Research Study Comparing a New Medicine "Fast-acting Insulin Aspart" to Another Already Available Medicine "NovoRapid"/"NovoLog" in People With Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Research Study Comparing a New Medicine "Fast-acting Insulin Aspart" to Another Already Available Medicine "NovoRapid"/"NovoLog" in People With Type 2 Diabetes (NCT03268005), a randomized phase 3 study. The primary outcome is Change in Glycosylated Haemoglobin (HbA1c).

2. Clinical Question

3. Trial Design

This was a randomized, quadruple-masked, parallel-group phase 3 study. It was sponsored by Novo Nordisk A/S and is listed as completed. The registry reports enrollment of 1264 participants with type 2 diabetes. The study compared faster-acting insulin aspart with insulin aspart (NovoRapid/NovoLog); both regimens included insulin degludec, with or without metformin.

The study description says the assigned treatment was decided by chance and that participants would use injections four times every day. It describes a study lasting about 8 months (34 weeks). The primary endpoint was assessed at week 16, using change from baseline at week 0. Its stated observation period ran from randomization until the last trial-related participant-site contact. The reported primary analysis includes participants with available data.

The registry lists the randomization and masking features, but does not report details such as the method used to generate or conceal allocation, which parties were masked, or how masking was maintained. The arm descriptions do not add details beyond the listed regimens. The registry record also does not report the trial’s original sample-size justification, its non-inferiority margin, or detailed analysis-plan specifications. These omissions limit what can be concluded about design choices from the posted record alone.

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: There was a 12-week run-in period primarily for optimisation of the basal insulin and reinforcement of subject training in trial procedures, diabetes education and dietary training. Out of the 1264 participants enrolled in the study, 173 were run-in failures and 1091 were randomised.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Overall number of participants analyzed = number of participants with available data.

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.

6. Primary Endpoint

The registry lists the primary outcome as Change in Glycosylated Haemoglobin (HbA1c), measured over Week 0, week 16.

The primary endpoint was change from baseline in glycosylated haemoglobin, or HbA1c, measured as a percentage of HbA1c. The registry specifies the baseline as week 0 and the endpoint assessment as week 16. The analysis population is described as participants with available data during the in-trial observation period.

In general, HbA1c is used as a measure of average blood glucose over a period of time, rather than as a single glucose measurement. A change from baseline expresses how the measured value at follow-up differs from the participant’s starting value. Comparing the groups’ mean changes can therefore describe their average change over the assessment period. That group comparison does not, by itself, show how much individual participants varied or whether the trial’s intended non-inferiority criterion was met; those questions depend on the analysis and its prespecified decision rules.

The registry reports group means and standard deviations for the primary endpoint, as well as a sponsor analysis of the treatment difference. The sponsor’s analysis is specifically classified as non-inferiority. The margin needed to interpret that classification against a decision threshold is not reported in the registry record.

7. Results Posted in the Registry

• NovoRapid: mean -0.09 versus -0.15 on control; difference 0.06 (95% CI -0.01 to 0.13), standardized effect d = 0.10; z-test p = 0.106.

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.

Change From Baseline in 1-hour PPG Increment

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • NovoRapid: mean 0.08 versus -0.43 on control; difference 0.51 (95% CI 0.20 to 0.82), standardized effect d = 0.20; z-test p = 0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for NovoRapid than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in 1,5-anhydroglucitol

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • NovoRapid: mean 0.89 versus 1.38 on control; difference -0.49 (95% CI -0.88 to -0.10), standardized effect d = -0.15; z-test p = 0.013.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for NovoRapid than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Fasting Plasma Glucose (FPG)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • NovoRapid: mean 0.68 versus 0.56 on control; difference 0.12 (95% CI -0.17 to 0.41), standardized effect d = 0.05; z-test p = 0.413.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between NovoRapid and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists a broad set of secondary outcomes, including post-meal glucose measures, fasting plasma glucose, 1,5-anhydroglucitol, HbA1c threshold outcomes, and self-measured glucose profiles. The posted continuous secondary results do not all point in the same direction or show the same degree of statistical evidence. This range of findings is a reminder that secondary measures address different aspects of glucose control and are not interchangeable with the primary HbA1c endpoint.

