period:

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: SAF

6. Primary Endpoint

The registry lists the primary outcome as Proportion of Participants With Absence of Active ROP and Unfavorable Structural Outcomes, measured over At 24 weeks after starting study treatment.

The primary endpoint is the proportion of participants with absence of active ROP and unfavorable structural outcomes at 24 weeks after starting study treatment. Active ROP meant ROP requiring treatment. The structural outcomes were retinal detachment, macular dragging, macular fold, or retrolental opacity.

This is a composite outcome: a participant's status depends on both disease activity and structural findings. In general, composite endpoints can capture more than one clinically relevant aspect of a condition in a single measure. They can also make interpretation less straightforward, because the overall result does not reveal how much each component contributed. The registry record does not report a separate primary result for each component, so the primary comparison cannot establish which component, if any, chiefly accounts for the observed proportion.

The endpoint is a proportion, not a time-to-event measure. It describes the share of participants meeting the defined outcome at the stated follow-up rather than the timing of an event during follow-up. The registry record gives arm proportions of 0.855 for aflibercept and 0.821 for laser photocoagulation.

The endpoint's precise role in the protocol, including why this composite and time point were selected over alternatives, is not reported in the registry record. In general, a predefined endpoint provides a common target for comparing assigned groups, but interpretation also depends on how the outcome was measured, how missing assessments were handled, and which participants were included. Those analysis details are not reported in the registry record.

7. Results Posted in the Registry

• Laser Photocoagulation: 0.0% (0/38) versus 1.3% (1/75) on control; risk difference -1.3 percentage points (95% CI -3.9 to 1.3); unadjusted two-proportion z-test p = 0.475.

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.

Percentage of Participants With Systemic TEAEs

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Laser Photocoagulation: 63.2% (24/38) versus 52.0% (39/75) on control; risk difference 11.2 percentage points (95% CI -7.9 to 30.2), risk ratio 1.21 (0.88–1.68); unadjusted two-proportion z-test p = 0.259.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Laser Photocoagulation and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants With Systemic SAEs

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Laser Photocoagulation: 36.8% (14/38) versus 24.0% (18/75) on control; risk difference 12.8 percentage points (95% CI -5.3 to 31.0), risk ratio 1.54 (0.86–2.74); unadjusted two-proportion z-test p = 0.152.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Laser Photocoagulation and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists a broad set of secondary questions spanning additional treatment, recurrence, ROP activity, ocular and systemic adverse events, serious adverse events, and plasma concentrations of free aflibercept. Results posted for only part of that set are available in the registry record. Findings for outcomes without a posted result are not reported in the registry record.

Across the reported secondary binary outcomes, the unadjusted contrasts are accompanied by uncertainty, and the listed intervals span both directions of difference. Taken together, they do not supply a single, consistent basis for a firm ranking of the treatment strategies. Safety outcomes and efficacy-related outcomes also address different questions; they should not be collapsed into one overall judgment.

These secondary findings complement rather than replace the primary result. The primary endpoint concerns the specified composite at 24 weeks, whereas secondary measures address other events and domains. A secondary outcome may help describe the broader experience recorded during the study, but it does not resolve the uncertainty around the primary comparison or answer questions beyond its own definition and timeframe.

Unadjusted secondary results carry less weight when many outcomes are examined because repeated testing can increase the chance of apparently notable findings arising by chance. The registry record does not report a multiplicity plan that would clarify how these results were prioritized or adjusted. Secondary estimates are therefore best treated as descriptive and exploratory unless a prespecified analysis plan establishes otherwise.

The registry lists 13 outcome measures and reports 12 posted secondary outcomes. A list of outcome names does not establish that complete results are available for every measure. The record also does not report whether outcome definitions or analysis populations differed across the secondary analyses. Interpretation should remain tied to the specific posted estimate, its interval, and its stated timeframe, without treating the collection of secondary findings as independent confirmation of the primary result.

