This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for A Study to Evaluate the Efficacy and Safety of MEDI0382 in the Treatment of Overweight and Obese Subjects With Type 2 Diabetes (NCT03235050), a randomized phase 2 study. The primary outcome is Change in HbA1c.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT03235050 |
| Phase | Phase 2 |
| Condition | Diabetes Mellitus, Type 2 |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Triple |
| Enrollment | 834 |
| Arms in the primary outcome | 4 |
| Lead sponsor | AstraZeneca |
| Results posted | Yes |
2. Clinical Question
Population
The trial concerned people with type 2 diabetes who were overweight or obese. The primary analysis is identified as the ITT population. The registry record does not report the detailed eligibility criteria or the operational definition used for the ITT population.
Intervention
The primary question concerned cotadutide, also called MEDI0382 in the trial title, at 100, 200, and 300 mcg. Trial arms paired MEDI0382 with metformin; the arm descriptions specify a total daily metformin dose of ≥1500 mg unless only a lower dose was tolerated.
Comparator
The primary comparisons were each MEDI0382 dose versus placebo. Placebo was also paired with metformin. Liraglutide plus metformin was included as an active-comparator arm, but the posted primary analyses compare MEDI0382 doses with placebo.
Primary question
How did HbA1c change from baseline to 14 weeks in the three MEDI0382 dose groups compared with placebo? HbA1c is a measure of blood glucose over time. The registry reports least-squares mean changes and ANCOVA estimates of the between-group differences, with 95% confidence intervals and p-values.
This question is a comparison of average change over a specified follow-up period, rather than a comparison of each participant’s result with a threshold. The registry lists other outcomes, including HbA1c target attainment, body-weight change, rescue or discontinuation for lack of glycaemic control, and safety. Those outcomes provide additional perspectives but are not the primary endpoint considered in the estimates. The registry record does not report the detailed eligibility criteria, the rationale for selecting the doses, or a prespecified threshold for a clinically important HbA1c difference.
3. Trial Design
This was a phase IIb, randomized, parallel-group trial described as double-blind and placebo-controlled, with an open-label active-comparator component. The registry also classifies masking as triple. It does not identify which groups of people were masked, or explain how masking was implemented across the blinded and open-label arms. The study was sponsored by AstraZeneca and is recorded as completed.
The registry lists an enrollment of 834. The primary endpoint was assessed from baseline to 14 weeks. Five intervention arms are described: MEDI0382 at low, mid, and high doses, placebo, and liraglutide. For each arm, the description includes metformin tablets at a total daily dose of ≥1500 mg unless only a lower dose was tolerated. The primary outcome description names 100, 200, and 300 μg of cotadutide versus placebo, while the posted arm and analysis labels use mcg. Those dose labels are reproduced as recorded.
The primary outcome is reported for the ITT population. The registry record does not describe the randomization sequence, allocation concealment, site structure, or the number of participants assigned to each arm at randomization. It also does not give details of the open-label procedures or specify how the active-comparator arm was incorporated into the main analysis. These details should not be inferred from the design label alone.
The registry lists 13 outcome measures and 83 analyses, and identifies 11 secondary outcomes as posted. That breadth indicates that the primary HbA1c comparison is only one part of the study’s reported outcome set. The registry record does not report a protocol-level statistical analysis plan, a sample-size justification, or whether interim analyses were planned or conducted. The enrollment figure therefore describes the registered trial, not the reasoning behind its planned size.