The secondary findings can add context to the primary result, but they do not independently settle the primary non-inferiority question. The primary endpoint was change in HbA1c at week 16, while secondary outcomes measure other markers or time points and may have different available-data counts. A secondary result can be supportive or exploratory without changing what the primary analysis estimates.

The recalculated secondary p-values are unadjusted. When a study evaluates many outcomes, some results may appear notable by chance even if there are no underlying differences. In general, multiplicity procedures can help control that risk, but the registry record does not report an adjustment strategy or a prespecified hierarchy for the secondary outcomes. Their p-values should therefore be read as descriptive, nominal evidence rather than as a collection of confirmatory tests.

The registry lists 52 outcome measures, while the detailed secondary entries and analyses are not a complete account of all outcomes in the record. The absence of a detailed posted result for a measure should not be interpreted as evidence that the groups were identical on it. The available secondary information is best considered alongside the primary endpoint and with caution about multiplicity and selective emphasis.

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

• Change From Baseline in 1-hour PPG Increment
• Change From Baseline in 1,5-anhydroglucitol
• Change From Baseline in Fasting Plasma Glucose (FPG)
• Participants Who Achieved HbA1c <7.0% (53 mmol/L) (Yes/No)
• Participants Who Achieved HbA1c <7.0% (53 mmol/L) Without Severe Hypoglycaemia Episodes (Yes/No)
• Change From Baseline (Week 0) in 30-minute, 1-hour, 2-hour, 3-hour and 4-hour Postprandial Glucose (PPG [Meal Test])
• Change From Baseline (Week 0) in 30-minute, 1-hour, 2-hour, 3-hour and 4-hour PPG Increment (Meal Test)
• Change From Baseline in Mean of the 7-9-7 Point Self-measured Plasma Glucose (SMPG) Profile

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: it is a measured HbA1c value, and the registry reports group means and standard deviations. A mean describes the group’s average observed change. A standard deviation describes how spread out individual values are around that average. These are different summaries: the mean concerns the group’s central result, while the standard deviation concerns variation among participants.

The unadjusted recalculation uses the posted group means and standard deviations to estimate a difference between group means. It does not reproduce the sponsor’s full analysis of variance model. In general, an unadjusted comparison treats the group summaries as the basis for the estimate and does not account for covariates or other model features unless those are already built into the summaries. The registry record does not report enough model detail to identify all features of the sponsor’s analysis beyond its stated use of analysis of variance.

The direction of a mean difference depends on the order used for subtraction. For the recalculated primary comparison, the estimate is expressed as NovoRapid minus faster aspart. A positive estimate therefore means the posted mean change was numerically higher in the NovoRapid group under that convention. The sponsor’s reported treatment difference uses a separately reported analysis; its numerical direction should be read in the context of that analysis rather than assumed to follow the recalculation’s subtraction order.

A confidence interval gives a range of values compatible with the data and the specified statistical procedure. In general, wider intervals indicate less precision, while narrower intervals indicate greater precision. A confidence interval is not a range containing a fixed proportion of individual participant outcomes, and it does not assign a probability to the true effect after the data have been observed. For a difference, an interval that includes zero is compatible with no difference under the interval’s method, but that is not equivalent to demonstrating equivalence or non-inferiority.

The sponsor labels its primary analysis as non-inferiority. In general, a non-inferiority analysis asks whether the result rules out a loss larger than a prespecified margin in the chosen direction. The conclusion depends on both the estimated treatment difference and the margin, with the confidence interval evaluated against the margin. The registry record does not report the margin, its justification, or the precise decision rule. It is therefore possible to describe the posted estimate and interval, but not to independently verify the non-inferiority conclusion from the registry record alone.

