A Study of Dulaglutide (LY2189265) in Chinese Participants With Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Dulaglutide (LY2189265) in Chinese Participants With Type 2 Diabetes (NCT04591626), a randomized phase 3 study. The primary outcome is Change From Baseline in Hemoglobin A1c (HbA1c).

2. Clinical Question

The clinical question concerns the additional effect of assigning dulaglutide rather than placebo while participants were also receiving basal insulin and, where used, metformin and/or acarbose. It is therefore a comparison of treatment strategies on a shared background regimen, not a comparison of dulaglutide with no diabetes treatment. Insulin glargine was titrated toward a target in both groups, according to the arm descriptions.

The main outcome was the change in HbA1c, a measure used to reflect average plasma glucose over a prolonged period. The posted primary analysis compares model-based mean changes at Week 28. The registry also lists secondary questions concerning achievement of HbA1c thresholds, body weight, fasting serum glucose, glucose measurements, and insulin glargine dose. Those outcomes add context but do not replace the primary comparison.

3. Trial Design

The registry describes a Phase 3, randomized, double-blind, parallel-group trial that is completed. Enrollment was 291. The study began on 2020-12-07, and primary completion was 2022-04-28. The record identifies Eli Lilly and Company as the sponsor.

Participants were assigned to weekly subcutaneous dulaglutide 1.5 mg or weekly subcutaneous placebo. In both groups, the assigned study treatment was added to titrated treat-to-target insulin glargine, with metformin and/or acarbose. This common background treatment is important to the interpretation: the comparison concerns adding dulaglutide versus adding placebo within the treatment setting described in the registry.

Random assignment is intended to make treatment groups comparable on average, including for factors that were not measured. Double masking is intended to reduce differences in care or assessment that could arise when treatment assignment is known. The parallel-group arrangement means participants were assigned to one treatment group rather than receiving both study assignments in sequence. These are general design purposes; the registry record does not describe implementation details such as the randomization procedure or how masking was maintained.

The primary outcome was assessed from baseline to Week 28. The primary analysis population is described as all randomized participants with baseline and at least one post-baseline HbA1c measurement. Outcome-specific group sizes for the primary comparison were 133 for dulaglutide and 142 for placebo. The registry record does not report the original sample-size justification or the trial's full protocol-level statistical plan.

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 with baseline and at least one post-baseline HbA1c data.

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.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in Hemoglobin A1c (HbA1c), measured over Baseline, Week 28.

The primary endpoint is change from baseline in hemoglobin A1c (HbA1c), measured at baseline and Week 28. HbA1c is the glycosylated fraction of hemoglobin A; the registry describes it as a measure used to identify average plasma glucose concentration over a prolonged period. The endpoint therefore captures change over time in a longer-term glycemic measure rather than a single glucose reading.

The reported group values are least-squares mean changes, estimated using a mixed model for repeated measures. The registry lists baseline, oral antihyperglycemic medication use, treatment, visit, and the treatment-by-visit interaction as model variables, with Type III sums of squares. Least-squares means are model-based group estimates; they are not necessarily identical to simple averages of observed participant changes.

Change from baseline makes the comparison about movement during the study rather than only the value at Week 28. The between-group estimate then expresses the difference in those modeled changes. A negative estimate for dulaglutide versus placebo indicates a more negative modeled change in the dulaglutide group under the reported contrast. The endpoint is useful for comparing groups, but it does not by itself describe each participant's experience or establish what happened after Week 28.

7. Results Posted in the Registry

• 1.5 mg Dulaglutide: mean -2.03 versus -1.08 on control; difference -0.95 (95% CI -1.16 to -0.74), standardized effect d = -1.09; z-test p = <0.001.

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.

Percentage of Participants Achieving HbA1c <7.0%

• 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:
  • 1.5 mg Dulaglutide: 75.9% (101/133) versus 33.8% (48/142) on control; risk difference 42.1 percentage points (95% CI 31.5 to 52.8), risk ratio 2.25 (1.75–2.88); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for 1.5 mg Dulaglutide than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Body Weight

• 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:
  • 1.5 mg Dulaglutide: mean -0.76 versus 0.42 on control; difference -1.18 (95% CI -1.79 to -0.57), standardized effect d = -0.46; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for 1.5 mg Dulaglutide than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Fasting Serum Glucose (FSG)

• 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:
  • 1.5 mg Dulaglutide: mean -58.47 versus -43.64 on control; difference -14.83 (95% CI -21.22 to -8.44), standardized effect d = -0.55; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for 1.5 mg Dulaglutide than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists secondary outcomes spanning glycemic thresholds, body weight, fasting serum glucose, glucose profiles, insulin dose, and combinations of glycemic, weight, and hypoglycemia criteria. This breadth means the secondary outcomes examine several related but distinct dimensions; they should not all be treated as alternate versions of the primary endpoint.

