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 Efficacy and Safety of Dapagliflozin in Combination With Metformin in Type 2 Diabetes Patients (NCT00660907), a randomized phase 3 study. The primary outcome is Adjusted Mean Change in HbA1c Levels.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT00660907 |
| Phase | Phase 3 |
| Condition | Type 2 Diabetes |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Quadruple |
| Enrollment | 1217 |
| Arms in the primary outcome | 2 |
| Lead sponsor | AstraZeneca |
| Results posted | Yes |
2. Clinical Question
Population
Adults with type 2 diabetes whose glycaemic control was inadequate on metformin therapy alone. The primary outcome population was the Full Analysis Set with non-missing baseline and Week 52 values, using last observation carried forward for missing Week 52 values.
Intervention
Dapagliflozin added to metformin. The registry describes the comparison as add-on treatment for patients already receiving metformin.
Comparator
Glipizide, a sulphonylurea, added to metformin. This active comparison asks how the two combinations performed relative to one another, rather than comparing either combination with no add-on treatment.
Primary question
How did adjusted mean HbA1c change from baseline to Week 52 with dapagliflozin plus metformin compared with glipizide plus metformin? The registry reports an ANCOVA comparison and classifies it as a non-inferiority or equivalence analysis.
The primary endpoint focuses on change in HbA1c, a continuous measure of glycaemic control. Both arms had a posted adjusted mean change of -0.52 percent, with a 95% confidence interval from -0.6 to -0.44 for each arm. The reported ANCOVA mean difference in final values was 0.00, with a 95% confidence interval from -0.11 to 0.11.
This is a comparison of treatment strategies in a defined study population and over a specified follow-up period. It does not, by itself, answer questions about outcomes beyond the reported endpoint, other patient groups, or treatment decisions for an individual. The registry lists body weight and hypoglycaemia among the secondary outcomes, but those are distinct questions from the primary HbA1c comparison.
3. Trial Design
The registry describes a Phase III, randomized, parallel-group, double-blind, active-controlled study of dapagliflozin plus metformin compared with glipizide plus metformin. It also records quadruple masking. The official title specifies a 52-week study followed by a 156-week extension period. The primary endpoint timeframe, however, is baseline to Week 52; the estimates discussed for that endpoint concern this period.
The study enrolled 1217 participants and is recorded as completed. The registry lists a start in 2008-03 and primary completion in 2009-12. The primary outcome description concerns participants with type 2 diabetes who had inadequate glycaemic control while taking 1500 mg/day or higher doses of metformin alone. The posted primary outcome analysis includes 400 participants in the dapagliflozin-plus-metformin arm and 401 in the glipizide-plus-metformin arm.
Random allocation is intended to create groups that can be compared while reducing systematic differences in how participants enter treatment groups. An active-controlled design compares the study treatment with another active treatment, here glipizide added to metformin. Parallel-group means participants are assigned to one treatment group rather than crossing between both study treatments. These are general descriptions of design terms, not claims about the success of allocation or masking in this particular trial.
The registry lists dapagliflozin, glipizide, and metformin hydrochloride as study drugs. Its arm descriptions identify the two combinations. The total enrollment and the analysis populations for particular outcomes are not interchangeable: the primary analysis counts differ from enrollment, and the safety groups have their own at-risk counts. Reasons for those differences are not reported in the registry record.
The registry record does not report details of sequence generation, allocation concealment, masking procedures, or how the extension period was analyzed. It therefore supports describing the registered design, but not filling in operational or analysis details that are absent from the record.
4. Participant Flow and Accounting
The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.
Before assignment: For participants with a metformin dose of less than 1500 mg/day, a change in metformin dose in the past 8 weeks or on an Oral Anti-Diabetic (OAD), an 8-week metformin-only dose stabilisation period occurred. A 2-week placebo lead in period occurred after the dose stabilisation period or after enrolment if dose stabilisation period was skipped.