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

• Proportion of Participants Requiring Intervention With a Second Treatment Modality
• Proportion of Participants With Recurrence of ROP
• Exploration of ROP Activity Scale Proposed by the International Neonatal Consortium
• Percentage of Participants With Ocular Treatment-emergent Adverse Events (TEAEs)
• Percentage of Participants With Ocular Serious Adverse Events (SAEs)
• Percentage of Participants With Systemic TEAEs
• Percentage of Participants With Systemic SAEs
• Concentrations of Free Aflibercept in Plasma

9. Statistical Interpretation

10. Statistical Methods Explained

What does a proportion measure?

A proportion is the number of participants meeting an outcome divided by the number assessed or included in the analysis. For the primary endpoint, the posted arm results are 0.855 for aflibercept and 0.821 for laser photocoagulation. These proportions summarize the observed primary outcome in each group. They do not by themselves account for differences in participant characteristics or explain why an outcome occurred.

For a binary outcome, each participant is classified according to whether the specified event or condition was recorded. Here, the endpoint is framed positively as absence of active ROP and unfavorable structural outcomes. The secondary adverse-event outcomes instead describe the percentage of participants with an event. Interpretation depends on the exact definition and timeframe for each measure; the registry lists different timeframes for different outcomes.

What is a risk difference?

A risk difference contrasts two group proportions by subtraction, with the sign depending on which group is subtracted from which. It is expressed on the proportion scale. A value of zero corresponds to equal proportions under the chosen orientation. The unadjusted primary recalculation reports -0.013 and a confidence interval from -0.039 to 0.013. Because the registry's sponsor-reported analysis gives a different summary, the two figures should not be assumed to use the same definition or analysis procedure. Details needed to align their estimands and coding are not reported in the registry record.

For secondary binary outcomes, the recalculations also include risk differences and, for some outcomes, relative measures such as risk ratios and odds ratios. A risk ratio compares proportions as a ratio, while an odds ratio compares odds. These scales answer different questions and can have different magnitudes. They should not be confused with a risk difference, and none of them alone explains the clinical importance of an outcome.

How should a p-value be read?

A p-value measures how unusual data at least as incompatible with a specified null hypothesis would be under the statistical model, if that model and its assumptions held. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of an effect. The recalculated primary p-value is 0.475. It should be interpreted alongside the effect estimate and interval, not as a stand-alone verdict.

The secondary recalculation p-values are unadjusted. When many outcomes or comparisons are examined, the chance of finding apparently unusual results by chance can increase. Without a reported multiplicity strategy, an isolated unadjusted p-value is generally less persuasive than a result from a prespecified primary analysis. The registry record does not report a multiplicity adjustment plan.

What does a confidence interval contribute?

A confidence interval gives a range of estimates compatible with the observed data under a specified procedure and its assumptions. It communicates precision as well as direction and scale. A narrower interval indicates greater statistical precision for that estimate; a wider interval leaves more uncertainty about the size and direction of the contrast. A confidence interval is not a guarantee that the true value lies within its endpoints, and its interpretation depends on the method used.

The recalculated primary interval spans negative and positive values. It therefore includes a zero risk difference. This does not prove equivalence. Establishing equivalence or non-inferiority requires a prespecified margin and a suitable analysis. The sponsor-reported interval is at 90%, whereas the recalculated interval is at 95%; intervals at different confidence levels are not directly interchangeable.

What is non-inferiority analysis?

In general, a non-inferiority analysis asks whether an intervention is not worse than a comparator by more than a prespecified acceptable margin. The margin must be defined in advance and applied with an appropriate analysis. The registry reports a difference in proportions of 0.034 and a 90% confidence interval from -0.08 to 0.162, and labels the analysis as non-inferiority. The margin, hypothesis direction, analysis population, and detailed method are not reported in the registry record. Without those details, the numerical interval cannot independently establish whether the non-inferiority criterion was met.

How should the planning illustrations be understood?

The planning illustrations are based on the observed effect, not on the trial's own design justification. The stated assumptions are a two-sided significance level of 0.05, equal allocation between arms, a normal approximation, and an effect set at a stated fraction of the observed effect. Such calculations illustrate the number of participants per arm that would be associated with specified power under those assumptions. They are not observed power, do not show what power this completed study had, and do not establish what sample size was originally planned. The trial's sample-size justification is not reported in the registry record.

