Efficacy and Safety of Semaglutide Once Weekly Versus Insulin Glargine Once Daily as add-on to Metformin With or Without Sulphonylurea in Insulin-naïve Subjects With Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy and Safety of Semaglutide Once Weekly Versus Insulin Glargine Once Daily as add-on to Metformin With or Without Sulphonylurea in Insulin-naïve Subjects With Type 2 Diabetes (NCT02128932), a randomized phase 3 study. The primary outcome is Change in HbA1c From Baseline.

2. Clinical Question

3. Trial Design

The registry describes a phase 3, randomized, parallel-group trial with no masking. The listed treatment groups were semaglutide 0.5 mg/week, semaglutide 1.0 mg/week, and insulin glargine. The trial was completed. Its enrollment is listed as 1089, while the primary outcome results report group sizes of 362, 360, and 360. These are registry enrollment and outcome-analysis figures, respectively; the record does not explain any difference between them.

The stated purpose was to compare the effect of once-weekly semaglutide at two dose levels with once-daily insulin glargine on glycaemic control after 30 weeks in insulin-naïve subjects with type 2 diabetes. The treatment context was add-on therapy to metformin, with or without sulphonylurea. The trial was conducted in Africa, North and South America, Asia, and Europe. The registry lists the start as 2014-08-04 and primary completion as 2015-09-03.

Randomization supports a treatment-group comparison by assignment, while the parallel design means participants were assigned to a treatment group rather than crossing between treatments. The registry says the primary analysis used the full analysis set, defined as randomized subjects who received at least one dose of randomized study treatment, and evaluated them according to randomized assignment. The record does not report the randomization method, allocation ratio, masking procedures beyond none, detailed eligibility criteria, dose titration rules, adherence, or the trial’s original sample-size justification. Those details should not be inferred from the posted outcome counts.

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: Insulin-naïve subjects diagnosed with type 2 diabetes and on stable diabetes treatment with metformin or metformin and SU (metformin ≥1500 mg or maximum tolerated dose and SU≥ half of maximum allowed dose according to national label) for at least 90 days before screening. Stable is defined as unchanged medication and unchanged dose.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The full analysis set (FAS) included all randomised subjects who had received at least one dose of randomized semaglutide (s.c.) or insulin glargine. Subjects in the FAS contributed to the evaluation based on the treatment assigned at randomisation.

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: The full analysis set (FAS) included all randomised subjects who had received at least one dose of randomised semaglutide (s.c.) or insulin glargine. Subjects in the FAS contributed to the evaluation based on the treatment assigned at randomisation.

6. Primary Endpoint

The registry lists the primary outcome as Change in HbA1c From Baseline, measured over Week 0, week 30.

The primary endpoint was change in HbA1c from baseline to week 30, expressed as a percentage. The registry identifies the outcome as a least-squares mean, with standard error as the reported measure of dispersion. HbA1c was selected as the measure of glycaemic control, consistent with the trial’s stated purpose of comparing the effect of the treatments on glycaemic control after 30 weeks.

A change score expresses the direction and magnitude of movement from the starting measurement: a negative value indicates a decrease from baseline, while a positive value indicates an increase. In the posted results, all three group means were negative. This describes the average modeled change in each group; it does not mean every participant had the same change. The least-squares mean is a model-based group estimate, not necessarily the simple arithmetic average of individual changes. The registry record does not report the full model specification or which covariates, if any, were included in the primary outcome model.

The endpoint is useful for comparing group-level changes over the stated follow-up period, but it does not by itself describe other aspects of treatment experience or establish how an individual participant’s outcome would differ. The record also lists secondary outcomes, including body weight and fasting plasma glucose, but these do not replace the stated primary endpoint.

7. Results Posted in the Registry

• Semaglutide 1.0 mg/Week: mean -1.64 versus -1.21 on control; difference -0.43 (95% CI -0.57 to -0.29), standardized effect d = -0.45; z-test p = <0.001.
• Insulin Glargine: mean -0.83 versus -1.21 on control; difference 0.38 (95% CI 0.24 to 0.52), standardized effect d = 0.40; z-test p = <0.001.

