Comparing the Efficacy and Safety of Biphasic Insulin Aspart 30 and Biphasic Human Insulin 30 on Blood Sugar Control in Subjects With Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Comparing the Efficacy and Safety of Biphasic Insulin Aspart 30 and Biphasic Human Insulin 30 on Blood Sugar Control in Subjects With Type 2 Diabetes (NCT00807092), a randomized phase 4 study. The primary outcome is Change in IAUC (Incremental Area Under the Curve) for Postprandial Glucose (0-4 Hours) Over 3 Main Meals.

2. Clinical Question

The question is comparative: it concerns the difference between two insulin-plus-metformin strategies rather than either regimen in isolation. The registry identifies a randomized parallel-group design, but masking was none. The primary outcome summarizes post-meal glucose exposure across breakfast, lunch, and dinner, rather than focusing on a single fasting measurement. The reported arm means both changed downward, with a very small difference between them in the posted summaries. The registry record does not report a clinical threshold for an important difference, so the size of the observed contrast cannot be classified against such a threshold from this record alone.

3. Trial Design

This was a phase 4, randomized, open-label, parallel-group trial sponsored by Novo Nordisk A/S. It was marked completed. The registry lists enrollment as 145 and identifies the study setting as Asia; the trial summary specifies Chinese insulin-naïve subjects with type 2 diabetes inadequately controlled on oral antidiabetic drug therapy. The record lists diabetes and type 2 diabetes mellitus as conditions.

The two groups received either BIAsp 30 or BHI 30, both in combination with metformin. BIAsp 30 was given subcutaneously twice daily before breakfast and dinner. BHI 30 was given subcutaneously twice daily 30 minutes before breakfast and dinner. Both arm descriptions report an initial total daily dose of 0.3 U or IU/kg body weight and individual dose adjustment over the first 4 weeks. Additional protocol detail for the BHI 30 arm is not reported in the registry record.

The recorded start was 2008-12 and the primary completion was 2009-10. The primary endpoint timeframe was week 0 and week 6. Continuous glucose monitoring was used to monitor blood glucose profiles for 72 hours at baseline and at the end of treatment. The registry specifies that the full analysis set included randomized subjects exposed to at least one dose of a trial product, and that last observation carried forward was used.

Randomization helps create comparable groups on average, but the record does not report the method used to generate or conceal the allocation sequence. Masking was none, so participants and investigators were not masked according to the registry entry. The record does not report the sample-size justification, the original design assumptions, or whether any interim analyses were planned. Those details cannot be inferred from enrollment or from the later planning illustrations.

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: All eligible subjects had their baseline blood glucose profiles monitored by continuous glucose monitoring system (CGMS) for 72 hours. Following CGMS uninstall and discontinuation of all previous oral anti-diabetic drug (OAD) treatment, subjects were randomised to one of two treatment groups.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The full analysis set using LOCF (Last Observation Carried Forward) consists of all randomised subjects who had been exposed to at least one dose of the trial products.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Change in IAUC (Incremental Area Under the Curve) for Postprandial Glucose (0-4 Hours) Over 3 Main Meals, measured over Week 0, week 6.

The primary endpoint was change in incremental area under the curve (IAUC) for postprandial glucose over three main meals. IAUC describes glucose exposure above the relevant baseline over a measurement period; in this trial the period was 0–4 hours after each meal. An area-based measure captures both the size and duration of post-meal glucose excursions, rather than representing only a single glucose reading.

Continuous glucose monitoring was performed for 72 hours at week 0 and week 6. The registry says IAUC was calculated using the trapezoidal method. For each monitoring period, the value used was the arithmetic mean of the three meal-specific incremental areas from day 1 and day 2. The endpoint therefore combines information from breakfast, lunch, and dinner and compares the measure at baseline with the measure at the end of treatment.

