Comparison of Biphasic Insulin Aspart 30 Versus Insulin Glargine Both in Combination With Metformin and Glimepiride in Subjects With Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Comparison of Biphasic Insulin Aspart 30 Versus Insulin Glargine Both in Combination With Metformin and Glimepiride in Subjects With Type 2 Diabetes (NCT00469092), a randomized phase 4 study. The primary outcome is Glycosylated Haemoglobin A1c (HbA1c).

2. Clinical Question

The registry record describes a comparison of two insulin regimens on a background of metformin and glimepiride. The treatment contrast is therefore between the once-daily insulin strategies, rather than between insulin treatment and no insulin treatment. Both groups received the same two named background medicines according to the trial title and summary.

The primary outcome was glycosylated haemoglobin A1c (HbA1c), measured in blood samples after 26 weeks of treatment. The posted group estimates are least-squares means: 7.08 for biphasic insulin aspart 30 and 7.23 for insulin glargine. The registry record also reports a treatment comparison as a mean difference. These values describe the posted endpoint and analysis; by themselves, they do not establish how the comparison should guide an individual treatment decision.

The registry record lists additional outcomes concerning self-measured plasma glucose profiles, achievement of an HbA1c target, treatment satisfaction, hypoglycaemic episodes and treatment-emergent adverse events. Their presence indicates that the trial assessed more than the primary laboratory measure. Details of the specific results for those secondary outcomes are not reported in the registry record presented for this analysis.

3. Trial Design

This was a completed phase 4, randomized, parallel-group trial with no masking. Its official title describes a multinational study with a 4 week run-in and a 26 weeks treat-to-target treatment period. The trial was conducted in Africa, Asia, Europe, Oceania and South America. The registry record lists enrollment of 480 subjects.

Participants were assigned to one of two once-daily insulin regimens: biphasic insulin aspart 30 or insulin glargine. Both regimens were used with metformin and glimepiride. The trial title identifies the participants as insulin-naive subjects with type 2 diabetes. Because masking was listed as none, the trial was open-label: participants and relevant trial personnel were not blinded to the assigned insulin regimen, as far as the registry description indicates.

The parallel-group structure means the treatment groups were followed as separate groups rather than participants crossing from one regimen to the other. Random allocation is intended to make the groups comparable at the start of the trial on average, although randomization does not guarantee that every measured or unmeasured characteristic will be balanced. The registry record does not report the allocation ratio, the sequence-generation method, or whether allocation concealment was used.

The primary outcome was HbA1c after 26 weeks of treatment. The primary analysis population was an intention-to-treat population with a specific inclusion definition: all randomized subjects exposed to trial drug, with at least one baseline HbA1c measurement and at least one post-randomization HbA1c measurement. Missing follow-up values were handled using last observation carried forward for this population. Further details about the visit schedule, dose-adjustment algorithm, treatment adherence, or handling of participants who discontinued treatment are not reported in the registry record.

The registry record lists five secondary outcome topics: 9-point self-measured plasma glucose profiles; the number of subjects reaching an HbA1c treatment target; treatment satisfaction measured with the Diabetes Medication Satisfaction Questionnaire; the number of hypoglycaemic episodes; and the number of subjects reporting treatment-emergent adverse events. Results for these outcomes are not described in the information available for this analysis. The registry record also does not report the trial's original sample-size justification or other details of its planned statistical analysis.

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 screening period of 1-2 weeks was followed by a run-in period of 4 weeks during which metformin was titrated to maximum 2550 mg and glimepiride to 4 mg, at the discretion of the investigator. Subjects who were already taking 4, 6 or 8 mg glimepiride continued on this dose. Doses were kept constant during the last week prior to randomisation.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intention to Treat (Last Observation Carried Forward) population. All randomised subjects exposed to trial drug, and who had at least a baseline HbA1c measurement and at least one post randomisation HbA1c measurement.

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 Glycosylated Haemoglobin A1c (HbA1c), measured over After 26 weeks of treatment.

