Long Term Comparative Effectiveness of Once Weekly Semaglutide Versus Standard of Care in a Real World Adult US Population With Type 2 Diabetes - a Randomized Pragmatic Trial: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Long Term Comparative Effectiveness of Once Weekly Semaglutide Versus Standard of Care in a Real World Adult US Population With Type 2 Diabetes - a Randomized Pragmatic Trial (NCT03596450), a randomized phase 4 study. The primary outcome is Number of Participants With Glycosylated Haemoglobin (HbA1c) Less Than 7.0 Percentage (%) (53 Millimoles Per Mole [mmol/Mol]) at Year 1 (Yes/No).

2. Clinical Question

The registry describes a randomized comparison of treatment strategies under routine-care conditions rather than a comparison in which treatment is fixed throughout follow-up. Participants were followed for 2 years regardless of changes in antidiabetic treatment. The primary endpoint is a yes-or-no classification at year 1: whether a participant’s HbA1c was below the stated threshold.

This framing matters for interpretation. The question concerns the proportion of participants meeting a specified HbA1c threshold at a particular follow-up point, not the average amount of HbA1c change, and not an outcome measured continuously over the full follow-up period. The registry lists changes in HbA1c and body weight among secondary outcomes, along with other HbA1c target measures. Those outcomes address different questions from the primary endpoint and should not be treated as interchangeable.

Arm-specific counts for the primary endpoint and a comparative effect estimate are not reported in the registry record. Consequently, the registry description establishes what was to be compared, but it does not show the magnitude or precision of a difference between assigned groups for this endpoint.

3. Trial Design

This was a phase 4, randomized, open-label, parallel-group trial with an enrollment of 1278. The registry lists Novo Nordisk A/S as sponsor and gives the status as completed. The recorded start date is 2018-07-13, and the primary completion date is 2022-06-09.

Participants were assigned to semaglutide or standard of care. The registry describes semaglutide as a subcutaneous treatment and standard of care as oral or injectable antidiabetic medication. In both groups, the assigned treatment was used as intensification in addition to metformin monotherapy in routine clinical practice. The trial followed participants for 2 years regardless of changes in antidiabetic treatment during the study.

The trial was open-label: the masking field is recorded as none. This means the registry does not describe blinding of participants or study personnel. Random allocation and masking are separate design features. Random allocation helps create groups that are comparable on average at the point of assignment; it does not itself conceal treatment after assignment or prevent later differences in care or outcome measurement.

The pragmatic, routine-practice setting is relevant to the comparison being asked. It allows treatment to change during follow-up, as described in the arm records, so the comparison is between assigned treatment strategies as implemented in that setting rather than an assurance that every participant stayed on the initial treatment. The registry does not report additional details about how adherence, treatment changes, or follow-up were handled in the statistical analysis.

The total enrollment is not the same as the number with an observed primary endpoint. The primary-outcome description says the analysis population included participants with available data for that measure, but the number analyzed for the endpoint is not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: A total of 1,278 participants were randomized 1:1 to receive either semaglutide or standard of care (SOC).

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: FAS included all randomized participants analyzed according to the treatment group to which they were assigned at randomization. Overall Number of Participants Analyzed = participants with available data for this outcome measure.

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: Full analysis set (FAS) included all randomized participants analyzed according to the treatment group to which they were assigned at randomization.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With Glycosylated Haemoglobin (HbA1c) Less Than 7.0 Percentage (%) (53 Millimoles Per Mole [mmol/Mol]) at Year 1 (Yes/No), measured over At year 1.

The primary endpoint counts participants whose HbA1c was below 7.0% (53 mmol/mol) at year 1. It is recorded as a binary outcome: participants either achieved the stated threshold or did not. The registry says that the number achieving the threshold is presented, and defines the complementary category as participants who did not achieve it.

A threshold endpoint can be useful because it translates a laboratory measure into an easily interpreted target-attainment outcome. It asks how many participants crossed the chosen cutoff, rather than how much HbA1c changed on average. That distinction is important: a binary classification does not retain all the information in the underlying continuous measurement. Participants just on either side of a cutoff are placed in different categories, while participants with different values on the same side are grouped together.