The unadjusted p-value for the primary recalculation is 0.106. In general, a p-value describes how unusual data at least as discrepant as those observed would be under a specified null hypothesis and model. It is not the probability that the null hypothesis is true, nor is it a measure of the magnitude or clinical importance of an effect. The recalculated p-value is not a substitute for the sponsor’s non-inferiority analysis, because it does not use a reported non-inferiority margin.

The registry lists 52 outcome measures and one analysis for the primary result. For the primary endpoint, available-data counts are reported, but the registry record does not report the approach to missing outcomes, imputation, sensitivity analyses, or a detailed definition of the analysis population beyond available data. Those choices can affect a trial estimate, especially when missingness differs between groups or relates to outcomes. Without the relevant details, the posted summaries cannot establish how such issues were handled.

For secondary continuous measures, the same general distinction between means, standard deviations, group differences, and uncertainty applies. Their recalculated p-values are unadjusted. When many outcomes are assessed, the chance of observing apparently notable results by chance can increase. General methods to address this include prespecified testing sequences or adjustments for multiple comparisons, but the registry record does not report whether such procedures were planned or used for the posted secondary outcomes.

Finally, the registry includes planning illustrations based on observed effects. These are not the trial’s original design calculations. The original sample-size assumptions, including any assumptions tied to the non-inferiority objective, are not reported in the registry record. The planning illustrations should therefore be read as hypothetical demonstrations of sample sizes under their stated assumptions, not as evidence about the trial’s original planning or the strength of its observed result.

11. Confidence Intervals and Precision

The sponsor’s primary estimate is -0.04, with a 95% confidence interval from -0.11 to 0.03. This interval spans zero. In general, a confidence interval expresses the uncertainty around an estimate under the method used to construct it; the endpoints indicate values that remain compatible with the data and that method. The interval should not be read as an assurance that the true effect lies within those endpoints, nor as a statement about the distribution of individual participant outcomes.

The unadjusted recalculation gives a mean difference of 0.06 and an unadjusted p-value of 0.106. The recalculated interval also includes zero. The sponsor estimate and the recalculation are based on different reported analyses, so their precision summaries should not be combined or treated as if they were the same estimate. The registry does not report the full sponsor model specification needed to reconstruct its interval.

The sponsor interval allows for effects on either side of zero under the stated contrast. Whether its range rules out an unacceptable loss depends on the non-inferiority margin and the direction in which that margin is defined. Neither the margin nor its justification is reported in the registry record. Thus, the interval communicates uncertainty about the treatment difference but does not, without the margin, permit an independent check of the non-inferiority decision.

Precision also differs from practical importance. A confidence interval can be interpreted statistically, but the registry does not provide a threshold for judging what size of HbA1c difference would be important in this study. The available estimates should therefore be read as descriptions of the posted group comparison and its uncertainty, not as a standalone judgment of clinical relevance.

12. Planning a Similar Trial: Sample Size

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

The planning figures are participants per arm, not total events. They are illustrations computed from the observed effect, using a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each planning row uses the stated fraction of the observed effect as the effect to detect. The trial’s original sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the illustrated participants per arm for a hypothetical study to have an 80% chance of detecting the specified effect under the stated assumptions, if that effect is the true effect and the statistical model applies. The 90% power column uses the same interpretation with a higher target chance of detection. These are prospective planning quantities, not descriptions of the probability that this completed study found a real effect.

Smaller effects require more participants to detect under otherwise unchanged assumptions. The displayed planning scenarios illustrate that relationship by considering the full observed effect and smaller fractions of it. They are not a reconstruction of the trial’s planned sample size, and they do not establish that the observed effect is stable or likely to recur. These figures 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

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 0.060; pooled standard deviation σ = 0.610.
2. n = 2 × (1.96 + 0.842)² × 0.610² / 0.060² = 1,623.2, rounded up to 1,624 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 0.06                      # observed difference in means
sigma <- 0.610128207209102                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(3268005)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 1,624 (80% power), 2,174 (90% power)
 75% of observed effect: n per arm = 2,886 (80% power), 3,864 (90% power)
 50% of observed effect: n per arm = 6,493 (80% power), 8,693 (90% power)
power.t.test: 1624.2 
Simulated power at n = 1,624 per arm: 0.803 (20000 replicates)

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

14. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary analysis as an analysis of variance treatment difference of -0.04, with a 95% confidence interval from -0.11 to 0.03 and p = 0.310. It is classified as a non-inferiority analysis. The registry record does not state the non-inferiority margin, so the posted estimate and interval cannot be checked against that decision boundary independently.