Among the secondary outcomes with numerical results described in the registry, the glycemic findings generally point in the same direction as the primary HbA1c comparison. The reported body-weight outcome addresses a separate dimension and does not measure glycemic control. Taken together, these outcomes provide context around the primary result, but they do not replace it or establish that every listed secondary objective had the same pattern. Numerical results for all listed secondary outcomes are not reported in the registry record.

Secondary results typically carry less evidential weight when many outcomes are examined and the approach to multiplicity is not reported. In general, testing multiple outcomes increases the chance that at least one result appears statistically notable by chance if each is assessed without an adjustment or a prespecified hierarchy. The registry record does not report whether or how multiplicity was handled. Unadjusted secondary p-values should therefore be read as outcome-specific evidence, not as if they were the result of a single isolated test with no broader set of analyses.

The secondary analyses are also subject to their own precision, definitions, and analysis populations. For example, a threshold outcome depends on the chosen cutoff, while a continuous change outcome retains information across the measurement scale. Those distinctions affect interpretation. The secondary outcome list alone does not establish that every measure was collected, analyzed, or reported in the same way.

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

• Percentage of Participants Achieving HbA1c <7.0%
• Change From Baseline in Body Weight
• Change From Baseline in Fasting Serum Glucose (FSG)
• Percentage of Participants Achieving HbA1c <7.0% With no Weight Gain (<0.1 kg) and Without Documented Symptomatic Hypoglycemia (Blood Glucose <3.0 mmol/L)
• Percentage of Participants Achieving HbA1c <7.0% Without Documented Symptomatic Hypoglycemia (Blood Glucose <3.0 mmol/L)
• Percentage of Participants Achieving HbA1c <7.0% Without Weight Gain (<0.1 kg)
• Change From Baseline in Blood Glucose From Daily Self-Monitored Blood Glucose (SMBG) Profile
• Change From Baseline in Daily Mean Insulin Glargine Doses

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is continuous: each participant's HbA1c change is measured on a numeric scale, and the analysis estimates group means and a difference between those means. The reported parameter is a least-squares mean, also called a model-adjusted mean. In general, this is a mean estimated from a statistical model that includes specified variables; it is not necessarily the same as an unadjusted arithmetic average.

How the repeated-measures model contributes

The registry identifies a mixed model for repeated measures (MMRM) for the primary analysis. In general, repeated-measures models use observations collected across visits while accounting for the fact that measurements from the same participant are related. This can use information from more than one post-baseline visit rather than treating each observation as unrelated. The record lists baseline, oral antihyperglycemic medication use, treatment, visit, and treatment-by-visit interaction as model variables, with Type III sums of squares. The registry record does not provide further detail about the covariance structure, estimation procedure, or model diagnostics.

The treatment-by-visit interaction allows the modeled treatment contrast to vary by visit. The primary timeframe includes baseline and Week 28, and the reported comparison concerns change to Week 28. Type III sums of squares are a model-based way to assess terms in the stated model; their inclusion does not, on its own, explain every implementation choice or establish that every possible source of bias has been addressed.

What the difference and standard error describe

The least-squares mean difference compares the modeled changes between groups. For the primary outcome, the sponsor-reported difference is -0.95 percentage points. The negative sign reflects the direction of the contrast as reported: the modeled change was more negative for dulaglutide than for placebo. The standard error describes the estimated sampling uncertainty of a model estimate. It is not the spread of individual participant outcomes and does not indicate how much participants varied from one another.

In general, a confidence interval combines an estimate with its uncertainty to give a range of values compatible with the data and the statistical model at a stated confidence level. A p-value measures how unusual data at least as incompatible with a specified null hypothesis would be under that hypothesis and the analysis assumptions. Neither statistic gives the probability that a hypothesis is true. The registry reports the primary p-value as p < 0.001.