The registry reports least-squares means with standard errors for the primary endpoint, in mmol/L. The analysis population was the full analysis set using last observation carried forward, defined as all randomized subjects who had received at least one dose of a trial product. The registry record does not report a threshold for a clinically important change or explain how missing monitoring data were handled beyond the stated carry-forward approach. The endpoint indicates a glucose-profile change over the monitoring period; by itself, it does not establish the clinical importance of a difference between regimens.

7. Results Posted in the Registry

• BHI 30: mean -1.98 versus -1.99 on control; difference 0.01 (95% CI -1.38 to 1.41), standardized effect d = 0.00; z-test p = 0.985.

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.

Mean FBG (Fasting Blood Glucose) Assessed by CGMS

• 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:
  • BHI 30: mean 6.41 versus 6.86 on control; difference -0.45 (95% CI -0.99 to 0.10), standardized effect d = -0.28; z-test p = 0.108.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between BHI 30 and the control. The interval gives the range of effects that remain compatible with the data.

Change in Mean FBG Assessed by CGMS

• 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:
  • BHI 30: mean -2.56 versus -2.11 on control; difference -0.45 (95% CI -0.99 to 0.10), standardized effect d = -0.28; z-test p = 0.108.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between BHI 30 and the control. The interval gives the range of effects that remain compatible with the data.

Change in FPG (Fasting Plasma Glucose)

• 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:
  • BHI 30: mean -3.46 versus -2.96 on control; difference -0.50 (95% CI -1.01 to 0.02), standardized effect d = -0.32; z-test p = 0.061.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between BHI 30 and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists a range of secondary outcomes, including fasting glucose, glucose profiles, post-meal increments, glycaemic excursions, and glycated albumin. Numerical summaries are available for three secondary outcome records. Among those records, the fasting glucose comparisons generally point toward lower fasting values or larger decreases in the BHI 30 group. The posted results do not form a uniformly decisive pattern: the estimates are uncertain, and the sponsor's reported adjusted findings do not provide consistent evidence of group separation across these measures.

These secondary measures add context to the primary meal-related IAUC outcome because they describe other aspects of glucose control. They do not replace the primary question, and similar direction across related outcomes does not establish a separate treatment effect. The primary comparison itself remains close to zero in the posted summaries, with intervals that include zero. The secondary data therefore do not turn the primary result into evidence of equivalence or of a clear difference.

Many outcomes were assessed, and the registry record does not report a multiplicity adjustment or a hierarchy for interpreting secondary tests. When numerous comparisons are examined, isolated nominal findings can occur by chance. Unadjusted secondary p-values carry less weight for confirmatory conclusions unless their role and multiplicity handling are specified. Numerical results for the other listed secondary outcomes are not reported in the registry record, so a complete outcome-by-outcome synthesis cannot be made from this record.

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

• Change in Mean IAUC for Postprandial Glucose (0-4 Hours) After Each Meal (Breakfast, Lunch, Dinner) Assessed by CGMS
• Mean FBG (Fasting Blood Glucose) Assessed by CGMS
• Change in Mean FBG Assessed by CGMS
• Change in FPG (Fasting Plasma Glucose)
• Change in 8-point SMBG (Self-monitored Blood Glucose) Profiles
• Change in Prandial Blood Glucose Increment
• Change in MAGE (Mean Amplitude of Glycaemic Excursions) Assessed by CGMS
• Change in GA (Glycated Albumin)

9. Statistical Interpretation

10. Statistical Methods Explained

Several statistical ideas help distinguish the values in the registry from the conclusions they can support. The primary outcome is continuous: it is a measured quantity that can take a range of values. The arm summaries are least-squares means, which are model-based averages, and their accompanying spread is reported as a standard error. A standard error describes uncertainty in an estimated mean; it is not the same thing as the spread of individual participants' measurements.

What does the primary comparison estimate?

The comparison concerns the difference in mean change between the BHI 30 and BIAsp 30 groups. The re-calculation uses the posted arm-level means and their standard errors to form a simple, unadjusted contrast. It does not reproduce a participant-level model. The posted re-calculated estimate is 0.013, while the sponsor's reported ANCOVA estimate is −0.013. A contrast's sign depends on which group is subtracted from which; the registry record does not state the sign convention in enough detail to reconcile the signs. The estimates are both close to zero.