The endpoint is measured at year 1, so it describes target attainment at that specified time point. It does not by itself describe the path of HbA1c before that assessment, whether the threshold was reached earlier, or whether it was maintained afterward. The registry lists HbA1c change and other target-attainment measures as secondary outcomes, which address related but distinct aspects of glycemic control.

Although the endpoint definition is reported, the number meeting the threshold in either group and the corresponding comparison are not reported in the registry record. Its clinical and statistical interpretation therefore remains limited to the prespecified meaning of the measure, rather than an observed between-group result.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Randomization and treatment comparison. The registry records randomized allocation to semaglutide or standard of care in a parallel-group design. In general, randomization is used to make the assigned groups comparable on average with respect to both measured and unmeasured characteristics at the start of a trial. It does not guarantee that the groups will be identical, nor does it prevent treatment changes after assignment. Here, participants were followed for 2 years regardless of changes in antidiabetic treatment, so the relevant comparison is between the assigned strategies in the context of those changes. The statistical estimand and the precise treatment of changes in the analysis are not reported in the registry record.

Binary endpoint. The primary outcome is whether HbA1c was below 7.0% (53 mmol/mol) at year 1. A binary outcome has two categories, achievement and non-achievement of the defined threshold. General statistical practice is to report the number of participants in each category, the number with outcome data, and a between-group estimate with an uncertainty interval. The registry defines the outcome and analysis population, but does not report the primary endpoint counts or a comparative estimate. No particular statistical model or test for this endpoint is reported in the registry record.

Analysis population and observed data. The registry describes the full analysis set as including all randomized participants analyzed according to their assigned group. It also states that the overall number analyzed for the outcome is the number of participants with available data for that outcome measure. These statements distinguish randomized assignment from outcome availability: an assigned participant may not contribute an observed value for a particular measure if that value is unavailable. The registry record does not give the number with available year-one HbA1c data by group, explain why data might be unavailable, or report an imputation or other missing-data procedure. It is therefore not possible to assess from this record how missing primary-endpoint observations were handled.

Precision and uncertainty. In general, a point estimate gives a single summary of a group difference, while a confidence interval communicates uncertainty around that estimate under the chosen method. A narrower interval indicates greater statistical precision, all else being equal, but does not establish that an estimate is important in clinical terms. The registry record does not report a primary-endpoint estimate or confidence interval, so neither the direction nor precision of the group comparison can be assessed. It also does not report the statistical test, model, covariate adjustment, or significance threshold used for the primary comparison.

Planning illustrations. The posted planning illustrations are described as using the observed effect, a two-sided significance level of 0.05, equal allocation, and a normal approximation, with the effect to detect set to a stated fraction of the observed effect. The record describes these as planning illustrations, not as the trial’s original design justification or as observed power. The illustration values and the trial’s own sample-size justification are not reported in the registry record. A planning calculation based on an observed effect is conditional on that effect and its assumptions; it does not establish the probability that the completed trial detected a true effect.

Dispersion and count outcomes. General methods for count outcomes may need to allow for over-dispersion when variability exceeds what a simple count model expects. That issue is not established for this primary endpoint, which is binary rather than a count of recurrent events. The registry does not report an over-dispersion assessment or model for the primary endpoint. The adverse-event counts are reported descriptively, and the record does not specify an inferential comparison for them.

What is documented. The registry reports 69 outcome measures and 1 analysis, but those figures alone do not identify the statistical method used for the primary endpoint or establish a multiplicity adjustment. The primary endpoint’s actual group results and the analysis details needed to reproduce or interpret a comparative estimate are not reported in the registry record. Accordingly, statistical explanations about binary outcomes, uncertainty, and missing observations are general descriptions, not claims about methods that the trial actually used.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 69 outcome measures, including the primary year-one HbA1c threshold, several other HbA1c target or change measures, and body-weight outcomes. When a study evaluates many outcomes or comparisons, the chance of at least one apparently unusual result can increase if each is interpreted as though it were the only analysis. General approaches to this issue include defining a hierarchy of outcomes or adjusting inference across a set of comparisons. Which approach, if any, was used for this trial is not reported in the registry record.

