Switching From Insulin Glargine to Insulin Degludec in Subjects With Type 2 Diabetes Mellitus (BEGIN™): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Switching From Insulin Glargine to Insulin Degludec in Subjects With Type 2 Diabetes Mellitus (BEGIN™) (NCT01135992), a non-randomized phase 3 study. The primary outcome is HbA1c (Glycosylated Haemoglobin).

2. Clinical Question

The stated aim is to assess the implications of switching from insulin glargine (IGlar) to insulin degludec (IDeg) in subjects with type 2 diabetes mellitus. The primary outcome is glycosylated haemoglobin, or HbA1c, expressed as a percentage of glycosylated haemoglobin. The registry specifies the mean and standard deviation at Week 4 and Week 16. Those definitions identify what was to be described, but do not provide an estimate of the observed mean, its variability, or a change over time.

The trial also lists fasting plasma glucose, change in body weight, treatment-emergent adverse events, confirmed hypoglycaemic episodes, and nocturnal confirmed hypoglycaemic episodes as secondary outcomes. Their names indicate the range of clinical measures considered, but numerical outcome results are not reported in the registry record. The arm labels are IGlar/IDeg and IDeg 3TW; their descriptions are blank, so their intended treatment sequences or roles cannot be reconstructed from the record.

Accordingly, the question that can be answered from the posted record is what outcome the trial set out to assess and how it was designed, not whether the observed HbA1c changed or how large any change was. The stated design also limits interpretation of any apparent contrast between treatment labels: allocation was non-randomized and the model was single-group, rather than a randomized comparison.

3. Trial Design

This was a phase 3 trial with a non-randomized allocation, no masking, and a single-group model. The registry reports enrollment of 143. It describes the trial as conducted in the United States and identifies type 2 diabetes mellitus as the condition under study. The record lists insulin degludec and insulin glargine as drug interventions, and gives the overall aim as assessing the implications of switching from glargine to degludec.

The listed start was 2010-06, primary completion was 2010-11, and status is completed. These dates describe the registry record’s trial timeline; they do not provide information about the duration of treatment for an individual participant. The primary outcome is HbA1c, reported as a mean with standard deviation at Week 4 and Week 16. Its specified analysis population is the Full Analysis Set, defined in the record as all subjects switched to IDeg 3TW treatment.

Two arm labels appear: IGlar/IDeg and IDeg 3TW. The registry does not supply descriptions for either arm. The non-randomized allocation and single-group model should guide interpretation rather than the presence of two labels: the record does not establish a randomized, balanced comparison between these labels. It also does not explain how participants came to receive a treatment, whether the labels refer to sequential treatment periods, or how any within-participant comparison was to be conducted. Those details are not reported in the registry record.

In general, random allocation helps make treatment groups comparable by balancing measured and unmeasured characteristics on average. Without randomization, observed differences can reflect the treatments, differences between participants, timing, or other factors. A single-group design can describe outcomes in the group studied, but by itself does not isolate a treatment effect from other explanations. The registry does not report a protocol, detailed analysis plan, or numeric primary-outcome results. The design therefore supports a cautious account of the study’s aim and outcome definitions, not a causal conclusion about switching.

4. Participant Flow and Accounting

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

Before assignment: The trial was conducted on subjects with type 2 diabetes mellitus currently treated with IGlar once daily (OD) and oral antidiabetic drug (OAD) therapy.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: FAS (Full Analysis Set) includes all subjects that were switched to IDeg 3TW treatment

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)-Includes 129 subjects only as 13 subjects did not receive IDeg.

6. Primary Endpoint

The registry lists the primary outcome as HbA1c (Glycosylated Haemoglobin), measured over Week 4 and Week 16.

The primary endpoint is HbA1c, defined in the registry as glycosylated haemoglobin and measured as a percentage of glycosylated haemoglobin. The record specifies a mean and standard deviation at Week 4 and Week 16. The primary-outcome population is the Full Analysis Set, which the record defines as all subjects switched to IDeg 3TW treatment.

In general, a continuous endpoint such as HbA1c can be described by its mean, while the standard deviation describes variation among individual measurements around that mean. Reporting both can show the average observed level and how dispersed participant values are. Measurements at stated time points can also help readers understand when outcomes were intended to be assessed. These are general statistical roles of the summaries; they do not establish what happened in this trial.