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
| Stage | Placebo | Liraglutide | MEDI0382 100 mcg | MEDI0382 200 mcg | MEDI0382 300 mcg |
|---|---|---|---|---|---|
| STARTED | 112 | 110 | 100 | 256 | 256 |
| COMPLETED | 108 | 105 | 96 | 242 | 245 |
| NOT COMPLETED | 4 | 5 | 4 | 14 | 11 |
Reasons for not completing this period:
| Reason | Placebo | Liraglutide | MEDI0382 100 mcg | MEDI0382 200 mcg | MEDI0382 300 mcg |
|---|---|---|---|---|---|
| Adverse Event | 0 | 0 | 1 | 4 | 1 |
| Death | 0 | 0 | 0 | 2 | 1 |
| Lost to Follow-up | 1 | 2 | 1 | 2 | 0 |
| Withdrawal by Subject | 3 | 2 | 2 | 5 | 7 |
| Other reasons | 0 | 1 | 0 | 1 | 2 |
Analysis population for the primary outcome, as defined in the registry: ITT population
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.
| Characteristic | Unit | Placebo | Liraglutide | MEDI0382 100 mcg | MEDI0382 200 mcg | MEDI0382 300 mcg | Total |
|---|---|---|---|---|---|---|---|
| Participants analysed | 112 | 110 | 100 | 256 | 256 | 834 | |
| Age, Continuous | Years | 57.3 (SD 9.5) | 55.5 (SD 9.8) | 57.6 (SD 9.9) | 57.3 (SD 9.9) | 56.3 (SD 10.2) | 56.8 (SD 9.9) |
| Sex: Female, Male: Female | Participants | 55 | 60 | 57 | 147 | 129 | 448 |
| Sex: Female, Male: Male | Participants | 57 | 50 | 43 | 109 | 127 | 386 |
| Race/Ethnicity, Customized: American Indian or Alaska Native | Participants | 3 | 3 | 0 | 5 | 0 | 11 |
| Race/Ethnicity, Customized: Asian | Participants | 1 | 1 | 0 | 3 | 1 | 6 |
| Race/Ethnicity, Customized: Black or African American | Participants | 0 | 3 | 1 | 3 | 3 | 10 |
| Race/Ethnicity, Customized: White | Participants | 107 | 103 | 99 | 245 | 252 | 806 |
| Race/Ethnicity, Customized: Other | Participants | 1 | 0 | 0 | 0 | 0 | 1 |
6. Primary Endpoint
The registry lists the primary outcome as Change in HbA1c, measured over From baseline to 14 weeks.
| Element | Registry entry |
|---|---|
| Result type | Least Squares Mean |
| Measure of spread | 95% Confidence Interval |
| Unit | percentage |
| Endpoint classification | Continuous Endpoint |
The primary endpoint is change in HbA1c from baseline to 14 weeks, expressed as a percentage. HbA1c reflects blood glucose over a period of time and is commonly used in diabetes trials as a continuous measure of glycaemic control. A continuous endpoint retains information about the degree of change, rather than reducing results to whether a participant crossed a single threshold.
The registry reports least-squares mean change for each group and identifies the dispersion as a 95% confidence interval. The group estimates are adjusted-model summaries, rather than simple arithmetic averages. The registry describes the primary population as ITT and lists ANCOVA for the comparisons between each MEDI0382 dose and placebo.
General statistical context: a confidence interval describes the uncertainty around an estimated quantity under the assumptions of the analysis. It is not a range containing every participant’s result, and it does not show how much individual responses varied. The endpoint is useful for comparing average changes between randomized groups, but an average alone cannot describe the full distribution of responses or establish whether a difference is important to patients. The registry record does not report a threshold for clinical importance or explain how missing endpoint measurements were handled.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
All four posted group estimates are reductions in HbA1c from baseline. The least-squares mean change was -0.18 for placebo, with a 95% confidence interval from -0.34 to -0.02. For MEDI0382 100 mcg, the estimate was -1.01, with a 95% confidence interval from -1.18 to -0.84. For 200 mcg, it was -1.22, with a 95% confidence interval from -1.33 to -1.11. For 300 mcg, it was -1.09, with a 95% confidence interval from -1.2 to -0.98.
