A Study for Patients With Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study for Patients With Type 2 Diabetes (NCT01027871), a randomized phase 2 study. The primary outcome is Fasting Blood Glucose (FBG) Level at Week 12 Endpoint as Measured by the 8-Point Self-Monitored Blood Glucose (SMBG) Profiles.

2. Clinical Question

3. Trial Design

This was a randomized, parallel-group phase 2 study sponsored by Eli Lilly and Company. The registry lists 289 participants enrolled and a completed status. The study began in 2010-01, with primary completion in 2010-12. The brief summary describes a comparison of fasting blood glucose after 12 weeks of treatment with LY2605541 or insulin glargine in people with type 2 diabetes.

The intervention included two LY2605541 dosing algorithms as well as the insulin glargine comparison group. Algorithm 1 included LY2605541 alongside pre-study insulin for the first several days. Algorithm 2 used only LY2605541, with the first dose doubled. The registry does not describe the insulin glargine dosing approach in the arm description. No masking was recorded, so the study is described as having none; further details about who knew each participant’s assignment are not reported in the registry record.

The primary outcome used data from 8-point self-monitored blood glucose profiles. Its analysis population was not simply everyone enrolled: participants needed to be randomized, take at least one dose, have a baseline measurement, and have at least one post-baseline measurement. The reported numbers analyzed differ by group and outcome, as expected when outcome-specific analysis populations are used. The registry record does not report the original sample-size justification or further details of the randomization procedure.

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.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomized participants who took at least one dose of study drug with non-missing baseline value and at least one non-missing post-baseline value.

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: All randomized participants who received at least one dose of study drug.

6. Primary Endpoint

The registry lists the primary outcome as Fasting Blood Glucose (FBG) Level at Week 12 Endpoint as Measured by the 8-Point Self-Monitored Blood Glucose (SMBG) Profiles, measured over Week 12.

The primary endpoint was fasting blood glucose at the Week 12 endpoint, measured using 8-point self-monitored blood glucose profiles. The profile included morning fasting blood glucose, pre-meal and post-meal measurements around the three main meals, bedtime blood glucose, and a measurement at 0300 hours. The endpoint analyzed the morning fasting value rather than treating all profile readings as interchangeable.

The reported group summaries were least-squares means from a mixed-model repeated-measures approach. The unit was millimoles per liter (mmol/L), and the registry reported standard errors for the group estimates. A least-squares mean is a model-based estimate that accounts for the terms included in the analysis, rather than just a raw arithmetic average. The model description lists treatment, dose conversion timing, stratification variables, visit, the visit-by-treatment interaction, and a participant random effect.

General: fasting blood glucose is a measure of glucose before eating and can be used to compare glycemic measurements across treatment groups. Repeated measures make it possible to analyze measurements collected at more than one visit while accounting for their relationship within a participant. The registry’s stated endpoint is specifically the Week 12 fasting value; it should not be confused with change from baseline, which is listed separately as a secondary outcome.

7. Results Posted in the Registry

• LY2605541 Algorithm 1: mean 7.15 versus 6.83 on control; difference 0.32 (95% CI -0.29 to 0.93), standardized effect d = 0.16; z-test p = 0.304.
• LY2605541 Algorithm 2: mean 6.84 versus 6.83 on control; difference 0.01 (95% CI -0.60 to 0.62), standardized effect d = 0.00; z-test p = 0.974.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Change From Baseline in Fasting Blood Glucose (FBG) at Week 12 Endpoint

• 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:
  • LY2605541 Combined: mean -1.97 versus -1.77 on control; difference -0.20 (95% CI -0.86 to 0.46), standardized effect d = -0.07; z-test p = 0.550.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between LY2605541 Combined and the control. The interval gives the range of effects that remain compatible with the data.