The recalculation is a simple, unadjusted comparison of the posted arm means and standard deviations. It gives a mean difference of 0.06, expressed as NovoRapid minus faster aspart, and an unadjusted p-value of 0.106. It does not replace or reproduce the sponsor’s analysis of variance result. The registry does not supply enough detail to explain the difference between the estimates, and it should not be attributed to a particular adjustment or model feature.

The discrepancy flag is false. Accordingly, the registry does not identify an inconsistency that would call for giving the recalculation precedence. For the trial’s stated non-inferiority analysis, the sponsor’s own result is the relevant reported analysis; the recalculation is a supplementary, unadjusted view of the posted group summaries.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies a primary endpoint and lists 52 outcome measures, but the record does not report whether the trial used a formal strategy to control multiplicity across secondary outcomes. It also does not report a hierarchy, gatekeeping sequence, or family-wise adjustment for the secondary analyses. Because testing many outcomes can increase the chance of at least one apparently unusual result, unadjusted secondary p-values warrant less weight than a prespecified primary analysis.

The record describes the primary analysis population as participants with available data and specifies an in-trial observation period from randomization until the last trial-related participant-site contact. It does not report how missing endpoint values were handled, whether any imputation was used, or whether sensitivity analyses assessed the effect of missing data. The reasons for missing data and their distribution across groups are also not reported in the registry record.

The available-data description does not establish whether the analysis followed an intention-to-treat, per-protocol, or other formally defined population. Those distinctions can matter in a non-inferiority trial, because exclusions and missing outcomes may influence the comparison. The registry record does not provide enough information to assess that possibility.

The sponsor analysis is described as analysis of variance and labeled non-inferiority, but the model specification, covariates, handling of repeated or incomplete measurements, and non-inferiority margin are not reported in the registry record. These omissions constrain independent assessment of the analysis plan and its robustness. They do not establish that the methods were inadequate; they mean that the posted record does not document the details needed to evaluate them.

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: hypoglycaemia (0.7%), hypoglycaemic unconsciousness (0.6%), accidental overdose (0.4%), acute coronary syndrome (0.4%), angina pectoris (0.4%).

The registry reports serious adverse events for 38 of 544 participants in the faster aspart group and 40 of 544 in the NovoRapid group. It also lists other adverse events for 32 of 544 and 33 of 544, respectively. Deaths are reported for 2 participants in the faster aspart group and 1 in the NovoRapid group. These are counts recorded in the registry; they do not, by themselves, show whether an event was caused by treatment.

The listed serious-event terms include hypoglycaemia (0.7%), hypoglycaemic unconsciousness (0.6%), accidental overdose (0.4%), acute coronary syndrome (0.4%), angina pectoris (0.4%), and cardiac failure (0.4%). The record does not provide a full narrative of these events or enough detail to interpret their timing, relatedness, duration, or clinical course. The term percentages should be read as posted rather than used to infer an unreported comparison.

Safety counts are descriptive. Differences in counts do not by themselves demonstrate that one regimen caused more or fewer events, and the registry record does not report a comparative statistical analysis of these adverse-event data.

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

The posted results describe average changes in HbA1c in two randomized groups receiving different forms of insulin aspart with insulin degludec, with or without metformin. The sponsor reports a treatment difference of -0.04, with a 95% confidence interval from -0.11 to 0.03. The interval includes zero, and the registry classifies the analysis as non-inferiority. Because the margin is not reported, the registry information is insufficient to independently verify the non-inferiority criterion.