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 in Body Weight From Baseline

• 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 errors.
• Result:
  • Semaglutide 1.0 mg/Week: mean -5.17 versus -3.47 on control; difference -1.70 (95% CI -2.37 to -1.03), standardized effect d = -0.37; z-test p = <0.001.
  • Insulin Glargine: mean 1.15 versus -3.47 on control; difference 4.62 (95% CI 3.97 to 5.27), standardized effect d = 1.03; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Semaglutide 1.0 mg/Week than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Insulin Glargine than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change in Fasting Plasma Glucose From Baseline

• 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 errors.
• Result:
  • Semaglutide 1.0 mg/Week: mean -49.21 versus -36.74 on control; difference -12.47 (95% CI -18.42 to -6.52), standardized effect d = -0.31; z-test p = <0.001.
  • Insulin Glargine: mean -38.18 versus -36.74 on control; difference -1.44 (95% CI -7.22 to 4.34), standardized effect d = -0.04; z-test p = 0.625.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Semaglutide 1.0 mg/Week than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Insulin Glargine and the control. The interval gives the range of effects that remain compatible with the data.

Change in Diastolic Blood Pressure.

• 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 errors.
• Result:
  • Semaglutide 1.0 mg/Week: mean -0.98 versus -1.38 on control; difference 0.40 (95% CI -0.81 to 1.61), standardized effect d = 0.05; z-test p = 0.516.
  • Insulin Glargine: mean -1.44 versus -1.38 on control; difference -0.06 (95% CI -1.22 to 1.10), standardized effect d = -0.01; z-test p = 0.920.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Semaglutide 1.0 mg/Week and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Insulin Glargine and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary outcomes cover several domains, including body weight, fasting plasma glucose, blood pressure, and patient-reported outcomes. The available continuous results indicate that the direction and size of changes vary by outcome and comparison; not every outcome follows the same pattern. Some comparisons have unadjusted p-values below 0.001, while others do not show a clear difference in the simple recalculation. These results provide additional context around the primary HbA1c finding, but they do not alter what the primary endpoint measures or replace its sponsor-reported analysis.

Secondary outcomes are typically interpreted with more caution because many questions are examined. When numerous outcomes or contrasts are tested, the chance of obtaining at least one apparently notable result by chance can increase. The registry record does not report a multiplicity adjustment or a hierarchy for interpreting secondary outcomes. The recalculated secondary comparisons are explicitly unadjusted. Their p-values therefore should not be read as if they were protected from the consequences of multiple testing.

The posted secondary results can be useful for describing the breadth of measured outcomes and identifying patterns that may merit further study. They are less definitive than a prespecified primary analysis when the analysis plan, multiplicity strategy, or status of each outcome as confirmatory or exploratory is unclear. The registry record lists seven secondary outcomes, while the posted numerical secondary results include a smaller set of continuous measures. Further details on the remaining outcomes and their analyses are not reported in the registry record. It is not appropriate to infer unreported results or to treat an absence of posted numerical detail as evidence of no effect.

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

• Change in Body Weight From Baseline
• Change in Fasting Plasma Glucose From Baseline
• Change in Diastolic Blood Pressure.
• Change in Systolic Blood Pressure.
• Change in Patient Reported Outcome (PRO) Questionnaire, Questionnaire SF-36v2™
• Change in Patient Reported Outcome Questionnaires. (PROs), Diabetes Treatment Satisfaction Questionnaire (DTSQs)
• Subjects Who Achieve HbA1c ≤6.5% (48 mmol/Mol), American Association of Clinical Endocrinologists (AACE)

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: it records a numerical change in HbA1c rather than classifying participants into categories. The registry reports least-squares means for change from baseline and standard errors. A least-squares mean is an estimate from a statistical model. It can account for model terms specified by the investigators, whereas a raw mean is calculated directly from the observed values. The registry record does not report enough detail to identify the terms in the sponsor’s primary model.

What does the standard error describe?

A standard error describes the statistical uncertainty in an estimated group mean. It is not the same as the spread of individual participant measurements. A smaller standard error generally indicates a more precisely estimated mean, given the model and data. The posted standard errors are 0.05 in each primary outcome group. The record does not give group standard deviations for this endpoint; the recalculated comparisons use an approximate pooled standard deviation derived from the standard errors and group sizes.

How should the mean differences be read?

A mean difference subtracts one group estimate from another in a stated order. The unadjusted recalculation contrasts semaglutide 1.0 mg/week with semaglutide 0.5 mg/week, and insulin glargine with semaglutide 0.5 mg/week. The recalculated estimates are -0.43 and 0.38. A negative difference means the first group in the stated contrast has the lower change value; a positive difference means it has the higher change value. Because each group mean is itself a change from baseline, the contrast describes the difference between group-level changes, not a change within a single participant.

The sponsor’s mixed-model contrasts instead compare each semaglutide group with insulin glargine. Those are different comparisons and produce different estimates. Model-based contrasts can differ from simple arithmetic contrasts when the model accounts for factors or data structures that the simple calculation does not. The sponsor’s own analysis takes precedence for reporting its analysis; the unadjusted recalculations are not a substitute for it.