The sponsor-reported ANCOVA comparisons express the difference between each MEDI0382 dose and placebo. The least-squares mean difference was -0.83 for 100 mcg, with a 95% confidence interval from -1.06 to -0.59 and p < 0.001. For 200 mcg, it was -1.04, with a 95% confidence interval from -1.23 to -0.85 and p < 0.001. For 300 mcg, it was -0.91, with a 95% confidence interval from -1.11 to -0.72 and p < 0.001. The negative signs indicate that the estimated change was lower in the MEDI0382 group than in the placebo group under the comparison’s coding.
Each reported confidence interval for the between-group estimate lies below zero, and each reported p-value is less than 0.001. These statistics describe the registered ANCOVA comparisons; they do not establish the size or importance of an effect for any individual. The estimates also do not show a steadily increasing difference across the three dose labels: the posted estimate for 200 mcg is more negative than those for 100 mcg and 300 mcg. This observation is descriptive and is not a formal test of a dose trend.
The analysis concerns change through 14 weeks. It does not, by itself, describe persistence beyond that timeframe, the results of the active-comparator arm, or the other outcomes in the registry. The registry record does not report a clinical-importance threshold, detailed covariate adjustments, or missing-data methods. Statistical evidence in these comparisons should therefore be read alongside those reporting limits and the trial’s other outcomes, without treating a p-value as a measure of benefit magnitude or as a complete account of the study.
9. Statistical Methods Explained
The primary endpoint is continuous: HbA1c change from baseline to 14 weeks. The registry reports a least-squares mean for each arm, along with a 95% confidence interval, and names ANCOVA as the method for the dose-versus-placebo analyses. The following statistical explanations are general; where the registry does not describe a specific implementation, that limitation is stated explicitly.
What does a least-squares mean represent?
General statistical explanation: a least-squares mean is an estimate produced by a statistical model. In an ANCOVA, the estimate can account for variables included in the model, so it may differ from a raw arithmetic group average. The registry record does not report which covariates were included, how they were coded, or the detailed model specification. It is therefore possible to describe the posted estimates as model-based, but not to identify exactly what adjustments they incorporate.
The arm estimates and the between-arm estimates answer related but distinct questions. An arm’s least-squares mean describes its modeled change. A least-squares mean difference compares two modeled arm estimates within the sponsor’s analysis. The reported comparisons are against placebo, rather than against no change within each treatment group. The analysis estimates and confidence intervals are therefore the relevant quantities for interpreting the stated primary comparisons.
How should confidence intervals be read?
General statistical explanation: a confidence interval conveys uncertainty around an estimate under the chosen model and its assumptions. A narrower interval indicates greater precision than a wider interval for the same type of estimate, but it does not establish that the estimate is clinically important. A 95% confidence interval is not a statement that 95% of individual participant outcomes fall inside its limits. For the three posted comparisons, the confidence intervals are reported alongside the point estimates so readers can assess their uncertainty, not only their direction.
General statistical explanation: the p-value measures how compatible the observed data are with a specified null hypothesis under the analysis assumptions. A small p-value is not the probability that the null hypothesis is true, nor does it quantify the magnitude or clinical value of a result. The registry reports p < 0.001 for each of the three dose-versus-placebo comparisons. The exact test details beyond the ANCOVA label are not reported in the registry record.
What does ANCOVA contribute?
General statistical explanation: analysis of covariance compares outcomes across groups while including one or more covariates in a model. Depending on the model and its assumptions, this can improve precision or account for baseline differences in variables included in the model. Randomization supports comparison between assigned groups, but the analysis still depends on appropriate modeling and on how measurements and missing observations are handled. The registry record does not state the covariates, the model diagnostics, or whether any transformations or alternative analyses were used.
The registered design is randomized and parallel-group. General statistical explanation: random allocation is intended to make groups comparable on average at the start of a trial, including for characteristics not measured. It does not guarantee exact balance in every realized sample, and it does not remove the need to define how outcomes are analyzed. The registry record does not give the randomization method or allocation-concealment details, so no more specific assessment of those procedures can be made from this record.
How do the recalculated comparisons relate to the reported model?