The unadjusted recalculation describes a mean difference of 0.06, expressed as NovoRapid minus faster aspart, with an unadjusted p-value of 0.106. It is not the sponsor’s non-inferiority test and should not be read as one. Neither estimate alone describes every participant’s response; group averages can conceal individual variation, which is reflected in the reported standard deviations.

These results do not establish that the treatments are identical. A result compatible with no difference under a particular analysis is not proof of equivalence, and non-inferiority requires comparison with a prespecified margin. The registry record does not report that margin. Likewise, the posted estimates do not establish the clinical importance of a difference, because the record does not provide a threshold for interpreting the size of an HbA1c change in this context.

The findings also do not determine what any individual should do. They are trial-level summaries of an endpoint assessed at week 16, with an available-data analysis as described in the registry. The design, missing-data approach, and full statistical plan are not documented in sufficient detail to support broader conclusions beyond the results reported.

18. Why This Trial Matters Statistically

This trial illustrates why the analysis label matters. A non-inferiority study is not interpreted simply by asking whether a conventional difference test is statistically significant. The non-inferiority conclusion depends on a prespecified margin and on whether the confidence interval rules out a loss beyond that margin. The registry labels the sponsor result as non-inferiority but does not report the margin, leaving an important part of the decision framework unavailable to readers.

It also demonstrates the distinction between a sponsor-reported model-based estimate and a recalculation from group summaries. The latter can help readers understand the numerical relationship between posted means, but it cannot recreate a full analysis of variance model or substitute for the stated primary analysis. Reporting both estimates is useful only when their different bases and purposes remain clear.

Finally, the record includes many outcome measures alongside the primary endpoint. Secondary measures may provide context, but unadjusted testing across numerous outcomes raises multiplicity concerns. Clear reporting of a testing hierarchy, missing-data methods, analysis populations, and the non-inferiority margin would help readers assess how the primary conclusion was reached and how much weight to place on secondary findings.

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 registry record does not report the non-inferiority margin or its rationale, so the sponsor’s stated non-inferiority analysis cannot be independently checked against its decision threshold. It also does not report the original sample-size calculation or design assumptions. The planning illustrations based on observed effects do not fill that gap.

The record gives an available-data description for the primary analysis but does not explain missing-data handling, imputation, sensitivity analyses, or the distribution and reasons for missing outcomes. It does not provide enough detail to determine whether the analysis population was intention-to-treat, per-protocol, or another defined population. These are important uncertainties when interpreting a non-inferiority comparison.

Further details of the sponsor’s statistical model, including covariate adjustment and other model specifications, are not reported in the registry record. Nor does the record report a multiplicity strategy for the secondary outcomes. As a result, the unadjusted secondary results should not be treated as a set of confirmatory claims.

The available results also do not provide detailed participant characteristics, event narratives, or a comparative statistical analysis of safety outcomes. The registry reports a study duration of about 34 weeks, while the primary endpoint is assessed at week 16; these results therefore describe the posted assessment, not unreported longer-term outcomes. These limits concern what can be assessed from the registry record and do not establish that additional information was absent from the full study documentation.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Research Study Comparing a New Medicine "Fast-acting Insulin Aspart" to Another Already Available Medicine "NovoRapid"/"NovoLog" in People With Type 2 Diabetes — NCT03268005.
• Registry record last updated 2022-01-11; status last verified 2022-01.
• Results first posted 2020-03-12.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The registry lists NCT03268005 as a completed, randomized, quadruple-masked, parallel-group phase 3 study with enrollment of 1264 adults with type 2 diabetes. It compared faster-acting insulin aspart with NovoRapid/NovoLog, both used with insulin degludec, with or without metformin.

The primary endpoint was change from baseline in HbA1c at week 16. The sponsor reported an analysis of variance treatment difference of -0.04, with a 95% confidence interval from -0.11 to 0.03 and p = 0.310, classified as non-inferiority. An unadjusted recalculation from posted group summaries gives a mean difference of 0.06 and p = 0.106. The non-inferiority margin and several analysis details are not reported in the registry record, limiting independent interpretation of the sponsor’s conclusion.