HbA1c is a measure of glycosylated haemoglobin in blood, reported in this trial as a percentage of total haemoglobin. It reflects glucose exposure over time rather than a single glucose reading. As a general statistical and clinical concept, this makes HbA1c useful for comparing longer-term glycaemic status between treatment groups at a planned follow-up point.

The registry record defines the primary endpoint as HbA1c measured after 26 weeks of treatment. It describes the group summary as a least-squares mean and gives the outcome unit as percentage of total haemoglobin. A least-squares mean is a model-based group estimate; its interpretation depends on the statistical model and any adjustments used. The full model specification and covariates are not reported in the registry record.

The posted least-squares mean values are 7.08 for biphasic insulin aspart 30 and 7.23 for insulin glargine. The registered primary analysis population included randomized subjects exposed to trial drug who had a baseline and at least one post-randomization HbA1c measurement, with last observation carried forward used for missing data. These choices describe how the posted primary result was defined; they do not show how every randomized participant's outcome would have appeared under complete follow-up.

HbA1c is one outcome among several listed in the registry record. It does not, on its own, summarize glucose patterns, treatment satisfaction, hypoglycaemic episodes or other adverse events. Those other outcomes are relevant to understanding the wider comparison, but their results are not reported in the registry information.

7. Results Posted in the Registry

• Glargine: mean 7.23 versus 7.08 on control; difference 0.15 (95% CI 0.14 to 0.16), standardized effect d = 2.14; z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is continuous: HbA1c is measured on a numeric scale, and the registry record reports group means. The posted values are least-squares means, which are estimated from a statistical model rather than necessarily being simple arithmetic averages. General explanation: a model-based mean can account for factors included in the model, but the estimate's meaning depends on the model specification. Covariates and other modeling details are not reported in the registry record.

What does a mean difference represent?

A mean difference expresses the contrast between two group-level estimates in the outcome's units. For the unadjusted recalculation, the comparison is ordered as insulin glargine minus biphasic insulin aspart 30; the estimate is 0.15 percentage points of total haemoglobin. A positive value under that ordering means the posted HbA1c estimate for insulin glargine is higher than the estimate for biphasic insulin aspart 30. Reversing the group order reverses the sign without changing the underlying separation between the estimates.

The recalculated comparison uses the posted group summaries and is not adjusted for additional variables. The registry's sponsor analysis instead reports linear regression and a net mean difference of -0.16. Its group order is opposite to the recalculation's order. Although sign convention explains the opposing directions, the registry information does not establish why the numerical estimates differ. The specific regression model, adjustment variables and other analysis choices are not reported in the registry record.

What does a confidence interval convey?

General explanation: a confidence interval is a range of values compatible with the observed data under a statistical model and its assumptions. It represents uncertainty in an estimated effect; a narrower interval indicates greater precision under that model, while a wider interval indicates less precision. It does not mean that the true value has a particular probability of falling within the observed interval after the data have been collected.

The sponsor analysis reports a 95% confidence interval from -0.30 to -0.02 for its net mean difference. This interval belongs to the sponsor's regression analysis and uses its own contrast convention. The recalculation is based on the posted summary statistics and is a separate unadjusted analysis. Intervals from analyses using different methods should not be combined or interpreted as if they were generated by the same model.

How should the p-values be read?

General explanation: a p-value quantifies how surprising data at least as extreme as those observed would be if a particular null hypothesis and statistical model were correct. Smaller values can indicate stronger evidence against that null model, but a p-value does not measure the size or importance of an effect. Its interpretation depends on the analysis that generated it and on the assumptions behind that analysis.

The recalculated unadjusted comparison has p < 0.001. The registry's sponsor analysis has p = 0.029. They arise from different analyses and should be attributed to those analyses rather than treated as competing versions of a single identical test. The registry record describes the sponsor result as a non-inferiority or equivalence analysis, but the margin, hypotheses and decision criteria are not reported in the registry record. Without those details, the p-value cannot by itself establish how the formal criterion was assessed.