A simple unadjusted recalculation from posted summary statistics is not the same estimand or procedure as an ANCOVA comparison. The recalculation does not reproduce model adjustments that are not available in the arm summaries. It can serve as a check on the scale and direction of the posted results, but it should not replace the sponsor’s analysis. No recalculated comparison values are stated in the registry record. The record does not flag a discrepancy between the sponsor’s analysis and a recalculation.
What do planning illustrations mean?
General statistical explanation: a sample-size planning illustration asks how many participants per arm might be needed under specified assumptions to detect a chosen effect with a chosen power. The registry’s planning assumptions are a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a stated fraction of the observed effect. Such illustrations are planning scenarios, not observed trial outcomes and not observed power. The registry record does not report the trial’s own sample-size justification or establish that its design used these illustration assumptions.
The primary endpoint is continuous, so the relevant planning interpretation is participants per arm for a specified power, where such a result is reported. The registry record does not report a planning value in the registry record for this analysis. No sample-size quantity should be inferred from the enrolled total. More generally, planning estimates depend on the assumed effect, variability, allocation, significance level, and analysis method; changing those assumptions changes the planning result.
10. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| ANCOVA | LS Mean Difference -0.83 | -1.06 to -0.59 | < 0.001 |
| ANCOVA | LS Mean Difference -1.04 | -1.23 to -0.85 | < 0.001 |
| ANCOVA | LS Mean Difference -0.91 | -1.11 to -0.72 | < 0.001 |
The sponsor’s own reported comparisons use ANCOVA and are presented as least-squares mean differences between each MEDI0382 dose and placebo. The corresponding estimates are -0.83 for 100 mcg, -1.04 for 200 mcg, and -0.91 for 300 mcg, each with a 95% confidence interval and p < 0.001. These model-based comparisons are not interchangeable with a simple unadjusted calculation from group summary statistics.
A recalculation from posted summary statistics is a separate, unadjusted check. It cannot reproduce adjustments made by an ANCOVA model when the model covariates and detailed specification are not reported. The registry record does not give recalculated comparison estimates in the material available for this analysis, so no numerical agreement or difference between the two approaches can be stated. It does not flag a discrepancy. If a recalculation and a sponsor analysis conflict, the sponsor’s own analysis takes precedence as the registered analysis, while the discrepancy should still be made clear to readers.
The registry record does not report enough detail to reproduce the sponsor’s model exactly. In particular, the covariates and the handling of missing outcomes are not reported in the registry record. The posted ANCOVA estimates should therefore be read as the sponsor-reported primary comparisons, with their stated uncertainty, rather than as a fully independently reproducible model specification.
11. Multiplicity, Missing Data and Analysis Populations
The registry lists 13 outcome measures, 83 analyses, and 11 secondary outcomes as posted. The primary endpoint includes three MEDI0382-versus-placebo comparisons. Multiple outcomes and comparisons create a multiplicity question: when many hypotheses are examined, the chance of at least one apparently notable result can increase if each is judged separately at the same threshold. General statistical explanation: approaches to multiplicity may adjust significance thresholds or focus inference on a prespecified hierarchy. The registry record does not report whether any such approach was planned or used, or how the primary comparisons were positioned relative to the secondary analyses.
The reported p-values and confidence intervals should consequently be interpreted as the results of the individual posted ANCOVA comparisons. The record does not state whether the confidence intervals were adjusted for multiplicity. It would be inappropriate to assume an adjustment, or to assume none, from the p-values alone. Nor does the number of analyses establish that every analysis represented a separate confirmatory hypothesis.
The primary population is identified as ITT. General statistical explanation: an intention-to-treat approach is generally intended to analyze participants according to randomized assignment, preserving the benefits of randomization. The registry record does not define its ITT population, describe any exclusions, or report the number of participants contributing to each endpoint estimate. The arm-specific denominators listed with the outcome estimates are not a substitute for a complete accounting of the analysis population and missing measurements.