What does a confidence interval add?

A confidence interval gives a range of values compatible with the estimate under the analysis model and its assumptions. It conveys precision as well as direction. An interval that excludes zero is evidence against a zero treatment difference under the corresponding test framework, but the interval does not establish clinical importance. The sponsor’s 95% confidence intervals for both reported comparisons lie below zero. They are model-based intervals, and their interpretation depends on the validity of the model and the assumptions used to handle the data.

What does a p-value say?

A p-value measures how unusual the observed result, or a more extreme one, would be if the null hypothesis and analysis assumptions held. It is not the chance that the null hypothesis is true, and it does not quantify the size of a treatment effect. The sponsor reports p < 0.001 for both comparisons; the simple recalculations report p < 0.001. These values belong to different analyses and should not be treated as interchangeable.

Why do model and design details matter?

The randomized parallel design gives a basis for comparing assigned groups, but the trial was not masked. Knowledge of assigned treatment can affect care, reporting, or assessment, depending on the outcome and study procedures. For a continuous laboratory measure, that concern differs from an outcome based entirely on subjective reporting, but the open-label design remains relevant context. The record identifies a full analysis set and assignment-based evaluation. The detailed statistical analysis plan, model specification, missing-data method, and any interim analyses are not reported in the registry record. These omissions limit how fully the sponsor analysis can be reconstructed from the registry summary.

Over-dispersion is a general concern when observed variability exceeds what a statistical model assumes. If present and unaccounted for, standard errors and p-values may be unreliable. The registry record does not report an over-dispersion assessment for this continuous endpoint. That concept should not be taken as evidence that over-dispersion occurred in this trial.

11. Confidence Intervals and Precision

Precision is reflected by the uncertainty around an estimate, not just by the estimate’s point value. For the sponsor’s mixed-model comparison of semaglutide 1.0 mg/week with insulin glargine, the treatment difference is -0.81 and the 95% confidence interval runs from -0.96 to -0.67. For semaglutide 0.5 mg/week versus insulin glargine, the treatment difference is -0.38 and the 95% confidence interval runs from -0.52 to -0.24. Both intervals lie below zero, so both reported estimates indicate a lower change value in the semaglutide group under the sponsor’s comparison convention.

The intervals quantify uncertainty under the mixed-model analysis; they do not capture every source of uncertainty. In particular, they do not automatically account for possible bias from open-label conduct, unreported analytic choices, multiplicity, or missing data. A narrow interval can still be conditional on assumptions that readers cannot verify from the registry record.

The recalculated comparisons are based on the posted least-squares means, standard errors, and group sizes, with an approximate pooled standard deviation. Their intervals are simple unadjusted recalculations rather than the sponsor’s mixed-model intervals. The recalculated mean differences are -0.43 for semaglutide 1.0 mg/week versus semaglutide 0.5 mg/week and 0.38 for insulin glargine versus semaglutide 0.5 mg/week. They should not be confused with the sponsor-reported contrasts against insulin glargine.

Confidence intervals address statistical precision, not whether a difference is large enough to matter in practice. The registry record does not report a threshold for a clinically important HbA1c difference. Nor does it report a non-inferiority margin, despite its classification of the sponsor analyses as non-inferiority or equivalence analyses. Those missing details constrain interpretation of the interval relative to the trial’s original decision criteria.

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 estimates express the number of participants per arm for 80% power or 90% power under an illustrative calculation. These are not total event counts: the primary endpoint and other continuous outcomes use participants per arm. The planning assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Each calculation uses a stated fraction of the observed effect as the effect to detect. Time-to-event calculations, if present, would instead give total events, but the posted outcomes are continuous.

The 80% power column means the illustrative design would have an 80% chance of detecting the specified effect at the stated significance level if the calculation’s assumptions held. The 90% column uses a higher probability of detecting that same effect, and therefore generally calls for more participants per arm. Rows that plan for 75% or 50% of the observed effect represent smaller target effects than planning for 100% of the observed effect. Smaller effects are harder to distinguish from random variation, so they require more participants per arm in these illustrations.

These estimates are based on the observed treatment effects and the stated simplifying assumptions, not on the trial’s original design calculation. The original sample-size justification and its assumptions are not reported in the registry record. The estimates are planning illustrations, not observed power: they do not describe the probability that this completed trial had of finding its observed result, and they should not be used to infer the trial’s design process. They also do not account for every practical or analytic feature that could affect sample-size requirements, such as missing outcomes or a different planned model.