Reported analysis and simple re-calculation

The registry's primary analysis is a mixed-model analysis, reporting a least-squares mean difference and its confidence interval. A separate simple, unadjusted re-calculation uses posted summary statistics and approximations; it is not the same as refitting the sponsor's repeated-measures model. The re-calculated mean difference is -0.95, in agreement with the sponsor's reported point estimate. Its interval is close to, but not identical with, the sponsor-reported interval. The registry record marks no discrepancy between these analyses, and the sponsor's model-based result remains the relevant reported analysis.

For some continuous outcomes, a pooled standard deviation and standardized effect are also calculated from summary information. Such calculations are descriptive approximations, not additional trial model results. In general, standardizing a mean difference expresses it relative to a measure of outcome variability, which can aid comparisons across scales; it does not change the original clinical units or resolve limitations in the underlying data.

Binary outcomes and other outcome types

Some secondary endpoints are percentages meeting a defined criterion. In general, a risk difference compares the proportions in two groups, while an odds ratio compares their odds. These quantities are not interchangeable: an odds ratio is not a percentage-point difference. The registry reports a logistic-regression odds ratio for the HbA1c threshold outcome. Its reported adjusted analysis and a simple contrast of percentages answer related, but not identical, statistical questions.

Other listed secondary outcomes include body-weight change, fasting serum glucose, daily self-monitored glucose profile, daily mean insulin glargine dose, and combinations of glycemic thresholds with weight or symptomatic hypoglycemia criteria. The record lists these outcome questions, but detailed numerical results for all of them are not included in the numerical results described in the record. No conclusions about unreported values can be drawn from the outcome titles alone.

In general, over-dispersion in a count or binary model means that variability is greater than the model's assumed distribution allows. It can make standard errors too small if ignored. The registry entries do not identify an over-dispersion adjustment, and the primary HbA1c outcome is continuous rather than a count outcome.

11. Confidence Intervals and Precision

The sponsor-reported primary difference is -0.95 percentage points, with a 95% confidence interval from -1.14 to -0.77. The interval lies entirely below zero and is relatively concentrated around the reported estimate, indicating that the model's estimated direction is not near the null value of no between-group difference. The interval communicates uncertainty about the average contrast, not a range containing a stated proportion of individual participant responses.

In general, a 95% confidence interval is generated by a procedure that, over repeated comparable studies and under its assumptions, would cover the true parameter in 95% of intervals. It is not correct to interpret an individual interval as assigning a 95% probability to the true effect. The interval's width reflects both the amount of information and the variability represented in the analysis. Precision is not the same as freedom from bias: an estimate can be precise while still depending on the appropriateness of the model and the quality of the data.

The arm-level standard errors reported for the primary least-squares means are 0.076 for dulaglutide and 0.073 for placebo. These describe uncertainty in the estimated group means, rather than participant-to-participant variability. The sponsor's reported interval comes from its mixed-model analysis; a simple re-calculation from posted summaries gives a close but not identical interval because it uses a different, unadjusted approximation.

Secondary outcomes have their own estimates and uncertainty, and their precision differs by outcome. A statistically clear primary comparison does not guarantee equally precise estimates for every secondary measure. The registry record does not report a clinical importance threshold for the primary difference, so the confidence interval alone cannot determine whether every value within it would be considered important in a clinical context.

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 sample-size illustrations give participants per arm for the continuous and binary outcomes discussed in the calculations; they are not total event counts. They are planning illustrations based on observed effects, not the original trial's design calculation. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect equal to the stated fraction of the observed effect. The trial's original design assumptions and sample-size justification are not reported in the registry record.

For each stated effect size, the 80% power column gives the estimated number of participants per arm for a design with an 80% chance of detecting that effect under the listed assumptions, if the effect is present and the model assumptions apply. The 90% power column describes the corresponding planning target for a 90% chance. These are prospective design concepts, not probabilities that the completed study has or has not detected a true effect. The values are not observed power.

The effect scenarios are labeled as 100%, 75%, and 50% of the observed effect. As the assumed effect becomes smaller, more participants are generally needed to distinguish it from random variation at the same power and significance level. The estimates therefore illustrate how sensitive planning can be to the effect size chosen. They do not establish that a future study would observe the same effect, and they should not be read as a description of the actual allocation or analysis in this trial. The original design's assumptions are not reported in the registry record.