Why can variability affect a comparison?

General explanation: statistical precision depends on both the amount of information and the variability in the outcome. When measurements vary more within groups, it can be harder to distinguish a group contrast from random variation. A standard deviation describes spread among observations; a standard error describes uncertainty in an estimated statistic. They are related but do not mean the same thing.

The registry identifies the primary outcome dispersion as standard deviation and lists a spread of 0.07 for each arm. The recalculation uses an approximate standard deviation, as noted in its calculation details. The summary information does not give participant-level data, so it cannot be used to examine the distribution, outliers or subgroup patterns. Nor does it identify all assumptions behind the approximate variability calculation.

What does randomization contribute?

General explanation: random allocation is designed to reduce systematic differences in baseline characteristics between treatment groups. It supports a fair comparison, particularly when assignment and follow-up are carried out as planned. It cannot prevent bias caused by missing outcomes, deviations from assigned treatment, unblinded assessment or selective reporting. The trial was randomized but open-label, and the registry record does not describe allocation concealment or the procedures used to limit other forms of bias.

The primary analysis population excluded randomized subjects who were not exposed to trial drug or who lacked the required HbA1c measurements. This eligibility definition should be considered alongside the label intention to treat. The record explicitly defines the population and the use of last observation carried forward; it does not report how many randomized participants failed to meet the analysis criteria or how many measurements were missing.

What is not specified?

The original sample-size justification, planned statistical analysis details, interim analyses, multiplicity strategy, non-inferiority margin and full model specification are not reported in the registry record. These items are important because they define how the investigators intended to quantify uncertainty, manage multiple outcomes, and interpret a non-inferiority or equivalence result. Their absence limits reconstruction of the prespecified analysis from the registry information alone.

General explanation: over-dispersion occurs when observed variability exceeds what a chosen statistical model expects. In count outcomes, for example, treating the data as less variable than they are can make standard errors too small and confidence intervals too narrow. The registry lists outcomes involving event counts, but the statistical models for those outcomes are not reported in the registry record. It is therefore not possible to determine from the available description whether over-dispersion was assessed or relevant to the analyses.

10. Confidence Intervals and Precision

The recalculated unadjusted mean difference is 0.15, with insulin glargine minus biphasic insulin aspart 30 as the comparison order. The reported group estimates are 7.23 and 7.08, respectively. The recalculation indicates that the group-level contrast is estimated with high statistical precision under its unadjusted assumptions, reflected by p < 0.001. This statement concerns statistical uncertainty in that calculation, not the certainty of a clinical conclusion.

The sponsor's regression analysis reports a 95% confidence interval from -0.30 to -0.02 for a net mean difference of -0.16. The interval is expressed using the sponsor analysis's contrast convention, opposite to the order used for the unadjusted recalculation. It is the interval associated with the sponsor's model, not an interval generated by the summary-statistic recalculation. The registry record does not provide the model details needed to independently reconstruct the sponsor interval.

General explanation: a confidence interval communicates the range of effect estimates compatible with the data under a stated method. Its precision depends on sample information, variability and modeling assumptions. It does not account automatically for bias, missing-data problems, outcome selection or uncertainty about the model itself. A narrow interval can describe a precisely estimated quantity while still leaving questions about whether that quantity answers the intended clinical question.

For this trial, the analysis population required exposure and specific HbA1c measurements, and missing values were handled by last observation carried forward. These choices affect which observations inform the estimate and how incomplete follow-up is represented. The registry record does not report the extent or pattern of missing data, so the potential influence of that missingness on precision cannot be assessed from these figures alone. The sample-size planning illustrations are separate from confidence intervals and do not quantify the observed precision.

11. 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 sample-size estimates are planning illustrations based on the observed effect, not the trial's own original design calculations. The listed figures are participants per arm, not total participants across both arms and not total events. The calculations use the assumptions stated for the illustration: a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each row uses the stated fraction of the observed effect as the effect to detect.