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 = -0.430; pooled standard deviation σ = 0.950.
2. n = 2 × (1.96 + 0.842)² × 0.950² / -0.430² = 76.6, rounded up to 77 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 <- -0.43                      # observed difference in means
sigma <- 0.950003654963729                     # 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(2128932)
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 = 77 (80% power), 103 (90% power)
 75% of observed effect: n per arm = 137 (80% power), 183 (90% power)
 50% of observed effect: n per arm = 307 (80% power), 411 (90% power)
power.t.test: 77.6 
Simulated power at n = 77 per arm: 0.799 (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 mixed-model treatment differences for semaglutide 1.0 mg/week versus insulin glargine and semaglutide 0.5 mg/week versus insulin glargine. Those estimates are -0.81 and -0.38, respectively, each with a 95% confidence interval and p < 0.001. The sponsor labels these analyses as non-inferiority or equivalence analyses. The non-inferiority margin and full testing details are not reported in the registry record.

The simple recalculations use the posted group means, standard errors, and group sizes. They compare semaglutide 1.0 mg/week with semaglutide 0.5 mg/week, and insulin glargine with semaglutide 0.5 mg/week; they are not reproductions of the sponsor’s two contrasts. The recalculated p-values are unadjusted and use an approximate pooled standard deviation. The record indicates no discrepancy between these recalculations and the posted summaries, but that does not make them equivalent to the sponsor’s model-based analysis.

For the sponsor’s inferential claims, its own analysis takes precedence. The recalculations offer a transparent, unadjusted check of contrasts that can be derived from the posted summaries. They should be interpreted as supplementary rather than as a replacement or correction of the sponsor’s analysis.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when investigators test several outcomes, treatment contrasts, time points, or subgroups. Without an adjustment or an ordered testing strategy, the overall chance of a false-positive finding can rise across the collection of tests. The registry lists multiple secondary outcomes and two sponsor-reported primary comparisons. The multiplicity approach, any prespecified outcome hierarchy, and whether secondary analyses were confirmatory are not reported in the registry record. The simple recalculated p-values are unadjusted and should be treated accordingly.

The record describes the primary analysis population as the full analysis set, including randomized subjects who received at least one dose of randomized treatment, analyzed according to treatment assigned at randomization. This is the population definition available in the registry. The handling of participants without a week 30 measurement, the amount and pattern of missing data, and any imputation or sensitivity analyses are not reported in the registry record. It is therefore not possible to assess from this summary how missing outcomes may have affected the estimates.

The sponsor used mixed models for the reported primary treatment comparisons. The complete model specification, assumptions about missingness, covariates, and any sensitivity analyses are not reported in the registry record. No interim analyses or stopping rules are reported in the registry record. These omissions do not establish that the analyses were inappropriate; they limit independent assessment of how robust the results are to analytic choices.

The record labels the sponsor’s comparisons as non-inferiority or equivalence analyses, but does not report the margin or detailed decision criteria. It also does not provide the full statistical analysis plan. Without these details, the reported estimates and intervals can be described, but the original confirmatory framework cannot be fully reconstructed from the registry entry alone.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: osteoarthritis (0.6%), transient ischaemic attack (0.6%), hyponatraemia (0.6%), pneumonia (0.6%), acute kidney injury (0.3%).

The safety summary reports participants with serious adverse events and participants with other adverse events by treatment group. Serious adverse events were reported for 22 participants in the semaglutide 0.5 mg/week group, 17 in the semaglutide 1.0 mg/week group, and 18 in the insulin glargine group. Other adverse events were reported for 172, 192, and 107 participants, respectively. These are counts, not estimates of comparative risk, and no comparative safety analysis is included in the posted record.

The listed serious adverse event terms include osteoarthritis, transient ischaemic attack, hyponatraemia, and pneumonia, each reported as 0.6%, and acute kidney injury and anaemia, each reported as 0.3%. The record does not provide enough detail in this summary to establish how event terms were selected for listing or how events were adjudicated. Death counts are not reported in the registry record.

Safety counts should be read in the context of the follow-up period, participant exposure, definitions, and completeness of reporting. The timeframe and further details needed for a more complete comparative safety assessment are not reported in the registry record. The counts alone do not establish causality or determine whether one treatment is safer.

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

The posted data describe average changes in HbA1c over the trial’s 30-week period for three randomized treatment groups. The sponsor’s mixed-model results indicate negative treatment differences for both semaglutide doses compared with insulin glargine, with reported 95% confidence intervals below zero. The primary group estimates also show negative average changes in all three groups. These are statistical descriptions of the analyzed groups, not a guarantee about an individual participant’s outcome.