The sponsor used ANCOVA, or analysis of covariance, for its reported primary analysis. In general, ANCOVA compares groups while adjusting for specified covariates, often including a baseline measure. Adjustment can improve precision or account for baseline imbalance when the model is appropriately specified. The registry names ANCOVA but does not report the covariates, model specification, or diagnostics. Those details are not reported in the registry record and should not be assumed.

How should the interval and p-value be read?

A confidence interval describes a range of values compatible with the estimate under the analysis and its assumptions. A narrower interval indicates greater precision; a wider interval indicates more uncertainty. The re-calculated interval for the primary comparison extends from −1.383 to 1.409. It includes zero, so the summary-statistic comparison does not distinguish a positive difference from a negative difference. The reported sponsor interval is a 95% confidence interval, from −1.332 to 1.306, and also includes zero.

The re-calculated p-value is 0.985. In general, a p-value measures how unusual data at least as incompatible with a null hypothesis would be under that null hypothesis and the specified analysis. It is not the probability that the null hypothesis is true, nor a measure of clinical importance. The sponsor did not report a primary p-value in the registry record. A missing p-value should not be reconstructed from rounded summaries and treated as the sponsor's result.

Why do adjusted and unadjusted estimates differ?

An unadjusted comparison is calculated from group summaries without the covariate adjustment used in ANCOVA. The adjusted result may therefore differ in estimate and precision. Also, the registry's arm values and the reported estimates are rounded. Re-calculations based on rounded summaries may not reproduce a model fitted to participant-level data. In this case the registry discrepancy flag is false, but that flag does not mean the methods are identical. The sponsor's ANCOVA remains the reported trial analysis, and the re-calculation should be understood as a separate descriptive check.

What does over-dispersion mean in general?

For count outcomes, over-dispersion means that observed variability exceeds what a simple model assumes. It can make standard errors too small if ignored. That concept does not directly define this trial's primary continuous glucose outcome, and the registry does not report an over-dispersion model for the primary analysis. For continuous measurements, the relevant issues include the outcome's variability, measurement process, model assumptions, and missing-data handling. These modeling details are not fully described in the registry record.

What role does missing-data handling play?

The stated primary analysis used the full analysis set and last observation carried forward (LOCF). In general, LOCF substitutes a participant's last observed value for a later missing value. It is simple to describe, but it assumes that the carried-forward value is an appropriate stand-in for the unobserved measurement. That assumption may not hold when outcomes change over time. The registry record does not report the amount or pattern of missing primary outcome data, reasons for missingness, or sensitivity analyses using alternative assumptions.

The full analysis set definition includes randomized subjects exposed to at least one dose. This is the population stated for the primary endpoint; it should not be casually relabeled as a strict intention-to-treat analysis without further detail. The registry lists the definition but does not provide a participant flow or an account of exclusions for each endpoint.

Why are secondary tests more tentative?

Secondary outcomes can give context about related measures, but conducting many comparisons increases the chance of at least one apparently notable result arising by chance. Multiplicity procedures can control this issue, but the registry record does not report whether any were used. The p-values posted for secondary ANCOVA analyses should therefore be read as individual results unless the record documents a family-wise or other adjustment. The primary endpoint remains the main prespecified question identified in the registry, while secondary results are supporting and exploratory in interpretation.

11. Confidence Intervals and Precision

Precision is about how tightly an estimate is determined by the data and analysis. The primary re-calculated mean difference is 0.013, but its interval extends from −1.383 to 1.409. The sponsor's ANCOVA estimate is −0.013, with a reported 95% confidence interval from −1.332 to 1.306. Both intervals include zero and include values in either direction. Thus, the point estimates alone do not convey the uncertainty: a near-zero estimate is accompanied by a range that permits materially different contrasts.