Missing data can affect both the estimated group changes and their uncertainty. General statistical explanation: an analysis that uses only observed measurements relies on assumptions about why outcomes are missing, while imputation or model-based approaches also require assumptions. The registry record does not report the amount or pattern of missing primary endpoint data, the method used to handle it, or sensitivity analyses assessing alternative assumptions. The analysis population label alone does not resolve these questions.
The registry record also does not report a detailed statistical analysis plan, the sample-size justification, or whether interim analyses were planned or conducted. Those omissions limit assessment of prespecification, multiplicity control, and the relationship between the study’s intended design and the posted analyses. They do not change the numerical estimates, but they constrain how fully the analysis can be evaluated.
12. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Placebo | 12 | 112 | 10.7% |
| Liraglutide | 8 | 110 | 7.3% |
| MEDI0382 100 mcg | 12 | 100 | 12.0% |
| MEDI0382 200 mcg | 33 | 256 | 12.9% |
| MEDI0382 300 mcg | 20 | 256 | 7.8% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Placebo | 0 of 112 |
| Liraglutide | 0 of 110 |
| MEDI0382 100 mcg | 0 of 100 |
| MEDI0382 200 mcg | 2 of 256 |
| MEDI0382 300 mcg | 1 of 256 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Placebo | 45 of 112 |
| Liraglutide | 36 of 110 |
| MEDI0382 100 mcg | 51 of 100 |
| MEDI0382 200 mcg | 159 of 256 |
| MEDI0382 300 mcg | 164 of 256 |
Most frequent serious adverse event terms
| Event term | Organ system | Placebo | Liraglutide | MEDI0382 100 mcg | MEDI0382 200 mcg | MEDI0382 300 mcg |
|---|---|---|---|---|---|---|
| Angina unstable | Cardiac disorders | 0 of 112 | 0 of 110 | 1 of 100 | 2 of 256 | 1 of 256 |
| Osteoarthritis | Musculoskeletal and connective tissue disorders | 0 of 112 | 0 of 110 | 1 of 100 | 3 of 256 | 0 of 256 |
| Angina pectoris | Cardiac disorders | 0 of 112 | 0 of 110 | 0 of 100 | 2 of 256 | 1 of 256 |
| Atrial fibrillation | Cardiac disorders | 1 of 112 | 0 of 110 | 1 of 100 | 0 of 256 | 1 of 256 |
| Cholelithiasis | Hepatobiliary disorders | 0 of 112 | 0 of 110 | 1 of 100 | 2 of 256 | 0 of 256 |
| Pneumonia | Infections and infestations | 0 of 112 | 1 of 110 | 2 of 100 | 0 of 256 | 0 of 256 |
| Acute myocardial infarction | Cardiac disorders | 1 of 112 | 0 of 110 | 0 of 100 | 1 of 256 | 0 of 256 |
| Atrial flutter | Cardiac disorders | 1 of 112 | 0 of 110 | 0 of 100 | 1 of 256 | 0 of 256 |
| Coronary artery disease | Cardiac disorders | 0 of 112 | 0 of 110 | 1 of 100 | 1 of 256 | 0 of 256 |
| Myocardial infarction | Cardiac disorders | 0 of 112 | 0 of 110 | 0 of 100 | 1 of 256 | 1 of 256 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Placebo | Liraglutide | MEDI0382 100 mcg | MEDI0382 200 mcg | MEDI0382 300 mcg |
|---|---|---|---|---|---|---|
| Nausea | Gastrointestinal disorders | 12 of 112 | 17 of 110 | 23 of 100 | 85 of 256 | 105 of 256 |
| Vomiting | Gastrointestinal disorders | 5 of 112 | 3 of 110 | 10 of 100 | 51 of 256 | 43 of 256 |
| Diarrhoea | Gastrointestinal disorders | 10 of 112 | 4 of 110 | 13 of 100 | 49 of 256 | 28 of 256 |
| Nasopharyngitis | Infections and infestations | 15 of 112 | 11 of 110 | 11 of 100 | 22 of 256 | 28 of 256 |
| Dyspepsia | Gastrointestinal disorders | 2 of 112 | 3 of 110 | 7 of 100 | 19 of 256 | 28 of 256 |
| Headache | Nervous system disorders | 1 of 112 | 5 of 110 | 3 of 100 | 15 of 256 | 19 of 256 |
| Back pain | Musculoskeletal and connective tissue disorders | 6 of 112 | 2 of 110 | 8 of 100 | 9 of 256 | 9 of 256 |
| Dizziness | Nervous system disorders | 3 of 112 | 0 of 110 | 4 of 100 | 11 of 256 | 15 of 256 |