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 = -0.950; pooled standard deviation σ = 0.873.
2. n = 2 × (1.96 + 0.842)² × 0.873² / -0.950² = 13.3, rounded up to 14 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 <- -0.95                      # observed difference in means
sigma <- 0.873082816458785                     # 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(4591626)
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 = 14 (80% power), 18 (90% power)
 75% of observed effect: n per arm = 24 (80% power), 32 (90% power)
 50% of observed effect: n per arm = 54 (80% power), 71 (90% power)
power.t.test: 14.3 
Simulated power at n = 14 per arm: 0.790 (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 sponsor-reported primary analysis is a mixed-model comparison of least-squares mean changes. It gives a difference of -0.95 percentage points, a 95% confidence interval from -1.14 to -0.77, and p < 0.001. This model-based result is the main reported estimate to use when describing the primary outcome.

A simple re-calculation from the posted group summaries also gives a mean difference of -0.95. Its interval is close to, but not identical with, the sponsor's interval. The re-calculation is unadjusted and uses approximations from summary statistics, whereas the sponsor's analysis uses the stated mixed model. Agreement in the point estimate does not make the methods interchangeable. The registry record reports no discrepancy, so the results do not call for a discrepancy warning; the sponsor's analysis remains the trial's own analysis.

For the HbA1c threshold secondary outcome, the sponsor reports a logistic-regression odds ratio of 10.91, with a 95% confidence interval from 5.35 to 22.28 and p < 0.001. For body-weight change and fasting serum glucose, the sponsor reports mixed-model differences. Those adjusted analyses should be distinguished from simple comparisons of group percentages or summary means, which may not use the same model or adjustment.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies one primary HbA1c endpoint and lists eight secondary outcomes. It reports a primary mixed-model analysis and numerical results for selected secondary outcomes. The registry record does not report a multiplicity adjustment, a hierarchy for testing secondary outcomes, or an overall strategy for controlling false-positive findings across outcomes. Without that information, individual secondary p-values should not be interpreted as though the entire family of outcome tests had been accounted for.

In general, multiplicity refers to making several statistical comparisons or tests. If many tests are conducted, the probability of at least one small p-value arising by chance can increase. A prespecified adjustment or testing sequence can address that issue, but the specific approach used in this trial is not reported in the registry record. The two-sided significance level stated for the planning illustrations is not evidence of the trial's original analysis threshold.

The primary analysis population is described as all randomized participants with baseline and at least one post-baseline HbA1c data. The registry record does not report how missing HbA1c data were handled in the model, whether additional sensitivity analyses were performed, or whether alternative analysis populations were used. The population description therefore should not be expanded into an assumption that every randomized participant contributed a Week 28 measurement.

The detailed statistical plan, including model diagnostics, missing-data assumptions, and sensitivity analyses, is not reported in the registry record. These omissions limit the ability to assess how dependent the estimated contrast is on assumptions about incomplete measurements. The results remain interpretable as the posted analysis, but the record does not provide enough detail to independently evaluate every analysis choice.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 1% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: chronic gastritis (1.4%), abdominal adhesions (0.7%), duodenal ulcer haemorrhage (0.7%), large intestine polyp (0.7%), cholecystitis acute (0.7%).

The registry reports serious adverse events among 144 participants at risk in the dulaglutide group and 147 at risk in the placebo group. Eight serious adverse events were reported in the dulaglutide group and 12 in the placebo group. These are counts of reported serious events, not measures of treatment effect by themselves; the record does not provide a causal assessment for each event.

Other adverse events were reported for 113 participants at risk in the dulaglutide group and 99 in the placebo group. The record also reports zero deaths in the dulaglutide group and one in the placebo group. These figures should be read in their stated categories and denominators. A count of participants with an event is not necessarily a count of all event episodes, and the registry record does not provide enough detail to infer severity, timing, or causal relation for each non-serious event.

The serious-event terms listed include chronic gastritis at 1.4%, and abdominal adhesions, duodenal ulcer haemorrhage, large intestine polyp, acute cholecystitis, and cholelithiasis at 0.7% each. These listed terms are descriptive reports; they do not establish that the study treatment caused them. Safety counts in this record should be interpreted as reported observations within the trial, not as a complete assessment of risks in other populations or over other time periods.

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

The primary result describes a model-based difference in HbA1c change between the randomized groups through Week 28, under the background treatment approach described in the registry. The estimate is negative for dulaglutide versus placebo, and its sponsor-reported confidence interval remains below zero. This is evidence of a difference in the measured group means in this trial, with uncertainty represented by the interval.