A confidence interval is not a guarantee that the true effect lies within its endpoints, nor does it show the probability that a particular effect is correct. In repeated use of a method under its assumptions, intervals at a stated confidence level have a long-run coverage interpretation. The sponsor's reported interval is explicitly 95%; the registry does not label a confidence level for the simple re-calculated interval in the comparison summary. The width of either interval should be considered alongside the outcome's scale and any clinical-importance threshold. Such a threshold is not reported in the registry record.

The sponsor's interval is adjusted using ANCOVA, while the re-calculated interval comes from posted summary statistics. They need not match exactly because they use different information and methods, and because the registry values are rounded. The record does not report a primary p-value from the sponsor analysis. The re-calculated p-value of 0.985 should not be treated as if it were the sponsor's inferential result.

For the secondary fasting glucose results, the reported intervals also show uncertainty, and some include zero while the sponsor's reported ANCOVA findings differ in whether their intervals include zero. These additional estimates do not remove uncertainty in the primary outcome. A failure to demonstrate a difference is not proof that the regimens are equivalent; equivalence would require a stated margin and a design and analysis suited to that question. No equivalence margin is reported in the registry record.

12. Planning a Similar Trial: Sample Size

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

The planning figures are participants per arm, not total participants or total events. They are illustrations based on the observed effect, not the original trial's design calculation. The registry's original sample-size assumptions are not reported in the registry record. The planning assumptions listed are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set as the stated fraction of the observed effect.

The 80% power column describes the participant count per arm that, under those assumptions, would give an 80% chance of detecting the chosen effect if that effect were the true effect and the analysis assumptions held. The 90% power column uses a higher target chance of detection and consequently requires more participants per arm for the same assumed effect. These are planning concepts, not guarantees about what a future study will find.

Rows for planning for 75% or 50% of the observed effect use smaller target effects than the row for 100% of the observed effect. Smaller effects are harder to distinguish from random variation, so the planning figures rise as the target effect gets smaller. The values should be read as illustrations anchored to an effect estimated in this trial, not as an independent estimate of a clinically important effect. The primary outcome's very small observed contrast leads to very large planning figures; this reflects the chosen planning inputs and should not be interpreted as a statement about the trial's actual power.

This is not observed power. Observed power is not a useful substitute for reading the estimate and its confidence interval; the planning figures instead show how sample requirements change under specified assumptions. They do not revise the trial's enrollment or provide evidence that the original design was adequate or inadequate.

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 = 0.013; pooled standard deviation σ = 4.056.
2. n = 2 × (1.96 + 0.842)² × 4.056² / 0.013² = 1,528,088.2, rounded up to 1,528,089 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 <- 0.0129999999999999                      # observed difference in means
sigma <- 4.05600734443677                     # 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(807092)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 1,528,089 (80% power), 2,045,677 (90% power)
 75% of observed effect: n per arm = 2,716,602 (80% power), 3,636,759 (90% power)
 50% of observed effect: n per arm = 6,112,353 (80% power), 8,182,707 (90% power)
power.t.test: 1528089 
Simulated power at n = 1,528,089 per arm: 0.797 (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's primary sponsor analysis is an ANCOVA with a net mean difference of −0.013 and a 95% confidence interval from −1.332 to 1.306. The sponsor p-value is not reported in the registry record. This is the reported adjusted analysis and should be distinguished from the unadjusted re-calculation, which gives a mean difference of 0.013, an interval from −1.383 to 1.409, and p = 0.985.

The re-calculation uses posted arm-level means and standard errors. It does not reproduce the sponsor's participant-level ANCOVA or its covariate adjustment. Rounding and different contrast sign conventions can also affect a direct comparison of displayed estimates. The sponsor and re-calculated estimates have opposite signs, but each is close to zero and each interval includes zero. The registry discrepancy flag is false; no material discrepancy is identified in the record. The two results nevertheless answer related questions using different analysis methods, so they should not be presented as if they were the same calculation.