The registry record does not report the mean or standard deviation at either time point, nor does it give a change estimate, confidence interval, or test result for HbA1c. It is therefore not possible to quantify the endpoint from the record. The definition and timing identify the planned outcome, but are not themselves evidence of an outcome difference or change.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a mean and standard deviation describe?

The registry defines the primary endpoint as a continuous measure and specifies that it be summarized by a mean and standard deviation. In general, the mean is a measure of the center of observed values. The standard deviation describes how spread out individual values are around that center. They answer different descriptive questions: the mean concerns the average level, while the standard deviation concerns participant-to-participant variability. Neither alone gives the precision of an estimated mean. Precision also depends on the amount and structure of the available data, which are not fully described for the primary outcome in the registry record.

The endpoint is specified at Week 4 and Week 16. Separate summaries at time points can describe observed levels, but a comparison of those summaries is not automatically the same as an analysis of within-participant change. That distinction depends on the analysis method and the availability of paired measurements. The registry record does not describe the model or calculation used to compare time points, and it does not report the numerical summaries.

What does the design imply for statistical interpretation?

The design is identified as non-randomized and single-group, with no masking. In general, randomization is a design method intended to reduce systematic differences between treatment groups. Non-randomized allocation does not provide that protection by design. Differences observed between treatment labels could be connected to participant characteristics or other factors as well as treatment. A single-group model also means the record does not establish a randomized control group against which to estimate a treatment contrast. Since the arm descriptions are blank, the exact relationship between the listed labels cannot be clarified from the registry record.

Masking is listed as none. In general, masking can reduce the influence of treatment knowledge on aspects of trial conduct or assessment, depending on the outcome and study procedures. The record does not explain which parties knew treatment assignments, how outcome measurements were collected, or what steps were used to limit measurement or reporting bias. These specifics are not reported in the registry record.

How should uncertainty be understood?

In general, an estimate is more useful when paired with a measure of uncertainty. A confidence interval is a range calculated using a statistical method that indicates values reasonably compatible with the observed data under the model’s assumptions. A narrow interval indicates greater statistical precision than a wide interval, although precision is not the same as freedom from bias. The registry record does not report a confidence interval or primary-outcome estimate, so neither the magnitude nor precision of the HbA1c result can be evaluated.

A p-value, in general, assesses how compatible data are with a specified null hypothesis under a chosen statistical model. It does not measure the size or clinical importance of an effect. No primary-outcome p-value or test is reported in the registry record. The significance threshold and detailed statistical plan for the trial are also not reported. The 0.05 significance level appears only in the assumptions for planning illustrations, not as a reported description of the trial’s own analysis plan.

What do the planning illustrations represent?

The record characterizes the sample-size calculations as planning illustrations computed from an observed effect, rather than as the trial’s own design. Their stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a stated fraction of the observed effect. In general, such calculations explore the number of participants that would be needed under particular assumptions to detect a specified effect with a chosen power. They are not observed outcomes and are not a measure of the trial’s achieved power. The record says that non-time-to-event calculations are expressed as participants per arm, while time-to-event calculations give a total number of events using the Schoenfeld method. The primary endpoint is continuous, and no trial-specific sample-size justification is reported in the registry record.

Because the design is listed as single-group and non-randomized, a planning illustration that assumes equal allocation between two arms should not be mistaken for a description of the observed trial design. Its assumptions are hypothetical inputs for a calculation. The record does not report a protocol-based sample-size justification, the target effect used in original planning, or any interim analysis plan. Those details are not reported in the registry record.

How should dispersion and outcome types be distinguished?

Standard deviation is a descriptive measure of dispersion for a continuous outcome. In general, greater dispersion means individual observations vary more around the mean. It is not interchangeable with a standard error or confidence interval: those concern uncertainty in an estimate rather than the spread of individual measurements. The registry specifies standard deviation for HbA1c, but does not give its value. It also lists event-rate outcomes, such as adverse events and hypoglycaemic episodes. In general, event counts or rates require denominators and a clear definition of the event and observation period. The registry’s listed safety denominators and affected counts provide some descriptive information, but no detailed analytic method or exposure-time information is reported.