The result does not say that every participant experienced the group-average change. Group means conceal variation among individuals, and the standard errors are not individual outcome ranges. The result also does not by itself establish the clinical importance of the difference. The registry record does not report a threshold for a clinically meaningful change, and a statistical comparison cannot supply that judgment on its own.

The comparison concerns adding dulaglutide rather than placebo to titrated insulin glargine with metformin and/or acarbose. It should not be interpreted as a comparison with no treatment or as an estimate of dulaglutide used alone. The record describes Chinese participants with type 2 diabetes; it does not establish that the same estimates apply to populations or treatment settings not represented in the trial.

The results are tied to the stated Week 28 assessment and the outcomes reported. They do not establish longer-term outcomes, effects on unreported outcomes, or how the findings would change under a different insulin titration or background-treatment approach. Nor do the adverse-event counts alone identify which events were caused by treatment. The registry provides useful group comparisons, but the limits of the population, follow-up, outcome reporting, and statistical detail remain relevant to interpretation.

18. Why This Trial Matters Statistically

This trial is statistically informative because it compares randomized groups on a prespecified primary measure using a model that accounts for repeated observations. The primary question is framed as a between-group difference in change, which is more directly interpretable than considering either group's change in isolation. The model-based estimate, its confidence interval, and the reported p-value communicate different parts of the result: direction and magnitude, uncertainty, and evidence against a null comparison under the model.

The study also illustrates why model-based results and simple re-calculations should be kept distinct. The sponsor reports a mixed-model analysis, while summary-statistic re-calculations rely on simplifying assumptions. Similar point estimates can coexist with differences in intervals because the methods use information differently. That distinction matters when readers assess what an analysis actually estimates.

Several secondary outcomes extend the statistical questions beyond HbA1c, but multiple outcomes create an interpretive challenge when multiplicity handling is not reported. The trial record also makes the analysis population explicit for the primary endpoint, while leaving important details about missing-data handling and the full statistical plan unreported. These are examples of why a result's numerical strength is only one part of an evidence assessment.

Finally, the planning illustrations demonstrate how sample-size needs depend on the effect targeted and the chosen design assumptions. They are not a retrospective measure of whether the completed study was adequately powered. Keeping planning calculations separate from observed results helps avoid treating a study's outcome as proof that its design had a particular probability of success.

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 population and treatment setting constrain interpretation. The registry describes Chinese participants with type 2 diabetes receiving titrated insulin glargine with metformin and/or acarbose. The record does not establish transportability to other populations, background regimens, or approaches to insulin titration.

The primary outcome is reported at Week 28. The registry record does not provide longer-term primary outcome results in the material, so conclusions should remain tied to that assessment period. The record also lists multiple secondary outcomes but does not provide numerical results for every listed outcome in the results. Selective availability of numerical detail limits a complete assessment of the full secondary-outcome pattern.

Important analysis details are not reported in the registry record, including the trial's original sample-size justification, multiplicity handling, missing-data method, full model specifications, and sensitivity analyses. The stated planning assumptions belong to illustrative calculations and do not fill in the trial's original design rationale. Without the missing details, independent assessment of those aspects is limited.

The reported adverse-event counts are descriptive and do not, by themselves, establish causation or provide a full comparison of event timing and severity. The registry record reports serious and other adverse-event counts, but a fuller safety interpretation requires details beyond these totals. Finally, statistical evidence for a difference does not itself determine clinical importance; a relevant threshold is not reported in the registry record.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of Dulaglutide (LY2189265) in Chinese Participants With Type 2 Diabetes — NCT04591626.
• Registry record last updated 2023-05-24; status last verified 2023-04.
• Results first posted 2023-05-24.
• 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

The registry describes a completed Phase 3, randomized, double-blind, parallel-group trial in Chinese participants with type 2 diabetes. It compared weekly dulaglutide 1.5 mg with weekly placebo, each added to titrated treat-to-target insulin glargine with metformin and/or acarbose. Enrollment was 291.

The primary endpoint was change in HbA1c from baseline to Week 28. The sponsor-reported mixed-model difference was -0.95 percentage points, with a 95% confidence interval from -1.14 to -0.77 and p < 0.001. The primary analysis population included randomized participants with baseline and at least one post-baseline HbA1c result.

The record lists eight secondary outcomes and reports selected numerical analyses, alongside serious and other adverse-event counts. Details on multiplicity, missing-data handling, and the original sample-size justification are not reported in the registry record.