The sponsor result is the trial's reported primary analysis. The re-calculation is an independent, simple summary-statistic comparison intended to show the approximate unadjusted contrast, not to replace the sponsor analysis or assign a sponsor p-value.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies one primary endpoint and lists multiple secondary outcomes. It does not report whether the secondary hypotheses were organized in a formal testing hierarchy, whether family-wise error was controlled, or whether any other multiplicity adjustment was used. Without that information, secondary p-values should be interpreted as individual comparisons rather than as a controlled set of confirmatory tests. The possibility of chance findings increases when many outcomes are tested, especially when no adjustment or prespecified ordering is documented.

The primary endpoint analysis population is described as the full analysis set: randomized subjects who received at least one dose of a trial product. LOCF was used for missing data. The registry record does not report the number of missing primary measurements, the reasons for missingness, the timing of withdrawals, or the amount of information carried forward. It also does not report sensitivity analyses to assess whether conclusions change under different missing-data assumptions.

LOCF is a method for filling a missing later measurement with the last recorded value. As a general statistical matter, that method can be biased if the outcome would have changed after the last observation. The size and direction of any such bias depend on why data are missing and how outcomes evolve, neither of which is fully. The registry's mention of LOCF does not establish that missingness had no effect.

The record names ANCOVA for the primary sponsor analysis but does not give the covariates or the full model specification. It also does not report the method of random sequence generation or allocation concealment. Interim analyses, stopping rules, and the original sample-size justification are not reported in the registry record. These omissions limit independent assessment of the analysis plan, selective reporting, and the robustness of the results.

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

Serious events with the highest arm-level rate: hypoglycaemic coma (2.7%), loose body in joint (1.4%), hepatic cancer metastatic (1.4%), pancreatic carcinoma (1.4%), soft tissue injury (1.4%).

The serious adverse event figures are reported by arm. For BIAsp 30, 1 serious adverse event was reported among 71 participants at risk; for BHI 30, 3 were reported among 73 participants at risk. No other adverse events were reported as affecting participants in either group in the registry's adverse-event summary. The registry lists deaths as null, so death information is not reported in the registry record.

The serious-event terms listed include hypoglycaemic coma at 2.7%, loose body in joint at 1.4%, hepatic cancer metastatic at 1.4%, pancreatic carcinoma at 1.4%, and soft tissue injury at 1.4%. The record does not provide enough detail to assign every listed term to a particular arm or to determine whether each term represents an event or an affected participant. It also does not provide event timing, severity beyond the serious classification, causal assessment, or clinical narratives.

These counts are descriptive. They should not be interpreted as a reliable comparative safety estimate on their own, particularly without details on follow-up, event ascertainment, and the small number of serious events. The registry record does not support conclusions about causality or overall safety differences.

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

The primary posted estimates are close to zero, and both the unadjusted re-calculated interval and the sponsor's reported 95% interval include zero. The data therefore do not establish a clear difference between the regimens on the primary measure in this trial. At the same time, the intervals cover effects in either direction, so the results do not demonstrate that the regimens are equivalent or that any possible difference is negligible.

Statistical evidence and clinical importance are different questions. A confidence interval describes uncertainty about an estimated contrast, whereas a clinical threshold indicates what size of difference would matter in practice. The registry record does not report such a threshold. Without it, the clinical relevance of the interval's possible values cannot be determined from the posted results alone.

The trial's endpoint reflects postprandial glucose exposure over a defined monitoring period, not every aspect of diabetes management or long-term health. The record reports a six-week endpoint timeframe and secondary glucose measures, but the numerical results available do not establish long-term outcomes. The safety information is also limited to registry event summaries, with little contextual detail.

These results should be understood as evidence from a randomized comparison in the described insulin-naïve Chinese population receiving metformin with one of two premixed insulin regimens. They do not automatically apply to other populations, background treatments, schedules, or settings. The registry provides no basis for medical advice or for a conclusion that either regimen works, fails, or is preferable. It supports a narrower statistical description: the primary estimated contrast is small relative to its uncertainty, while the record leaves important questions about precision, missing data, clinical importance, and safety context unresolved.