In general, power calculations depend on the assumed effect, outcome frequency, analysis method, significance level, and allocation. Planning from an observed effect can be especially sensitive to the uncertainty of that effect. The illustrations are therefore hypothetical planning aids, not evidence that the study was adequately or inadequately designed.

What analysis assumptions remain unclear?

For a binary endpoint, common approaches model the probability of an event or compare observed proportions. The registry does not report the detailed method for the sponsor's primary analysis, the handling of missing outcomes, the analysis population, or whether covariates were used. The recalculated results are explicitly unadjusted summaries from posted data. They do not substitute for a full analysis plan or recover information that the registry record does not report.

11. Confidence Intervals and Precision

The primary arm proportions are 0.855 for aflibercept and 0.821 for laser photocoagulation. The unadjusted risk-difference estimate is -0.013, with a 95% confidence interval from -0.039 to 0.013. The interval crosses zero and covers values on both sides of it. This means the estimate is not highly precise about the direction of the contrast under this calculation: values compatible with the data include contrasts with either sign.

A confidence interval is a range generated by a statistical procedure that, under its assumptions, reflects sampling uncertainty. It is not a range of individual participant outcomes, nor does it incorporate every possible source of uncertainty. The interval's width is affected by the number of observations, the variability of the outcome, and the analysis method. It does not account automatically for bias from outcome measurement, missing data, or limitations in the study design.

The registry's non-inferiority analysis reports a difference in proportions of 0.034 and a 90% confidence interval from -0.08 to 0.162. This is a separate reported analysis and confidence level from the unadjusted recalculation. The non-inferiority margin and analysis details are not reported in the registry record, so the interval cannot be interpreted as a stand-alone non-inferiority decision.

In general, a confidence interval that includes zero does not establish that two groups are equivalent. Equivalence and non-inferiority require prespecified margins and appropriate procedures. Conversely, a confidence interval that excludes zero would not by itself show that an effect is clinically important. Precision, direction, magnitude, design, and outcome meaning all matter.

The secondary outcomes also have uncertainty, and some have broad intervals. Their estimates should not be treated as exact rankings of the treatment strategies. The small group sizes recorded for the primary comparison, particularly the 38.0 participants in the laser arm, limit how much precision can be expected for comparisons of uncommon outcomes. The registry record does not report whether adjustment or other methods were used in sponsor analyses to address this uncertainty.

12. Analyses Reported by the Sponsor

The registry reports a sponsor analysis labeled non-inferiority, with a difference in proportions of 0.034 and a 90% confidence interval from -0.08 to 0.162. It does not report a p-value for that analysis. The registry also contains an unadjusted recalculation of the primary contrast: risk difference -0.013, 95% confidence interval -0.039 to 0.013, and p = 0.475.

These estimates are not identical summaries. A difference in proportions and a risk difference may sound alike, but their interpretation depends on the groups' coding, contrast direction, outcome definition, analysis population, and statistical method. The registry record does not report enough detail to establish whether the sponsor analysis and recalculation use precisely the same estimand or coding. They should therefore be read as distinct results rather than combined or treated as interchangeable.

The sponsor's analysis is the registry-reported analysis for the non-inferiority question. Its interpretation depends on a prespecified margin, the direction of the contrast, and the analysis method; the margin and these details are not reported in the registry record. The reported interval alone cannot show whether the margin criterion was met.

The recalculation is a simple unadjusted comparison derived from the posted summary statistics. It provides a transparent descriptive contrast but does not reproduce an analysis whose full method is unavailable. Neither summary should be used to infer the study's original sample-size rationale or the handling of missing outcomes, which are not reported in the registry record.

13. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study evaluates multiple outcomes, comparisons, or analyses. Even when each test is conducted using a conventional threshold, the chance of at least one apparently unusual result can rise as more tests are considered. This is a general statistical concern, not evidence that any particular secondary result in this study is false.