The 80% power column gives the approximate number of participants per arm for a design that would have an 80% probability of detecting the specified effect under the calculation assumptions, if that effect and model were correct. The 90% power column gives the corresponding number per arm for a higher probability of detecting that effect. The figures are 4 and 5 participants per arm when planning for 100% of the observed effect; 7 and 9 when planning for 75% of the observed effect; and 14 and 19 when planning for 50% of the observed effect, respectively.

Smaller target effects generally require more participants because a smaller separation between groups is harder to distinguish from random variation. These estimates are illustrations, not a reconstruction of the trial's recruitment plan. The trial's original sample-size justification and design assumptions are not reported in the registry record. The values are not observed power.

The sample-size description distinguishes these continuous-outcome calculations from time-to-event calculations, for which figures would refer to total events. The primary HbA1c endpoint is continuous, so the estimates refer to participants per arm. They do not indicate that the trial enrolled those numbers for each planning scenario; the registry record lists actual enrollment of 480.

12. 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 = 0.150; pooled standard deviation σ = 0.070.
2. n = 2 × (1.96 + 0.842)² × 0.070² / 0.150² = 3.4, rounded up to 4 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 <- 0.15                      # observed difference in means
sigma <- 0.07                     # 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(469092)
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 = 4 (80% power), 5 (90% power)
 75% of observed effect: n per arm = 7 (80% power), 9 (90% power)
 50% of observed effect: n per arm = 14 (80% power), 19 (90% power)
power.t.test: 4.6 
Simulated power at n = 4 per arm: 0.717 (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.

13. Analyses Reported by the Sponsor

The registry's sponsor analysis uses linear regression and reports a net mean difference of -0.16, with a 95% confidence interval from -0.30 to -0.02 and p = 0.029. It is classified in the registry record as a non-inferiority or equivalence analysis. The specified margin and the decision criteria for that analysis are not reported in the registry record.

The recalculated comparison is an unadjusted mean difference of 0.15, ordered as insulin glargine minus biphasic insulin aspart 30, with p < 0.001. The sponsor's negative estimate uses the opposite direction for the group contrast, so the signs should not be read as indicating conflicting directions without first aligning the comparison order. The calculations also differ in method: the recalculation uses posted summary statistics, while the sponsor result is identified as regression.

The registry does not report enough detail to explain the difference in the estimates or reproduce the sponsor model. The sponsor result should therefore be presented as the trial's reported analysis, while the recalculation should be identified as an unadjusted analysis of the posted summaries. Neither result should be substituted for the other or interpreted without its method and contrast direction.

14. Multiplicity, Missing Data and Analysis Populations

The registry record lists one primary endpoint, HbA1c after 26 weeks, and five secondary outcome topics. Multiple outcomes create a multiplicity question: when many hypotheses are tested, the chance of at least one apparently notable result can increase. General explanation: statistical procedures can control or otherwise address this issue, but the appropriate approach depends on which comparisons were planned and how the analyses were organized. A multiplicity strategy is not reported in the registry record.

The registry record includes 12 analyses and 6 outcome measures, but does not specify which analyses were prespecified, whether all planned analyses were reported, or whether the p-values were adjusted for multiple testing. The presence of several listed outcomes does not itself establish that multiplicity adjustment was or was not performed. Without the analysis plan, the reported p-values should be interpreted as results of their respective analyses rather than as a complete family-wise assessment.

The primary analysis population is described as intention to treat with a specific qualification: randomized subjects had to be exposed to trial drug and have at least a baseline HbA1c measurement and one post-randomization HbA1c measurement. Missing HbA1c values were handled using last observation carried forward. This method carries the last observed value forward to represent a later missing value. General explanation: it can be sensitive to the timing of dropout and to whether the outcome would in fact remain stable after the last measurement. It does not automatically remove bias from missing data.