The registry also reports 1 analysis. That count does not, on its own, establish which outcomes or comparisons were included in that analysis, whether outcomes were examined in multiple ways, or whether any multiplicity procedure was used. The statistical analysis plan and multiplicity strategy are not reported in the registry record. Thus, the listed number of outcomes should be considered when interpreting any collection of results, but it does not permit a specific adjustment to be inferred.

For the primary endpoint, the registry describes a full analysis set based on randomized assignment and available outcome data. The number with available data is not reported. The registry does not explain the extent or reasons for missing year-one HbA1c observations, or describe imputation, sensitivity analyses, or another method for addressing missing data. In general, missingness can affect a comparison if availability of outcome data is related to treatment assignment or to the outcome itself. Without group-specific availability and the analysis approach, the potential effect of missing data cannot be assessed.

The record also does not report whether an interim analysis was planned or performed. No interim-monitoring procedure should be inferred from the completed status alone. Overall, the outcome list and analysis-population definition are documented, but the multiplicity and missing-data procedures needed to judge the full inferential context are not reported in the registry record.

11. 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

Serious events with the highest arm-level rate: coronavirus infection (0.8%), acute respiratory failure (0.8%), death (0.5%), pneumonia (0.3%), acute myocardial infarction (0.3%).

The registry lists 634 participants at risk for serious adverse events in the semaglutide group and 621 in the standard-of-care group. It lists 38 and 39 participants, respectively, as affected by serious adverse events. For the other-adverse-event fields, the affected counts are recorded as 0 in each group, with the same at-risk group counts. The registry lists 9 deaths in each group.

These are descriptive registry figures. They do not by themselves establish whether an event was caused by a study treatment, whether the groups differ statistically, or how the events relate to exposure duration. The record does not report an inferential comparison for these safety counts. The number at risk is the denominator listed for each group; it should not be confused with the number of events or with an estimate of treatment effect.

The registry lists serious-event terms with percentages: Coronavirus infection, 0.8%; Acute respiratory failure, 0.8%; Death, 0.5%; Pneumonia, 0.3%; Acute myocardial infarction, 0.3%; and Coronary artery disease, 0.3%. The record does not specify the group attribution or denominator for these term percentages. The listed term Death should also be read separately from the group-level death counts; the registry does not provide enough detail to reconcile the term percentage with those counts.

Safety counts are not a substitute for the primary HbA1c comparison. They describe reported adverse-event categories, while the primary endpoint concerns target attainment at year 1.

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

The registry record establishes the purpose and broad structure of the trial: randomized assignment to semaglutide or standard of care, both as treatment intensification with metformin monotherapy in routine practice, and a primary endpoint based on HbA1c <7.0% at year 1. It also reports enrollment, treatment descriptions, follow-up duration, and selected safety counts. These registry entries describe what was studied and how the primary endpoint was defined.

The record does not report the primary outcome counts by treatment group, a group difference, a confidence interval, or a p-value. It therefore does not show whether target attainment was more or less frequent in either assigned group, how large a difference might have been, or how precisely a difference was estimated. A defined outcome is not itself an outcome result. Likewise, the trial’s completed status does not supply the missing comparative estimate.

Even if endpoint results were available, a binary threshold would answer a specific question about target attainment at year 1. It would not capture every feature of HbA1c values, such as variation within the categories, or describe the full course of HbA1c over follow-up. The listed secondary outcomes address other measures, but their results and relationship to the primary analysis are not reported in the registry record.

The routine-care design and treatment changes during follow-up also shape the question. The comparison concerns groups assigned to treatment strategies in a pragmatic setting, with follow-up continuing regardless of changes in antidiabetic treatment. It should not be read as evidence that all participants remained on their initial therapy. The registry does not report the analysis details needed to determine precisely how such changes informed the endpoint comparison.

The posted information can support a description of the trial design and the meaning of the endpoint. It cannot support a quantitative claim about comparative efficacy for the primary outcome or a conclusion about the clinical importance of any difference. The safety figures should likewise be read as registry-reported counts and terms, not as evidence of causation or as a statistical comparison between treatments.