The registry lists 13 outcome measures and reports 12 posted secondary outcomes. It does not report a multiplicity adjustment strategy, a hierarchy for testing outcomes, or which secondary analyses were confirmatory rather than exploratory. The primary outcome is identified, but the broader testing plan is not reported in the registry record. Unadjusted secondary p-values should consequently be interpreted cautiously and should not be treated as definitive standalone evidence.

Important details about missing data are also not reported in the registry record. The record does not explain whether all randomized participants had primary outcome assessments, how missing assessments were handled, or whether any sensitivity analyses were used. For a proportion, excluding participants without an outcome can affect the estimate if missingness is related to outcome status or treatment. This is a general possibility; the registry record does not show whether it occurred.

The analysis population is not reported in enough detail to determine whether the primary result includes all randomized participants or a subset defined by treatment received, follow-up, or other criteria. This matters particularly for a non-inferiority question, where different analysis populations can have different implications. The registry also does not report whether any protocol deviations affected inclusion or how those deviations were handled.

The protocol's sample-size justification and any interim analyses are not reported in the registry record. The separate planning illustrations use the observed effect and stated assumptions; they do not establish the trial's original design rationale. Overall, multiplicity, missing-data handling, analysis populations, and interim monitoring remain limitations on a complete appraisal of the statistical plan.

14. 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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: retinal detachment (5.3%), infantile apnoea (5.3%), retinal haemorrhage (2.7%), bronchiolitis (2.7%), pulmonary valve stenosis (2.6%).

The registry reports serious adverse events affecting 9 participants in the aflibercept group, with 75 at risk, and 10 participants in the laser group, with 38 at risk. It also records 3 deaths in the aflibercept group and 0 in the laser group. These are counts in the registry record; they do not establish that treatment caused an event or death.

The listed serious adverse-event terms include retinal detachment and infantile apnoea, each recorded as 5.3%, and retinal haemorrhage and bronchiolitis, each recorded as 2.7%. Pulmonary valve stenosis and necrotising colitis are each recorded as 2.6%. These term-level percentages describe the posted record and should not be added together or interpreted as mutually exclusive groups without further information.

Serious adverse events are distinct from all adverse events, and event counts are not necessarily the same as participant counts when a participant can experience more than one event. The registry record does not report the causal attribution of these events, their timing, or enough detail to determine how overlapping events were counted. The group sizes differ, and crude counts alone do not provide a like-for-like comparison of risk.

Safety findings should be read alongside the study's timeframe and the limitations of the posted data. The registry lists systemic and ocular adverse-event outcomes, but not every safety detail is reported in the registry record. These results describe what was recorded; they do not establish a complete safety profile or support medical advice.

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

The primary results describe the proportion of participants recorded as having absence of active ROP and unfavorable structural outcomes at 24 weeks after starting treatment. The posted proportions are 0.855 for aflibercept and 0.821 for laser photocoagulation. The unadjusted recalculation estimates a risk difference of -0.013, with a 95% confidence interval from -0.039 to 0.013 and p = 0.475.

Those statistics do not establish that the treatments are equivalent. A result that is compatible with no difference is not the same as proof that differences are absent or unimportant. Nor do the arm proportions and recalculated contrast establish a non-inferiority conclusion. That conclusion requires a prespecified margin and an analysis that applies it; the margin and detailed method are not reported in the registry record.

The endpoint is a composite, so its overall result does not say which component drove the outcome. It also concerns a defined assessment time and does not answer every question about later outcomes, recurrence, additional treatment, adverse events, or drug concentrations. Those questions require their own results and interpretation.

The study's open-label design, the difference in the recorded group sizes, and missing detail about outcome handling and analysis populations limit how fully the results can be assessed. These points do not invalidate the posted estimates, but they qualify the conclusions that can be drawn from them. Confidence intervals express sampling uncertainty under a statistical procedure; they do not cover every potential source of bias or uncertainty.