18. Why This Trial Matters Statistically

This trial illustrates why a treatment comparison cannot be summarized by its point estimate alone. The primary arm means are similar, but the interval is broad enough to include differences in both directions. Reporting the estimate together with its uncertainty gives a more informative account than treating a near-zero point estimate as proof of no difference.

It also shows the distinction between a sponsor's adjusted analysis and an unadjusted re-calculation from group summaries. The reported ANCOVA and simple re-calculation are related but not interchangeable. Their displayed point estimates use opposite signs, while both are close to zero and both intervals include zero. Readers need the analysis method and contrast direction to interpret such estimates correctly.

The endpoint's construction matters as well. IAUC combines meal-specific glucose exposure across breakfast, lunch, and dinner and compares monitoring periods at baseline and the end of treatment. That is a different question from fasting glucose alone. Secondary measures can help describe the pattern, but multiple outcomes and unreported multiplicity handling limit how strongly they can support confirmatory claims.

Finally, the record highlights how missing-data methods, analysis-population definitions, and safety-event context affect interpretation. LOCF is reported, but the extent and pattern of missing data are not. The serious-event counts are available, but the event narratives and causal assessments are not. These are not peripheral details: they help determine how precise, robust, and clinically interpretable the statistical comparison can be.

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 short endpoint timeframe is week 6, and the primary outcome is a glucose-monitoring measure rather than a long-term clinical outcome. The registry record does not report whether effects were sustained beyond the recorded treatment endpoint. It also does not report a clinical-importance threshold for IAUC, making it difficult to judge whether the interval's possible differences matter clinically.

The trial was open-label, with masking listed as none. The primary endpoint used continuous glucose monitoring, but the registry does not report whether outcome processing or analysis was masked. The method of random sequence generation and allocation concealment is also not reported in the registry record.

The primary analysis used LOCF, but the record does not describe the amount or pattern of missingness or report sensitivity analyses. ANCOVA is named for the sponsor's analysis, but the covariates and full model details are not reported. The sponsor's primary p-value is not reported. The available record therefore does not allow full reconstruction of the primary analysis from the registry summary.

Only some secondary outcomes have numerical results in the registry record, and multiplicity handling is not reported. Safety reporting includes counts but lacks event narratives, timing, and causal assessments; deaths are not reported. Finally, the trial population is described as Chinese insulin-naïve subjects with type 2 diabetes inadequately controlled on oral antidiabetic drugs, limiting how confidently results can be generalized beyond that population and regimen context.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Comparing the Efficacy and Safety of Biphasic Insulin Aspart 30 and Biphasic Human Insulin 30 on Blood Sugar Control in Subjects With Type 2 Diabetes — NCT00807092.
• Registry record last updated 2015-03-04; status last verified 2015-02.
• Results first posted 2011-04-11.
• 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

NCT00807092 was a completed phase 4 randomized, open-label, parallel-group trial sponsored by Novo Nordisk A/S. It compared BIAsp 30 plus metformin with BHI 30 plus metformin in insulin-naïve Chinese subjects with type 2 diabetes inadequately controlled on oral antidiabetic drug therapy. The record lists enrollment as 145 and gives a start of 2008-12 and primary completion of 2009-10.

The primary endpoint was change in postprandial glucose IAUC over three main meals, measured by continuous glucose monitoring at week 0 and week 6. The sponsor reported an ANCOVA net mean difference of −0.013, with a 95% confidence interval from −1.332 to 1.306; a primary p-value is not reported. The re-calculated unadjusted estimate is 0.013, with an interval from −1.383 to 1.409 and p = 0.985.

The main statistical reading is uncertainty around a primary estimate near zero, not proof of equivalence. Missing-data sensitivity analyses, multiplicity handling, the original sample-size justification, and further detail on safety events are not reported in the registry record.