Overall Study
| Stage | Dapagliflozin Plus Metformin | Glipizide Plus Metformin |
|---|---|---|
| STARTED | 406 | 408 |
| COMPLETED | 322 | 314 |
| NOT COMPLETED | 84 | 94 |
Reasons for not completing this period:
| Reason | Dapagliflozin Plus Metformin | Glipizide Plus Metformin |
|---|---|---|
| Incorrect Enrollment | 1 | 1 |
| Adverse Event | 33 | 19 |
| Subject No Longer Meets Study Criteria | 6 | 27 |
| Withdrawal by Subject | 23 | 32 |
| Lost to Follow-up | 3 | 3 |
| Poor/Non-Compliance | 5 | 1 |
| Safety | 1 | 0 |
| Death | 1 | 3 |
| Various | 11 | 8 |
Analysis population for the primary outcome, as defined in the registry: Full Analysis Set, participants with non-missing baseline and Week 52 (LOCF) values
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: Full Analysis Set defined as all randomized participants (as randomized) who received at least one dose of double-blind study medication, who have a non-missing baseline value and at least one post-baseline efficacy value for at least one efficacy variable during double-blind treatment period.
| Characteristic | Unit | Dapagliflozin Plus Metformin | Glipizide Plus Metformin | Total |
|---|---|---|---|---|
| Participants analysed | 400 | 401 | 801 | |
| Age, Continuous | years | 58.1 (SD 9.37) | 58.6 (SD 9.80) | 58.4 (SD 9.58) |
| Sex: Female, Male: Female | Participants | 179 | 181 | 360 |
| Sex: Female, Male: Male | Participants | 221 | 220 | 441 |
| Race/Ethnicity, Customized: American Indian or Alaska Native | Participants | 0 | 0 | 0 |
| Race/Ethnicity, Customized: Asian | Participants | 27 | 34 | 61 |
| Race/Ethnicity, Customized: Native Hawaiian or Other Pacific Islander | Participants | 0 | 0 | 0 |
| Race/Ethnicity, Customized: Black or African American | Participants | 26 | 24 | 50 |
| Race/Ethnicity, Customized: White | Participants | 327 | 323 | 650 |
| Race/Ethnicity, Customized: More than one race | Participants | 0 | 0 | 0 |
| Race/Ethnicity, Customized: Unknown or Not Reported | Participants | 20 | 20 | 40 |
| BMI | kg/m2 | 31.71 (SD 5.104) | 31.23 (SD 5.053) | 31.47 (SD 5.081) |
| HbA1c | percent | 7.69 (SD 0.855) | 7.74 (SD 0.886) | 7.72 (SD 0.870) |
| FPG | ng/mL | 162.24 (SD 37.796) | 163.91 (SD 41.559) | 163.07 (SD 39.708) |
6. Primary Endpoint
The registry lists the primary outcome as Adjusted Mean Change in HbA1c Levels, measured over Baseline to Week 52.
| Element | Registry entry |
|---|---|
| Result type | Least Squares Mean |
| Measure of spread | 95% Confidence Interval |
| Unit | percent |
| Endpoint classification | Continuous Endpoint |
The primary endpoint is adjusted mean change in HbA1c from baseline to Week 52, expressed in percent. HbA1c is a continuous measure used to assess glycaemic control. A change score describes how the measured value at follow-up differs from baseline; the adjusted mean is an estimate based on the analysis model rather than a simple count of participants meeting a threshold.
Continuous outcomes retain information about the size and direction of changes across participants. This can be useful when the scientific question concerns the average change in a measurement, rather than only whether a person crosses a predefined category. In general, an adjusted estimate may account for variables included in the statistical model. The registry names ANCOVA for the comparison, but the covariates in that model are not reported in the registry record.
The posted arm-level adjusted mean changes are -0.52 percent in both groups. Each arm’s 95% confidence interval is -0.6 to -0.44. The registry’s stated population is the Full Analysis Set with non-missing baseline and Week 52 values, with last observation carried forward used for Week 52 values. Thus, the endpoint estimate is tied to that analysis population and missing-data approach; it should not be treated as an estimate from every enrolled participant without qualification.
HbA1c is only one aspect of the study’s broader assessment. Body weight and hypoglycaemia are among the listed secondary outcomes, but they answer different questions and do not replace the primary endpoint.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The posted adjusted mean change in HbA1c was -0.52 percent for dapagliflozin plus metformin and -0.52 percent for glipizide plus metformin. The reported ANCOVA mean difference in final values was 0.00, with a 95% confidence interval from -0.11 to 0.11. These estimates are very similar at the precision displayed in the registry. The arm-level confidence intervals are also the same at the displayed precision.
A confidence interval gives a range of values compatible with the data and the assumptions of the analysis, in the sense used by frequentist statistical methods. It is not a guarantee that the true value lies in that range. For the reported between-arm difference, the interval extends in both directions from zero. That means the posted interval is compatible with a small difference favoring either arm, as well as no difference on this scale. The registry’s rounding also matters: values are reported to limited precision, so apparent equality at the displayed digits does not show that every underlying estimate was exactly identical.