Accordingly, the registry results are best understood as group-level estimates from a completed randomized comparison, not as a guarantee about outcomes for an individual baby. They do not constitute medical advice, and the statistics alone do not determine a treatment choice.

16. Why This Trial Matters Statistically

This trial illustrates why reading a clinical comparison requires more than looking at the two arm proportions. The primary outcome is a composite, the design is open-label, and the registry reports both a sponsor non-inferiority analysis and a simple unadjusted recalculation. Understanding what each estimate represents is essential before comparing their numerical values.

It also shows the distinction between a difference that is not clearly separated from zero and evidence of equivalence or non-inferiority. The recalculated confidence interval spans zero, while the sponsor analysis is labeled non-inferiority. A non-inferiority conclusion depends on a margin and analysis plan, details not reported in the registry record. The label alone does not supply those missing elements.

The study further illustrates the importance of uncertainty for binary outcomes with relatively limited group sizes. Point estimates provide a compact description, but intervals show that a range of contrasts is compatible with the data. This is especially important when readers are tempted to treat small numerical differences as precise rankings.

Finally, the record combines a primary efficacy-related endpoint with secondary outcomes covering recurrence, further treatment, safety, and pharmacokinetics. Those domains are related to the study's broader objectives but are not interchangeable. Multiple outcomes create interpretive challenges, particularly when multiplicity handling and analysis populations are not reported. The case is therefore useful for distinguishing what the registry quantifies from what its posted statistical details leave unresolved.

17. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record does not report the primary analysis method in enough detail to reproduce or fully interpret the sponsor's analysis. It provides a sponsor-reported difference in proportions and a separate unadjusted recalculation, but does not supply the definitions and coding needed to reconcile the estimates. The non-inferiority margin and decision rule are also not reported.

The study was open-label, with masking listed as none. The record does not say whether outcome assessors were masked, whether subjective assessments could have been influenced by knowledge of assignment, or how treatment decisions and retreatments were standardized. These are possible design considerations, not claims that bias occurred.

The registry does not report a detailed participant flow, the analysis population for each outcome, missing-data handling, protocol deviations, or sensitivity analyses. It also does not report the original sample-size justification or interim analyses. Those omissions prevent a full appraisal of whether the planned methods and analysis populations were appropriate for the stated non-inferiority objective.

The primary outcome combines absence of active ROP with absence of several unfavorable structural outcomes. Separate component results are not reported in the registry record, making it difficult to assess the contribution of each component. The primary follow-up is at 24 weeks after starting treatment; the reported primary result alone does not establish longer-term outcomes.

The total enrollment and arm sizes constrain the information available for less common outcomes, and the group sizes are unequal. Secondary findings are unadjusted, and the registry does not report a multiplicity strategy. Safety event counts and listed terms do not include enough context about attribution, timing, or overlap to establish comparative causes or a complete safety profile.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Aflibercept for Retinopathy of Prematurity - Intravitreal Injection Versus Laser Therapy — NCT04004208.
• Registry record last updated 2022-05-06; status last verified 2022-05.
• Results first posted 2022-05-06.
• 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.

21. Record Summary

NCT04004208 is a completed phase 3, randomized, open-label, parallel, two-arm study sponsored by Bayer. It enrolled 113 participants with retinopathy of prematurity and compared intravitreal aflibercept 0.4 mg with laser photocoagulation. The recorded study start date is 2019-09-25, and the primary completion date is 2021-02-12.

The primary endpoint was absence of active ROP and unfavorable structural outcomes at 24 weeks after starting treatment. The posted proportions were 0.855 for aflibercept and 0.821 for laser photocoagulation. The registry reports a non-inferiority analysis and also permits an unadjusted recalculation; their estimates and intervals should be interpreted as distinct summaries because the analysis details needed to align them are not reported.

Secondary outcomes address additional treatment, recurrence, ROP activity, adverse events, and plasma concentrations. Safety results include serious adverse-event counts and deaths. Important statistical-plan details—including the non-inferiority margin, missing-data methods, analysis populations, and multiplicity handling—are not reported in the registry record.