Over-dispersion is a general concern when count data vary more than a simple statistical model expects; ignoring it can make uncertainty estimates too optimistic. Whether that issue applies depends on the data and model. The registry record does not describe a count-data model or report enough outcome results to assess over-dispersion for the listed event outcomes.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists HbA1c as the primary outcome and names several secondary outcomes: fasting plasma glucose, change in body weight, treatment-emergent adverse events, confirmed hypoglycaemic episodes, and nocturnal confirmed hypoglycaemic episodes. The primary HbA1c outcome has two stated time points, Week 4 and Week 16. Considering multiple outcomes or time points can raise a multiplicity issue: in general, the more statistical tests conducted, the greater the chance of at least one apparently notable result arising by chance if no adjustment or hierarchy is used. Whether an adjustment was planned, which outcomes were confirmatory, and how time points were handled are not reported in the registry record.

The record lists zero analyses. It does not report outcome estimates, test results, or an analysis plan for the endpoints. It is therefore not possible to determine whether hypotheses were prespecified in a protocol, whether analyses were adjusted for multiplicity, or whether any analyses were exploratory. The significance level of 0.05 appears among assumptions for planning illustrations; that does not establish the threshold used for trial outcome analyses.

The primary analysis population is described as the Full Analysis Set, defined as all subjects switched to IDeg 3TW treatment. This identifies the population named for HbA1c, but the registry does not say how missing outcome measurements were handled, whether participants were excluded after enrollment, or whether sensitivity analyses were performed. The number contributing data at each time point is not reported in the registry record. In general, missing data can affect both the amount of information available and the interpretation of an estimate, particularly if the reasons for missingness relate to participant characteristics or outcomes. Without a reported approach and outcome-level denominators, that issue cannot be evaluated for this trial.

The registry also does not specify other analysis populations, such as a safety population, or define how participants were included in analyses of secondary outcomes. These details are not reported in the registry record. The stated population definition should not be expanded into assumptions about all enrolled participants or about how every endpoint was analyzed.

11. 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

Serious events with the highest arm-level rate: cholelithiaisis (ideg) (0.8%), carotid artery occlusion (ideg) (0.8%), cholelithiasis (total) (0.7%), lumbar spinal stenosis (iglar) (0.7%), lumbar spinal stenosis (total) (0.7%).

The registry’s serious-adverse-event fields list the IGlar/IDeg group with 142 at risk and 3 affected. For the other adverse-event category, the fields list 142 at risk and 0 affected. These are the counts recorded in the registry fields; they should be read as descriptive safety entries, not as a randomized comparison. The design is identified as non-randomized and single-group, and the arm descriptions are blank, so a direct comparative safety interpretation is not supported by the record.

The listed serious terms include “Cholelithiaisis (IDeg)” at 0.8%, “Carotid artery occlusion (IDeg)” at 0.8%, “Cholelithiasis (Total)” at 0.7%, “Lumbar spinal stenosis (IGlar)” at 0.7%, “Lumbar spinal stenosis (Total)” at 0.7%, and “Carotid artery occlusion (Total)” at 0.7%. The registry lists these terms and percentages but does not provide enough detail to resolve their relationship to the group-level affected counts or to infer comparative risk. The death field is not reported in the registry record.

In general, safety counts need context, including how events and affected participants are defined, the observation period, and the number exposed. The listed at-risk counts offer denominators for the recorded fields, but the registry does not report detailed exposure time or a comparative analysis. These entries should not be read as proof that an event was caused by a treatment or that an unlisted event did not occur.

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

The record establishes that NCT01135992 was a completed phase 3 trial intended to assess implications of switching from insulin glargine to insulin degludec in subjects with type 2 diabetes mellitus. It specifies HbA1c as the primary outcome, measured at Week 4 and Week 16 and summarized by mean and standard deviation. It also identifies the Full Analysis Set as subjects switched to IDeg 3TW treatment. These details describe the study question and endpoint definition.

The record does not report the numerical HbA1c results, the secondary-outcome results, or primary-outcome estimates of uncertainty. It therefore does not show the size or direction of any observed HbA1c difference or change, nor how precisely any such estimate was measured. The endpoint name and the fact that the trial is completed cannot stand in for results.

The design further limits the interpretation of any possible contrast: allocation is non-randomized, the model is single-group, and the arm descriptions are empty. In general, a design without random assignment cannot rule out differences in participant characteristics or other influences as explanations for observed outcomes. The record does not establish a like-for-like control comparison. No causal conclusion about the effect of switching can be drawn from the information reported in the registry.