| Urinary tract infection | Infections and infestations | 6 of 112 | 2 of 110 | 3 of 100 | 13 of 256 | 7 of 256 |
| Decreased appetite | Metabolism and nutrition disorders | 1 of 112 | 1 of 110 | 3 of 100 | 12 of 256 | 14 of 256 |
Serious events with the highest arm-level rate: pneumonia (2.0%), osteoarthritis (1.2%), angina unstable (1.0%), atrial fibrillation (1.0%), coronary artery disease (1.0%).
The registry reports serious adverse events and other adverse events by arm, with counts of participants at risk and counts affected. For serious adverse events, the posted counts are 12 of 112 for placebo, 8 of 110 for liraglutide, 12 of 100 for MEDI0382 100 mcg, 33 of 256 for MEDI0382 200 mcg, and 20 of 256 for MEDI0382 300 mcg. The registry reports deaths of 0, 0, 0, 2, and 1 in those respective groups.
These are descriptive counts, not estimates of causal effects. Differences in group size and the number of participants affected should be considered when reading across arms; no risk comparison or statistical test for safety is reported in the registry record. The record also lists pneumonia at 2.0%, osteoarthritis at 1.2%, and angina unstable, atrial fibrillation, coronary artery disease, and gastric polyps at 1.0% each among the top serious-event terms. The registry record does not provide enough context in these figures to establish whether an event was caused by an intervention. They should be read as reported safety observations, not as proof of safety or harm.
13. What the Result Does, and Does Not, Mean
The posted primary results describe model-based mean HbA1c changes through 14 weeks and ANCOVA comparisons of MEDI0382 doses with placebo. Each reported dose-versus-placebo estimate is negative, and each reported 95% confidence interval lies below zero. The corresponding p-values are each p < 0.001. Together, these results indicate that the trial’s reported analyses found differences in average HbA1c change between each listed MEDI0382 dose and placebo over the stated period.
Those findings do not tell readers how much participants varied around the group averages, whether every participant experienced a change in the same direction, or how durable any change was after 14 weeks. The group confidence intervals quantify uncertainty in estimated averages under the analysis; they are not individual response ranges. The registry record does not provide a threshold for a clinically important difference, so statistical evidence alone cannot determine the clinical importance of the estimates.
The primary comparison also does not establish how MEDI0382 compared with liraglutide. Although liraglutide is listed as an active comparator, the posted primary analyses are comparisons against placebo. Nor do these estimates answer the secondary questions about weight, HbA1c target attainment, rescue or discontinuation, or the full safety profile. Those outcomes require their own reported estimates and methods.
The registry record does not report the detailed ANCOVA specification, missing-data handling, or multiplicity strategy. These omissions limit independent evaluation of model assumptions and the interpretation of results across the full set of analyses. The figures are best understood as the sponsor-reported estimates for the stated primary endpoint and comparisons, within the follow-up and reporting information available in the registry.
General statistical explanation: a trial estimate is conditional on its design, population, endpoint definition, follow-up, and analysis assumptions. It is not a guarantee of an individual outcome, and it cannot by itself settle questions that were not measured or reported. The findings should therefore be kept within the boundaries of this trial’s registered primary question.