Change From Baseline in Hemoglobin A1c (HbA1c) at Week 12 Endpoint

• 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:
  • LY2605541 Combined: mean -0.74 versus -0.64 on control; difference -0.10 (95% CI -0.30 to 0.10), standardized effect d = -0.13; z-test p = 0.317.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between LY2605541 Combined and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants With Hypoglycemia From Baseline Through Week 12

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • LY2605541 Combined: 53.8% (105/195) versus 63.4% (59/93) on control; risk difference -9.6 percentage points (95% CI -21.6 to 2.4), risk ratio 0.85 (0.69–1.04); unadjusted two-proportion z-test p = 0.124.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between LY2605541 Combined and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary results add information about change from baseline in fasting blood glucose, change from baseline in HbA1c, and the proportion of participants with hypoglycemia. The continuous outcomes show negative estimated changes from baseline in both the insulin glargine and combined LY2605541 groups. The reported comparisons for change in fasting blood glucose and HbA1c have confidence intervals that include zero. For hypoglycemia, the observed percentages were 63.4% with insulin glargine and 54.0% with combined LY2605541; the reported Fisher exact test has p = 0.162.

Overall, these secondary results do not remove the uncertainty in the primary endpoint comparison. They offer related but distinct information: fasting blood glucose change and HbA1c capture different measures and summaries, while hypoglycemia is a safety-related outcome. The record includes other secondary outcomes, including additional glucose measures and insulin dose, but not all their results are represented in the posted secondary results.

General: when many outcomes are tested, the chance of at least one apparently unusual result can increase even if no true effect exists. A multiplicity procedure can control that error rate, but the registry record does not report whether a multiplicity adjustment was prespecified or applied. The listed secondary p-values are therefore best treated as exploratory evidence rather than as a set of independent confirmatory tests. Their interpretation should also consider the analysis populations, outcome definitions, and confidence intervals, rather than relying on whether an individual p-value crosses a conventional threshold.

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

• Change From Baseline in Fasting Blood Glucose (FBG) at Week 12 Endpoint
• Change From Baseline in Hemoglobin A1c (HbA1c) at Week 12 Endpoint
• Percentage of Participants With HbA1c <7.0% and ≤6.5% at Week 12 Endpoint
• Percentage of Participants With HbA1c <7.0% and HbA1c ≤6.5% at Week 12 Endpoint Who Did Not Experience a Hypoglycemic Episode During Treatment
• 8-Point Self-Monitored Blood Glucose (SMBG) Measures at Week 12 Endpoint
• Daily Basal Insulin Dose at Week 2 and Week 12
• Percentage of Participants With Hypoglycemia From Baseline Through Week 12
• Rate of Hypoglycemia Per 30 Days From Baseline Through Week 12

9. Statistical Interpretation

10. Statistical Methods Explained

The primary analysis is described as a mixed-model repeated-measures (MMRM) approach. It estimates least-squares means for fasting blood glucose and compares treatment groups while using repeated measurements over visits. The registry lists fixed effects for treatment, dose conversion (pre-interim analysis and post-interim analysis), stratification variables, visit, and the interaction between visit and treatment. It also lists a random effect for participant. The registry does not report further details of the model specification, such as the covariance structure or the method used to estimate degrees of freedom.

General: repeated measurements from the same person are usually correlated, so treating each measurement as independent can misrepresent uncertainty. A repeated-measures model accounts for that within-participant structure. Fixed effects represent modeled factors whose levels are of direct interest, while a participant random effect represents participant-level variation. The specific interpretation of model terms depends on the full model and its assumptions; those details are not all reported in the registry record.

The model’s least-squares means are adjusted estimates, not simple unadjusted group averages. The posted primary means have standard errors of 0.22 mmol/L in each group. A standard error describes the estimated uncertainty of a mean, rather than the spread of individual participant measurements. The registry does not report standard deviations for the primary group results.

The unadjusted recalculations compare the posted group estimates directly. They are useful as a transparent descriptive check, but they do not reproduce the MMRM. The recalculation also uses an approximate pooled standard deviation inferred from the posted standard errors, as indicated by the analysis metadata. It does not incorporate the model’s listed covariates, repeated visits, or participant-level structure. Its p-values and intervals should therefore not be treated as substitutes for the sponsor’s fitted model.

For Algorithm 2 versus insulin glargine, the sponsor’s reported analysis gives a least-squares mean difference of 0.17, a 90% confidence interval from -0.18 to 0.52, and p = 0.433. The simple recalculation gives a difference of 0.01 and p = 0.974. These are not contradictory estimates of the same analysis: they arise from different approaches, and the registry’s model-based result is the relevant reported analysis. The registry record does not give a corresponding sponsor-reported comparison for Algorithm 1.