13. Why This Trial Matters Statistically

This trial is statistically interesting because it combines random allocation with a pragmatic, routine-practice setting. Randomization creates a basis for comparing assigned groups, while routine-care treatment and continued follow-up despite treatment changes make the interpretation depend on how the assigned strategy is defined and analyzed. That combination is relevant to how clinical trials can ask comparative questions under practice conditions, but the registry record does not report the estimand or detailed methods needed to characterize the analysis fully.

The primary outcome also illustrates a common design choice: a continuous laboratory measure is converted into a binary target-attainment endpoint. This can make results straightforward to communicate as counts meeting or not meeting a threshold. The tradeoff, in general, is loss of information about the range of values within each category. The record lists continuous HbA1c change and other target outcomes among secondary measures, which reflects that related outcomes can answer different statistical questions.

Another important statistical issue is outcome availability. The registry defines the analysis set by randomized group but says that the analyzed number for an outcome consists of participants with available data. Without the group-specific number observed and the missing-data method, readers cannot judge how much of the randomized population contributed to the primary endpoint or assess the sensitivity of the comparison to missing observations.

The registry lists 69 outcome measures and 1 analysis, making the absence of reported multiplicity procedures relevant to interpretation. A count of outcomes is not enough to determine how results were selected or adjusted, and the record does not report that strategy. Most importantly, the primary result itself is not reported in the registry record. The statistical significance of the trial’s question lies in the design and measurement issues it raises; the posted record does not permit a numerical judgment of the primary comparison.

14. 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 central limitation for interpreting this trial from the registry record is that the primary endpoint is defined but its arm-specific results are not reported. There is no reported count meeting the HbA1c threshold in either group, no comparative estimate, and no confidence interval or p-value. The primary comparison’s direction, magnitude, and precision therefore cannot be assessed.

The registry describes the analysis population as participants analyzed according to randomized assignment, with the analyzed number for the outcome restricted to participants with available data. It does not report how many participants had year-one HbA1c data in each group, the reasons for any missing observations, or how missing data were handled. This leaves uncertainty about the observed endpoint population and the possible influence of missingness.

The trial is recorded as open-label. That design feature may be relevant when considering how care and outcome assessment occurred, but the registry does not report outcome-assessment procedures or whether the HbA1c assessment was conducted in a way that could limit knowledge of assigned treatment. The routine-care descriptions also allow antidiabetic treatment changes during the 2-year follow-up. Details about adherence, actual treatment received, and the analysis of changes are not reported in the registry record.

The registry lists 69 outcome measures, but does not report the multiplicity strategy or detailed statistical analysis plan. It also does not report the trial’s own sample-size justification or interim-analysis procedures. The serious adverse-event figures are descriptive, and the record does not report a statistical comparison or establish causation. These limitations concern what can be concluded from the registry record; they do not establish that a particular method was or was not used outside the information recorded there.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Long Term Comparative Effectiveness of Once Weekly Semaglutide Versus Standard of Care in a Real World Adult US Population With Type 2 Diabetes - a Randomized Pragmatic Trial — NCT03596450.
• Registry record last updated 2026-09-30; status last verified 2026-09.
• Results first posted 2023-08-15.
• 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.

18. Record Summary

NCT03596450 is a completed phase 4, randomized, open-label, parallel-group trial with enrollment of 1278. It compared semaglutide with standard of care as treatment intensification added to metformin monotherapy in routine clinical practice. Participants were followed for 2 years regardless of changes in antidiabetic treatment.

The primary endpoint was the number of participants with HbA1c <7.0% (53 mmol/mol) at year 1. The registry describes a full analysis set based on randomized assignment and available outcome data, but does not report the primary endpoint counts by group, a comparative effect estimate, or its uncertainty. It lists 69 outcome measures and 1 analysis; multiplicity and missing-data procedures are not reported in the registry record.

Serious adverse-event counts are listed as 38 in the semaglutide group and 39 in the standard-of-care group, with 9 deaths in each group. These figures are descriptive and do not establish causation or a treatment comparison. The registry record supports a summary of the design and endpoint definition, but not a quantitative interpretation of the primary result.