The serious-adverse-event fields and listed serious terms provide limited descriptive safety information. They do not establish comparative risks or causation, and the death field is not reported. Secondary outcomes are named but not accompanied by numerical findings. Taken together, the registry record supports a description of the trial’s stated aim, design, outcome definitions, and selected safety entries. It does not support a conclusion about efficacy, the size or precision of an effect, or a comparative benefit-risk judgment.

13. Why This Trial Matters Statistically

This trial is statistically informative as an example of why the design label and the outcome results must be read together. The registry names two insulin treatments and two arm labels, but identifies allocation as non-randomized and the model as single-group. Those design features matter because they determine what kinds of comparisons can be interpreted. A treatment label alone does not demonstrate that groups were formed to support a causal contrast, especially when the arm descriptions are empty.

The primary endpoint is clearly defined as mean HbA1c with standard deviation at Week 4 and Week 16 in a stated Full Analysis Set. Clear outcome definitions help readers understand what investigators intended to summarize. But a definition is distinct from a result: no numerical HbA1c findings are reported in the registry record. This makes the distinction between a registered question and an answer particularly important. Readers cannot assess the observed magnitude, variability, or precision without the outcome estimates and denominators.

The record also illustrates the limits of interpreting safety entries in isolation. Counts, at-risk figures, and listed event terms offer descriptive information, but do not alone establish that a treatment caused an event or that one treatment has a higher risk than another. Study design, exposure, event definitions, and analysis methods all inform such comparisons; several of those details are not reported in the registry record.

Finally, the planning assumptions should not be confused with the trial’s own design or observed statistical performance. The registry explicitly describes the sample-size calculations as planning illustrations. General statistical practice distinguishes a hypothetical sample-size calculation from an analysis of collected data. Keeping those distinctions clear prevents a planning estimate from being interpreted as observed power or as evidence that a result was present.

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 interpretation is that the registry record does not report numerical primary or secondary outcome results. Although HbA1c is defined at Week 4 and Week 16, no means, standard deviations, changes, confidence intervals, or test results are given. The record therefore does not support an assessment of effect magnitude or precision.

The design is recorded as non-randomized and single-group, with no masking. The arm labels are present, but their descriptions are empty. The record does not explain treatment sequence, allocation procedures, or whether the labels represent any comparable treatment contrast. This prevents a clear reconstruction of how the listed interventions relate to the primary endpoint population.

Other trial details needed to assess potential bias and analysis choices are not reported in the registry record. These include a detailed statistical analysis plan, missing-data methods, outcome-level denominators, interim analyses, and the original sample-size justification. The record defines the Full Analysis Set for the primary outcome but does not state how missing measurements or exclusions were handled. It also does not report whether multiple outcomes or time points were handled using a prespecified hierarchy or adjustment.

The safety information is limited to selected fields and listed serious terms. The death field is not reported, and the registry does not provide detailed exposure time or a comparative safety analysis. The available entries should therefore be treated as descriptive rather than comprehensive or causal evidence. These limitations arise from what the registry record reports; they do not establish that additional information was absent from other trial documents.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Switching From Insulin Glargine to Insulin Degludec in Subjects With Type 2 Diabetes Mellitus (BEGIN™) — NCT01135992.
• Registry record last updated 2016-01-22; status last verified 2015-12.
• Results first posted 2016-01-22.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

18. Record Summary

NCT01135992, titled “Switching From Insulin Glargine to Insulin Degludec in Subjects With Type 2 Diabetes Mellitus (BEGIN™),” is a completed phase 3 trial conducted in the United States. The registry lists enrollment of 143, non-randomized allocation, no masking, and a single-group model. The stated aim was to assess the implications of switching from insulin glargine to insulin degludec in subjects with type 2 diabetes mellitus.

The primary endpoint is HbA1c, expressed as a percentage of glycosylated haemoglobin and summarized by mean and standard deviation at Week 4 and Week 16. The specified Full Analysis Set includes all subjects switched to IDeg 3TW treatment. Secondary outcomes include fasting plasma glucose, body weight, adverse events, and hypoglycaemic episodes.

The registry record does not report numerical efficacy results or a detailed analysis plan. Serious-adverse-event fields list 3 affected among 142 at risk in the IGlar/IDeg group and 0 affected among 142 at risk for the other adverse-event category. The death field is not reported in the registry record.