General: an interval estimate expresses uncertainty about an effect estimate under the assumptions of the analysis. An interval that includes zero is compatible with no difference as well as with differences in either direction, within the interval’s stated confidence level. It does not establish that all values within the interval are equally likely or that the groups are equivalent. The registry-reported primary interval is explicitly 90%; the confidence level for the simple recalculated intervals is not stated in the registry entries and should not be inferred.

Several secondary outcomes are continuous measures and are reported using least-squares means and standard errors. The registry also lists a binary outcome, the percentage of participants with hypoglycemia, with a Fisher exact test in the reported analysis. General: for a binary outcome, a risk difference describes the difference between group proportions, whereas a Fisher exact test evaluates evidence against a null association under its test assumptions. A test result does not establish causality or show whether an outcome is clinically important.

Over-dispersion is a general issue for count outcomes when the observed variability exceeds what a simple count model assumes. The listed secondary outcomes include a rate of hypoglycemia per 30 days, but no results for that outcome are included in the posted secondary results. It is therefore not possible to assess the distribution of those counts or whether over-dispersion was considered. The registry record does not report the full statistical analysis plan, model diagnostics, or procedures for handling every outcome type.

11. Confidence Intervals and Precision

The sponsor-reported primary comparison has a 90% confidence interval from -0.18 to 0.52 for the least-squares mean difference between LY2605541 Algorithm 2 and insulin glargine. Because this interval includes zero, the reported result is compatible with no difference as well as with differences in either direction under the model. The point estimate is 0.17, but the interval shows that the estimate is not precise enough to pin down a single value.

The recalculated comparisons are also uncertain. For Algorithm 1 versus insulin glargine, the recalculated mean difference is 0.32 mmol/L, with an interval from -0.290 to 0.930. For Algorithm 2 versus insulin glargine, the recalculated mean difference is 0.01 mmol/L, with an interval from -0.600 to 0.620. These are simple unadjusted recalculations, not the registry’s MMRM estimates. The confidence level for these recalculated intervals is not reported in the registry record.

General: a confidence interval is a range produced by a statistical procedure that, under its assumptions, reflects sampling uncertainty. A wider interval indicates less precision than a narrower one, but precision is not the same as freedom from bias. An interval can be wide because the data are limited or variable, and an analysis can still be affected by design or measurement limitations that the interval does not capture.

The group standard errors for the primary endpoint are 0.22 mmol/L. They describe uncertainty around each model-based group mean and are not standard deviations of participant outcomes. The enrollment count of 289 is not the same as the number contributing to each endpoint analysis. The registry’s primary analysis population required baseline and post-baseline values, and the group counts for the primary endpoint were 89, 86, and 91. The record does not explain all reasons that participants may be absent from a particular endpoint analysis.

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 values are illustrations based on the observed effect, not the trial’s original design assumptions. For the continuous outcomes and the binary hypoglycemia outcome shown in the planning information, the figures are participants per arm. The planning assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is the stated fraction of the observed effect. Time-to-event planning values, when present, are total event counts using the Schoenfeld method. The original design assumptions are not reported in the registry record.

The 80% power column gives the planned number of participants per arm for a hypothetical study to have an 80% chance of detecting the specified effect under those assumptions. The 90% power column gives the corresponding planning number for a 90% chance. These are not guarantees about the result of a future study: the calculations depend on the assumed effect and the statistical model used for planning.

The planning rows consider 100%, 75%, and 50% of the observed effect. Smaller effects require more participants per arm to achieve the same target power. For example, the primary Algorithm 1 comparison gives 650 participants per arm for 80% power when planning for 100% of the observed effect, and 2,598 participants per arm when planning for 50% of it. The very small observed difference for Algorithm 2 leads to much larger planning figures: 683,880 participants per arm for 80% power at 100% of that observed effect. Such values are highly sensitive to the estimated effect and should not be interpreted as a recommendation for a trial design.

These calculations are not observed power. They are forward-looking illustrations based on the observed effect sizes, and the registry record does not report that they were used to design this trial.