The registry classifies the reported ANCOVA analysis as non-inferiority or equivalence and gives p < 0.001. This p-value should be read in the context of that analysis classification, not automatically as a conventional test of whether the mean difference is zero. A non-inferiority or equivalence test uses a prespecified boundary to define an acceptable range. The boundary, hypothesis details, and interpretation of the p-value are not reported in the registry record. Without those details, the p-value cannot independently establish what margin was tested or whether a particular formal criterion was met.
The result therefore supports a narrow statistical description: the posted arm estimates are closely aligned, and the sponsor-reported model gives a mean difference of 0.00 with the stated confidence interval and p-value. It does not establish exact equality. Nor does the p-value alone show how clinically important a difference would be. The registry does not report a threshold for clinical importance.
The analysis concerns a continuous endpoint at Week 52 in a specified analysis population. It does not establish the comparative results for every secondary outcome, for the full extension period, or for all enrolled participants. Those questions require their own outcome estimates and analysis details. A close estimate for HbA1c should not be extended into an unsupported claim about all benefits or harms.
9. Statistical Methods Explained
What does the mean change describe?
A mean summarizes the central value of measurements in a group. For a change-from-baseline endpoint, each participant’s follow-up measurement is considered relative to that participant’s baseline measurement, and the group result summarizes those changes. The posted values are adjusted means, not counts of participants with a particular outcome. In this study, both arms have an adjusted mean change of -0.52 percent from baseline to Week 52.
Adjustment is a general statistical approach in which a model estimates a group result while accounting for variables specified in that model. It can improve precision or address imbalances in modeled factors, depending on the design and data. The registry identifies ANCOVA for the comparison but does not report the model covariates, their coding, or further modeling details. It is therefore not possible to determine from the registry record exactly what adjustments were made.
How should the confidence interval be read?
A confidence interval expresses uncertainty around an estimate under the assumptions of the statistical method. The arm-level 95% confidence intervals are -0.6 to -0.44 for both groups. The between-arm ANCOVA estimate is 0.00, with a 95% confidence interval from -0.11 to 0.11. The latter interval describes uncertainty in the comparative estimate, which is the relevant interval for interpreting the difference between arms.
General interpretation: a narrower interval indicates greater precision than a wider interval on the same scale, while an interval spanning zero does not demonstrate a nonzero difference at the corresponding conventional two-sided level. That observation does not settle a non-inferiority or equivalence question, because those questions compare the interval with a margin rather than relying only on whether it crosses zero. The margin is not reported in the registry record.
What does ANCOVA contribute?
ANCOVA is a general model-based method for comparing an outcome between groups while accounting for one or more covariates. For a clinical trial endpoint, the model may use baseline information and other prespecified variables, but the registry record does not identify the covariates for this analysis. The reported result is a mean difference in final values, rather than a reported difference between the two posted change-from-baseline estimates. These are related ways of expressing a comparison, but they should not be conflated when describing the registry’s analysis.
Model-based estimates depend on the chosen model and the data included. A confidence interval reflects uncertainty under that model; it does not capture every possible source of bias or uncertainty, such as missing-data assumptions or departures from model assumptions. The registry does not report diagnostics or sensitivity analyses for the ANCOVA.
Why does the analysis label matter?
In a conventional superiority test, the primary question is whether the data are inconsistent with a null hypothesis of no difference, according to a prespecified testing procedure. Non-inferiority and equivalence questions are different. Non-inferiority asks whether a treatment is not worse than a comparator by more than a defined margin; equivalence asks whether a difference falls within a defined range. These are general concepts. The registry groups the analysis under a non-inferiority-or-equivalence classification, but it does not report the margin or enough hypothesis details to distinguish the intended criterion from that label alone.
The reported p-value is p < 0.001. It should not be casually interpreted as evidence that the arms differ, given the listed analysis classification and the reported estimate and interval. A p-value is the probability, under a stated null hypothesis and model, of results at least as incompatible with that hypothesis as those observed. Its meaning depends on which null hypothesis was tested. Since the precise hypothesis and margin are not reported in the registry record, the p-value’s exact inferential role cannot be reconstructed from the registry entry.
What is the role of the analysis population?
The registry identifies the primary analysis population as the Full Analysis Set, restricted to participants with non-missing baseline and Week 52 values, with last observation carried forward for missing Week 52 values. A Full Analysis Set is a trial-specific analysis population; the registry does not provide a fuller definition of membership. The non-missing-value criterion means the analysis is not simply an analysis of every enrolled participant regardless of endpoint availability.