The results do not, by themselves, establish the clinical importance of the differences. The registry record does not report a threshold for a clinically important change in HbA1c or the non-inferiority margin used by the sponsor. A small p-value is not a measure of clinical value, and a confidence interval is not a statement about the probability that a treatment is beneficial. The record also does not permit a full independent evaluation of the model, missing-data assumptions, or multiplicity handling.

The trial was randomized but open-label. Randomization helps make treatment groups comparable in expectation, while lack of masking can affect treatment behavior, reporting, or assessment. The registry identifies the full analysis set and says participants were evaluated according to randomized assignment after receiving at least one dose. It does not report detailed adherence, withdrawals, or missing-outcome patterns.

These findings should therefore be understood as the posted statistical results for this trial, population, endpoint, and timeframe. They do not establish results beyond the measured follow-up or for people outside the described study population. The record supports description of the observed group estimates and sponsor analyses, but does not provide every protocol and analysis detail needed to resolve all questions about confirmatory interpretation.

18. Why This Trial Matters Statistically

This trial is statistically informative because it reports a continuous primary endpoint in three randomized parallel groups and includes both sponsor-model results and enough summary statistics for simple unadjusted recalculations. That combination lets readers distinguish a model-based comparison from a calculation based directly on group estimates. It also makes clear that the contrast being examined matters: the sponsor compares each semaglutide dose with insulin glargine, while the posted recalculation uses semaglutide 0.5 mg/week as the reference for its two contrasts.

The entry also illustrates why estimates, uncertainty, and analysis design should be read together. Point estimates describe direction and magnitude; confidence intervals communicate precision under the chosen model; p-values summarize compatibility with a null hypothesis under assumptions. None of these alone describes clinical importance. The registry’s classification of the sponsor tests as non-inferiority or equivalence raises the importance of knowing the prespecified margin, which is not reported in the registry record.

Finally, the registry includes a broad set of secondary measures and safety counts but does not report the multiplicity strategy, complete missing-data approach, or full statistical analysis plan. This distinction between available results and unavailable analytic detail is important in interpreting trial registrations: the numerical record can support careful description, while still leaving limits on independent verification and on the strength of confirmatory conclusions.

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, full statistical analysis plan, detailed mixed-model specification, multiplicity strategy, or missing-data methods. These are especially relevant because the sponsor’s primary analyses are labeled as non-inferiority or equivalence analyses and multiple secondary outcomes are listed. Without the margin and testing framework, the original decision criteria cannot be fully evaluated from the registry record.

The trial was open-label, and the record does not describe additional safeguards against potential effects of treatment knowledge. The available summary also does not report detailed participant flow, withdrawals, adherence, exposure, or completeness of week 30 measurements. Although the primary analysis population is defined, the effect of missing outcome data cannot be assessed from the posted summary.

The unadjusted recalculations are based on least-squares means, standard errors, and group sizes, with an approximate pooled standard deviation. They are not a substitute for the sponsor’s mixed-model analysis and may not reproduce its assumptions or adjustments. The recalculated secondary comparisons also do not account for multiplicity. The registry reports a selected set of numerical secondary outcomes, not detailed results for every listed secondary measure.

Finally, the registry provides group-level results at the stated endpoint and timeframe. It does not report a clinically important threshold for the primary outcome in this record, and the posted summary does not establish how results would generalize beyond the study population or follow-up period. These limitations qualify interpretation of the statistical evidence; they do not, on their own, determine the direction or validity of the underlying results.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Efficacy and Safety of Semaglutide Once Weekly Versus Insulin Glargine Once Daily as add-on to Metformin With or Without Sulphonylurea in Insulin-naïve Subjects With Type 2 Diabetes — NCT02128932.
• Registry record last updated 2019-06-13; status last verified 2019-05.
• Results first posted 2018-03-08.
• 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

NCT02128932 is a completed phase 3 randomized, open-label, parallel-group trial comparing semaglutide once weekly at two dose levels with insulin glargine once daily in insulin-naïve subjects with type 2 diabetes receiving metformin, with or without sulphonylurea. The primary endpoint was change in HbA1c from baseline to week 30, analyzed in the full analysis set as described in the registry.

The sponsor’s mixed-model results report treatment differences against insulin glargine for both semaglutide doses, with 95% confidence intervals and p < 0.001. Simple recalculations from posted summary statistics address different contrasts and are unadjusted. Secondary outcomes and adverse event counts add context, but multiplicity, missing-data methods, the non-inferiority margin, and the full analysis plan are not reported in the registry record.