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 = 0.320; pooled standard deviation σ = 2.058.
2. n = 2 × (1.96 + 0.842)² × 2.058² / 0.320² = 649.4, rounded up to 650 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 <- 0.32                      # observed difference in means
sigma <- 2.05821742280488                     # 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(1027871)
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 = 650 (80% power), 870 (90% power)
 75% of observed effect: n per arm = 1,155 (80% power), 1,546 (90% power)
 50% of observed effect: n per arm = 2,598 (80% power), 3,478 (90% power)
power.t.test: 650.4 
Simulated power at n = 650 per arm: 0.798 (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 sponsor’s reported primary analysis is a mixed-model analysis of LY2605541 Algorithm 2 versus insulin glargine. It reports a least-squares mean difference of 0.17, with a 90% confidence interval from -0.18 to 0.52 and p = 0.433. This is the registry’s own model-based result and should take precedence when describing the sponsor’s analysis.

The recalculated comparison for the same groups uses the posted endpoint summaries without the sponsor’s full repeated-measures model. It gives a mean difference of 0.01 and an unadjusted p-value of 0.974. The difference between the estimates reflects that the recalculation and reported analysis are not the same statistical procedure. The recalculation should be read as a simple check on summary statistics, not as a correction to the sponsor’s analysis.

The registry record does not report a sponsor-model comparison for Algorithm 1 versus insulin glargine. The unadjusted recalculation for that comparison is not a substitute for a missing model-based result. No discrepancy flag is recorded for the posted calculations, but that does not make the recalculations equivalent to the sponsor’s model.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists 21 outcome measures and 54 analyses. The posted results include a primary outcome and multiple secondary outcomes, creating the possibility of many statistical comparisons. The registry record does not report a multiplicity strategy, such as a hierarchy of hypotheses or an adjustment to account for testing multiple outcomes. Without that information, the reported p-values should not be interpreted as if each were the only analysis performed.

The record describes an MMRM for the primary endpoint and names fixed effects and a participant random effect. It also describes the primary analysis population as randomized participants who received at least one dose and had both a non-missing baseline value and at least one non-missing post-baseline value. This definition makes clear that the primary analysis did not necessarily include everyone enrolled. The registry record does not report the number excluded for each reason or provide a detailed account of missing-data assumptions and sensitivity analyses.

General: repeated-measures models can use available outcome measurements under assumptions about the missingness process. The validity of an estimate depends partly on whether those assumptions are reasonable. A model description alone does not establish how missing observations were handled in every circumstance, and the record does not report a complete missing-data analysis.

The endpoint-specific numbers analyzed vary. For example, the primary group counts are 89, 86, and 91, while the secondary analyses have their own group counts. The registry record does not explain every difference in analysis population or whether all deviations from the intended analysis were prespecified. It also does not report a full statistical analysis plan, details of any interim analyses, or the original sample-size justification. The primary endpoint model description mentions dose conversion before and after interim analysis as a factor; details of any interim analysis are not reported in the registry record.

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 2% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: angina unstable (1.1%), coronary artery disease (1.1%), myocardial infarction (1.0%), cholecystitis (1.0%), foreign body (1.0%).

The serious adverse event figures are counts of affected participants alongside the number at risk in each safety group: 4 of 98 for LY2605541 Dosing Algorithm 1, 1 of 98 for LY2605541 Dosing Algorithm 2, and 2 of 93 for insulin glargine. The record also reports counts for other adverse events: 44 of 98, 47 of 98, and 45 of 93, respectively. These are descriptive figures and do not establish that an event was caused by a treatment.

The registry lists serious adverse event terms including unstable angina, coronary artery disease, myocardial infarction, cholecystitis, and foreign body, each with a recorded percentage of 1.1% or 1.0%. It does not provide additional detail about the circumstances or relatedness of these terms in the registry record. Death counts are not reported in the registry record; a missing death count should not be read as zero.

Safety counts can differ from efficacy analysis counts because the outcomes may use different analysis populations or denominators. The record’s safety denominators should therefore be read as reported, rather than substituted for the primary endpoint group counts.

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

The primary result is a comparison of model-based fasting blood glucose estimates at Week 12, alongside unadjusted recalculations from posted summaries. For Algorithm 2 versus insulin glargine, the sponsor-reported difference is 0.17, with a 90% confidence interval from -0.18 to 0.52 and p = 0.433. For Algorithm 1 versus insulin glargine, the recalculated unadjusted difference is 0.32 mmol/L, with an unadjusted p-value of 0.304. The intervals and tests do not establish a clear difference, and lack of a statistically significant result does not establish equivalence.