Last observation carried forward is a general imputation approach that substitutes an earlier recorded value when the endpoint value at the target visit is missing. It can be easy to describe, but it assumes that carrying the earlier measurement forward is a defensible representation of the missing value. That assumption may not hold in every setting. The registry does not report the amount or pattern of missingness, the timing of carried-forward measurements, or sensitivity analyses using other approaches.
The posted primary outcome counts are 400 for dapagliflozin plus metformin and 401 for glipizide plus metformin. The safety at-risk counts are 406 and 408, respectively. Those counts refer to different reported analysis contexts and should not be substituted for one another. The registry record does not explain why the counts differ.
How should planning illustrations be understood?
General sample-size planning translates a target effect, variability assumptions, significance level, allocation, and desired power into an approximate number of participants. The stated planning approach uses a two-sided significance level of 0.05, equal allocation, a normal approximation, and effects defined as fractions of the observed effect. Such calculations are illustrations based on assumptions and the observed estimate; they are not the trial’s own design justification and are not observed power. The registry record does not report the original sample-size justification or the actual numerical planning illustrations in the information. Planning results should not be read as evidence that the trial had a particular probability of detecting an effect after the results were observed.
10. Analyses Reported by the Sponsor
| Method | Estimate | 95% CI | P-value |
|---|---|---|---|
| ANCOVA | Mean Difference (Final Values) 0.00 | -0.11 to 0.11 | < 0.001 |
The registry reports the sponsor’s ANCOVA comparison as a mean difference in final values of 0.00, with a 95% confidence interval from -0.11 to 0.11 and p < 0.001. It classifies this analysis as non-inferiority or equivalence. The arm-level adjusted mean changes are both -0.52 percent, each with a 95% confidence interval from -0.6 to -0.44. These displayed values are consistent in showing closely aligned arm estimates, but they are not identical descriptions of the same statistic: the arm-level values are adjusted changes, while the reported comparison is a mean difference in final values from ANCOVA.
The p-value should be understood through the sponsor-reported analysis framework. It is not appropriate to treat it automatically as a test of a zero difference, because the registry labels the analysis as non-inferiority or equivalence. The margin and full hypothesis details are not reported in the registry record, so the specific interpretation of that p-value cannot be fully checked from the entry alone.
The posted estimates are rounded, and no additional unadjusted reconstruction is needed to describe the sponsor’s analysis. The registry record does not report a conflicting sponsor estimate. No recalculated result should be taken as a substitute for the sponsor’s model-based analysis or as a formal reanalysis of participant-level data.
11. Multiplicity, Missing Data and Analysis Populations
The registry lists a primary outcome of adjusted mean change in HbA1c and three secondary outcomes: adjusted mean change in body weight, the proportion with at least one episode of hypoglycaemia, and the proportion with body weight reduction of at least 5%. It records four outcome measures and four analyses. These listings show that more than one outcome was assessed, but they do not explain how the study handled the increased chance of false-positive findings that can arise when multiple hypotheses are tested.
Multiplicity methods are general statistical procedures for controlling or managing error across several comparisons. Examples include hierarchical testing, adjustment of significance thresholds, and treating secondary analyses as exploratory. The registry record does not report whether any such strategy was prespecified, which outcomes were included in any adjustment, or whether the secondary outcomes were intended to support confirmatory claims. The primary result should therefore be distinguished from conclusions based on multiple secondary outcomes.
The primary outcome population is described as the Full Analysis Set with non-missing baseline and Week 52 values, using last observation carried forward for missing Week 52 values. This identifies an analysis rule, but it does not give the number of missing endpoint values, their distribution across arms, the reasons for missingness, or the sensitivity of the result to other missing-data assumptions. The effect of missing data cannot be judged fully without those details.
Last observation carried forward is a general imputation method. It assumes that an earlier observation can stand in for a later missing one. If outcomes change over time after the last observation, the carried-forward value may not reflect the unobserved Week 52 value. That possibility does not establish the direction or size of any bias in this trial; the necessary information is not reported in the registry record.
The endpoint population counts and safety at-risk counts are distinct. The registry reports 400 and 401 participants for the primary endpoint arms, compared with 406 and 408 at risk in the safety groups. The record does not explain these differences. It also does not report a detailed definition of the Full Analysis Set, sensitivity analyses, a missing-data model, or the protocol’s multiplicity plan. These omissions limit independent assessment of the robustness and confirmatory status of the reported findings.