The registry record does not report the number of missing HbA1c observations, reasons for missingness, timing of discontinuations, or sensitivity analyses using alternative missing-data assumptions. It also does not report whether analyses were conducted in other populations, such as all randomized participants regardless of exposure or measurement availability. These omissions limit assessment of how robust the primary estimate is to the analysis-population definition and missing-data method.

The record identifies a non-inferiority or equivalence classification for the sponsor's linear regression analysis, but the margin, hypothesis formulation and protocol decision rule are not reported in the registry record. For such an analysis, the prespecified margin and the direction of the confidence interval relative to that margin are essential to interpretation. The reported classification can be quoted, but the missing criteria prevent an independent check of the formal conclusion.

The registry record does not report interim analyses, stopping rules, a full statistical analysis plan, or whether the outcome and analysis definitions were changed after the trial began. These details should not be inferred from the fact that the trial is completed or from the reported results.

15. 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: hypoglycaemia (0.8%), cerebrovascular accident (0.8%), angina unstable (0.4%), biliary colic (0.4%), gastroenteritis (0.4%).

The registry reports serious adverse event figures for 231 subjects at risk in the biphasic insulin aspart 30 group, with 13 affected, and 238 subjects at risk in the insulin glargine group, with 10 affected. These are the posted serious adverse event counts and their listed at-risk denominators. They should be read as descriptive counts for the safety groups, not as a randomized comparison adjusted for follow-up time or other factors.

The registry also lists serious adverse event terms and percentages: hypoglycaemia, 0.8%; cerebrovascular accident, 0.8%; angina unstable, 0.4%; biliary colic, 0.4%; gastroenteritis, 0.4%; and gastroenteritis Escherichia Coli, 0.4%. These entries describe the reported serious-event terms; they do not provide a full account of every event's timing, severity, relatedness or outcome.

For other adverse events, the registry lists 56 affected subjects among 231 at risk in the biphasic insulin aspart 30 group and 63 among 238 at risk in the insulin glargine group. Death counts are not reported in the registry record. The safety figures should be interpreted alongside the open-label design, the trial duration and the absence of detailed event-level information in the registry record. They do not, on their own, establish a causal difference between regimens.

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

The primary endpoint results describe HbA1c after 26 weeks of treatment in two randomized groups receiving different once-daily insulin regimens alongside metformin and glimepiride. The posted least-squares means were 7.08 for biphasic insulin aspart 30 and 7.23 for insulin glargine. The unadjusted recalculation expresses their contrast as 0.15 for insulin glargine minus biphasic insulin aspart 30. The sponsor analysis reports a net mean difference of -0.16 using the opposite contrast direction and a regression method.

These values characterize the observed endpoint under the analyses and population definitions recorded. They do not by themselves show which regimen would be preferable for a particular person, establish a difference in every clinically relevant outcome, or resolve the balance of treatment benefits and harms. The HbA1c outcome does not replace the separate information that would be needed about glucose profiles, target attainment, treatment satisfaction, hypoglycaemia and other adverse events.

The registry describes the sponsor analysis as non-inferiority or equivalence, but does not report the margin or decision rule. Without those details, the formal classification cannot be independently evaluated from the numerical result alone. The unadjusted recalculation is not a substitute for that analysis; it is a distinct summary-statistic comparison.

General explanation: statistical evidence and clinical importance are related but different. A statistically precise estimate can still require context about the size of a meaningful difference, the outcomes that matter, the duration of follow-up and the certainty of the analysis assumptions. The registry record does not provide a clinical-importance threshold for HbA1c or enough information to assess all those considerations.

The trial was randomized and parallel-group, which supports a comparative interpretation, but it was open-label. The primary population also required exposure and specified HbA1c measurements, and missing values were handled using last observation carried forward. Those features should be kept in view when interpreting the result. The registry record does not provide sufficient information to determine the impact of missing data, adherence, or any unreported analysis choices.