The result does not determine whether either treatment is clinically preferable. The registry record does not state a threshold for clinical importance, and the statistical comparisons alone do not address every outcome relevant to treatment decisions. The trial was a phase 2 study, and its posted primary outcome focuses on a measured fasting glucose value at Week 12. Longer-term outcomes are not reported in the record.

General: statistical evidence concerns the compatibility of the observed data with an analysis model and its assumptions. It does not, by itself, establish causation, clinical importance, or how a result applies to people outside the trial’s enrolled population. Interpretation also depends on design features, outcome measurement, missing data, and the analysis plan.

The secondary results provide additional measurements, but should not be treated as independent confirmation without knowing how multiplicity was handled. The safety counts are descriptive and do not establish comparative risk or causality. The registry record does not report all details needed to evaluate missing-data assumptions, the original sample-size rationale, or the full analysis plan. These limitations matter when considering how far the posted estimates can be generalized.

18. Why This Trial Matters Statistically

This trial illustrates why an endpoint estimate, an uncertainty interval, and a model description should be read together. The sponsor’s primary analysis is a mixed-model repeated-measures comparison, while the recalculated comparisons are unadjusted summaries. Although both use the posted results, they answer the comparison through different statistical approaches. Reporting which estimate comes from the registry’s model and which is a simple recalculation prevents them from being mistaken for interchangeable results.

The primary analysis also shows why enrollment and analysis counts are not the same quantity. The trial enrolled 289 participants, while the primary endpoint population required a dose, baseline data, and post-baseline data. The model’s least-squares means and standard errors reflect an analysis of the eligible outcome data, not simply the enrolled total.

The planning illustrations demonstrate how strongly sample-size calculations depend on the effect chosen for planning. A smaller target effect requires more participants per arm to achieve a given power under the stated assumptions. Because the calculations use observed effects and are not the trial’s original design justification, they are useful for understanding precision and planning sensitivity, not for evaluating the trial’s design as if they had been specified in advance.

Finally, the registry lists many outcome measures and analyses, but does not report how multiplicity was handled. This is a central statistical issue: evidence from multiple outcomes should be interpreted in the context of the full set of tests, not by selecting isolated results. The record provides useful model and outcome details, while leaving important planning and analysis information unreported.

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 study was randomized and parallel-group, but the registry records no masking. The record does not provide further details about how randomization was implemented or whether other measures were used to reduce bias. The short 12-week treatment description and Week 12 primary endpoint limit what can be concluded about outcomes beyond that time point; longer-term results are not reported in the registry record.

The primary endpoint analysis population required participants to have baseline and post-baseline measurements after taking at least one dose. The record does not give a detailed accounting of missing observations or explain the assumptions used for all missing-data patterns. Group counts also vary across outcomes, and the reasons for those differences are not reported in full.

The registry describes the primary MMRM terms but does not report all model details, diagnostics, or sensitivity analyses. The sponsor-reported primary comparison is available for Algorithm 2 versus insulin glargine, but a corresponding sponsor-model result for Algorithm 1 is not reported in the registry record. The unadjusted recalculation for Algorithm 1 cannot fill that gap.

The record lists 21 outcome measures and 54 analyses but does not report a multiplicity procedure or the complete statistical analysis plan. It also does not report the original sample-size justification. The planning illustrations based on observed effects cannot answer whether the trial was designed with adequate precision for a prespecified clinically important difference.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study for Patients With Type 2 Diabetes — NCT01027871.
• Registry record last updated 2018-06-08; status last verified 2018-05.
• Results first posted 2018-06-08.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

NCT01027871 is a completed phase 2 randomized, parallel-group study sponsored by Eli Lilly and Company. It enrolled 289 people with type 2 diabetes and compared two LY2605541 dosing algorithms with insulin glargine. The brief study summary describes 12 weeks of treatment, and the primary endpoint was Week 12 fasting blood glucose from 8-point self-monitored blood glucose profiles.

The primary endpoint was analyzed with a mixed-model repeated-measures approach. The sponsor-reported comparison for LY2605541 Algorithm 2 versus insulin glargine was 0.17, with a 90% confidence interval from -0.18 to 0.52 and p = 0.433. Other posted results include secondary glycemic and hypoglycemia outcomes, as well as adverse-event counts.

The registry record does not report the original sample-size justification, a multiplicity strategy, or full details of missing-data handling and the statistical analysis plan.