14. Why This Trial Matters Statistically
This trial illustrates how a clinical study can report several related but distinct statistical quantities. The group-level least-squares means describe modeled HbA1c change, while the ANCOVA estimates address the between-group comparisons that define the primary question. Confidence intervals add information about precision, and p-values address compatibility with a statistical null under the analysis assumptions. None of these quantities alone describes every aspect of the outcome.
The record also makes visible the difference between an analysis result and its reproducibility. The ANCOVA method is named, but its covariates and detailed specification are not reported. A simple unadjusted recalculation from summaries is not a substitute for the sponsor’s model. Clear separation of those approaches helps prevent readers from treating unlike estimates as if they were the same calculation.
Multiplicity and missing-data reporting matter because the study includes numerous outcomes and analyses, while the registry record does not describe adjustment procedures or the handling of missing primary endpoint measurements. These details shape how confidently readers can interpret the reported comparisons and relate them to the broader set of outcomes. Their absence is a limit on interpretation, not evidence that a particular method was or was not used.
Finally, the registry includes both placebo and active-comparator arms, and reports safety counts alongside efficacy outcomes. Keeping the primary placebo comparisons distinct from the active-comparator question and from descriptive safety observations prevents over-reading one result as a complete account of the study. This is a useful setting for reading model-based estimates with attention to endpoint, timeframe, population, uncertainty, and reporting limits.
15. 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 record gives the primary endpoint timeframe as baseline to 14 weeks, so the estimates do not describe outcomes beyond that period. It identifies the primary population as ITT but does not define the population, detail exclusions, or state how many participants contributed measurements to each estimate. The registry record also does not report the number or pattern of missing endpoint observations, how missing data were handled, or sensitivity analyses.
ANCOVA is named for the comparisons, but the covariates, model specification, diagnostics, and any alternative analyses are not reported in the registry record. This makes the sponsor’s analysis difficult to reproduce from the record alone. The registry does not report a multiplicity strategy despite listing 13 outcome measures and 83 analyses, and it does not clarify whether the stated confidence intervals account for multiple comparisons.
The record does not report the trial’s sample-size justification, the assumptions behind the original design, or whether interim analyses were planned or conducted. It also does not describe randomization procedures or allocation concealment, identify the groups covered by triple masking, or explain the implementation of the open-label active-comparator component. These are reporting limitations and should not be filled in by inference.
The primary posted comparisons are against placebo; the registry record does not provide a primary estimate comparing MEDI0382 with liraglutide. No clinical-importance threshold is reported. Safety figures are descriptive counts, and the registry record does not provide a causal analysis of adverse events. These limits constrain the scope of conclusions that can be drawn from the posted primary result and safety summary.
16. Related Tutorials
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: A Study to Evaluate the Efficacy and Safety of MEDI0382 in the Treatment of Overweight and Obese Subjects With Type 2 Diabetes — NCT03235050.
- Registry record last updated 2020-08-17; status last verified 2020-07.
- Results first posted 2020-07-20.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
19. Record Summary
NCT03235050 is a completed, randomized phase IIb parallel-group trial sponsored by AstraZeneca. The registry describes a double-blind, placebo-controlled study with an open-label active-comparator component and lists triple masking. Enrollment was 834. The trial studied MEDI0382 with metformin, alongside placebo with metformin and liraglutide with metformin.
The primary endpoint was HbA1c change from baseline to 14 weeks in the ITT population. The registry reports least-squares mean estimates and ANCOVA comparisons of MEDI0382 100, 200, and 300 mcg with placebo. Each reported comparison has p < 0.001 and a 95% confidence interval below zero. The registry also lists secondary outcomes related to HbA1c, body weight, rescue or discontinuation, and safety.
The record does not report key details needed to reproduce or fully assess the analysis, including covariates, missing-data handling, multiplicity procedures, and the trial’s sample-size justification. The findings should be read as results for the registered endpoint, population, comparison, and timeframe.