17. Why This Trial Matters Statistically

This trial offers a useful statistical example because it reports both group-level endpoint estimates and a sponsor analysis classified as non-inferiority or equivalence. The posted least-squares means and the recalculated unadjusted contrast are relatively straightforward to describe, while the sponsor analysis illustrates why contrast direction and model choice must be stated explicitly. The same treatment comparison can carry opposite signs when the order of groups is reversed.

It also demonstrates why a small p-value should not be treated as a complete interpretation. The recalculated p-value and sponsor p-value come from different methods, and the registry record does not provide the model details or non-inferiority margin needed to reconstruct the sponsor's formal decision. A p-value does not describe clinical importance, and a non-inferiority or equivalence label requires an explicit criterion to be fully assessed.

The endpoint and analysis-population definitions are also important. HbA1c is a continuous outcome measured after 26 weeks, but the primary analysis population required trial-drug exposure and specific measurements. Last observation carried forward was used for missing data. These details shape the meaning of the estimate and underscore why precision calculations cannot answer questions about bias or robustness to missingness.

Finally, the registry lists secondary outcomes and safety information, but the information available for this analysis does not report the secondary results or detailed analysis plans. The statistical interpretation is therefore strongest when confined to the posted primary endpoint, the reported methods and the stated limits of what the registry record documents.

18. 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 trial was open-label, with no masking listed. Knowledge of assigned treatment can affect participant behavior, treatment decisions or the assessment of outcomes that involve judgment or reporting. HbA1c is a laboratory measure, but open-label conduct can still influence aspects of care and follow-up. The registry record does not describe procedures used to reduce those risks.

The study included a 4 week run-in and 26 weeks of treatment. The information available does not describe longer-term outcomes, so the posted primary comparison is limited to the endpoint after 26 weeks. The trial was conducted across Africa, Asia, Europe, Oceania and South America, but country-level enrollment and regional subgroup results are not reported in the registry record.

The primary analysis population required trial-drug exposure and both baseline and post-randomization HbA1c measurements. Last observation carried forward was used for missing data. The numbers and reasons for missing measurements, withdrawals and treatment discontinuations are not reported in the registry record, nor are sensitivity analyses using other assumptions. The extent to which those choices affected the result therefore cannot be assessed.

The registry record does not report the original sample-size justification, full statistical analysis plan, interim analyses, covariate adjustment details, or the non-inferiority margin. It consequently does not permit a complete audit of the planned analysis or independent verification of the sponsor's non-inferiority or equivalence classification. The posted unadjusted recalculation is not a replacement for missing protocol details.

Finally, the registry lists secondary outcomes and adverse events, but detailed results for the listed secondary outcomes are not reported in the registry information used. The available safety summary does not include event-level descriptions, exposure-adjusted rates or reported death counts. These limitations leave important questions about the broader outcome profile unresolved.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Comparison of Biphasic Insulin Aspart 30 Versus Insulin Glargine Both in Combination With Metformin and Glimepiride in Subjects With Type 2 Diabetes — NCT00469092.
• Registry record last updated 2017-02-23; status last verified 2017-01.
• Results first posted 2010-01-06.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

22. Record Summary

NCT00469092 is a completed phase 4 randomized, open-label, parallel-group trial with listed enrollment of 480. It compared once-daily biphasic insulin aspart 30 with once-daily insulin glargine, both in combination with metformin and glimepiride, in insulin-naive subjects with type 2 diabetes. The official title describes a 4 week run-in and 26 weeks of treat-to-target treatment.

The primary endpoint was HbA1c after 26 weeks. Posted least-squares means were 7.08 and 7.23, respectively. The sponsor's linear regression analysis reported a net mean difference of -0.16, a 95% confidence interval from -0.30 to -0.02, and p = 0.029; the registry classifies it as non-inferiority or equivalence. The margin and full statistical plan are not reported in the registry record.

The primary analysis used a defined intention-to-treat population and last observation carried forward for missing values. The registry lists five secondary outcome topics, but their results are not reported in the information considered for this summary.