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 |
|---|---|---|---|
| Dapagliflozin Plus Metformin | 35 | 406 | 8.6% |
| Glipizide Plus Metformin | 46 | 408 | 11.3% |
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 |
|---|---|
| Dapagliflozin Plus Metformin | 163 of 406 |
| Glipizide Plus Metformin | 260 of 408 |
Most frequent serious adverse event terms
| Event term | Organ system | Dapagliflozin Plus Metformin | Glipizide Plus Metformin |
|---|---|---|---|
| Myocardial Infarction | Cardiac disorders | 1 of 406 | 3 of 408 |
| Prostate Cancer | Neoplasms benign, malignant and unspecified (incl cysts and polyps) | 3 of 406 | 1 of 408 |
| Hypoglycaemia | Metabolism and nutrition disorders | 0 of 406 | 3 of 408 |
| Anaemia | Blood and lymphatic system disorders | 0 of 406 | 2 of 408 |
| Angina Pectoris | Cardiac disorders | 1 of 406 | 1 of 408 |
| Coronary Artery Disease | Cardiac disorders | 1 of 406 | 1 of 408 |
| Retinal Detachment | Eye disorders | 1 of 406 | 1 of 408 |
| Diverticulitis | Infections and infestations | 0 of 406 | 2 of 408 |
| Bronchitis | Infections and infestations | 1 of 406 | 1 of 408 |
| Pneumonia | Infections and infestations | 1 of 406 | 1 of 408 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Dapagliflozin Plus Metformin | Glipizide Plus Metformin |
|---|---|---|---|
| Hypoglycemia | Endocrine disorders | 14 of 406 | 162 of 408 |
| Nasopharyngitis | Infections and infestations | 43 of 406 | 61 of 408 |
| Hypertension | Vascular disorders | 30 of 406 | 35 of 408 |
| Influenza | Infections and infestations | 30 of 406 | 30 of 408 |
| Upper Respiratory Tract Infection | Infections and infestations | 24 of 406 | 31 of 408 |
| Dizziness | Nervous system disorders | 15 of 406 | 37 of 408 |
| Urinary Tract Infection | Infections and infestations | 30 of 406 | 17 of 408 |
| Diarrhoea | Gastrointestinal disorders | 19 of 406 | 26 of 408 |
| Headache | Nervous system disorders | 21 of 406 | 17 of 408 |
| Arthralgia | Musculoskeletal and connective tissue disorders | 11 of 406 | 21 of 408 |
Serious events with the highest arm-level rate: prostate cancer (0.7%), myocardial infarction (0.7%), hypoglycaemia (0.7%), uterine prolapse (0.5%), anaemia (0.5%).
The registry reports serious adverse events among 35 of 406 participants at risk in the dapagliflozin-plus-metformin group and 46 of 408 at risk in the glipizide-plus-metformin group. It also reports 163 and 260 participants, respectively, in the other adverse-event category. The record labels these counts as affected participants and gives the corresponding at-risk counts. They are descriptive safety figures; they are not, by themselves, a complete comparison of event severity, timing, duration, or relationship to treatment.
The listed serious adverse-event terms include Prostate Cancer, Myocardial Infarction, and Hypoglycaemia at 0.7% each, and Uterine Prolapse, Anaemia, and Diverticulitis at 0.5% each. The registry entry does not assign these listed percentages to a particular treatment group in the information reported, so they should not be attributed to one arm. The number of deaths is not reported in the registry record.
General interpretation: adverse-event counts depend on the definitions used, the period of observation, ascertainment, and the number of participants at risk. A count comparison alone does not establish causation or explain clinical importance. The registry record does not provide event-level details or enough context to infer why events occurred. These figures describe reported events in the study groups and should be read separately from the primary HbA1c endpoint.
13. What the Result Does, and Does Not, Mean
The central statistical observation is that the posted adjusted mean HbA1c changes are the same at the displayed precision: -0.52 percent in each arm. The sponsor-reported ANCOVA mean difference in final values is 0.00, with a 95% confidence interval from -0.11 to 0.11. This conveys a close estimated comparison, while the interval indicates uncertainty around the between-arm estimate.
That finding does not prove that the treatments are exactly equal. An estimate of zero at the displayed precision is not proof of a zero underlying difference, and a confidence interval is not a guarantee about the true value. The interval includes values on either side of zero. Formal interpretation as non-inferiority or equivalence depends on the margin and testing details, which are not reported in the registry record.
The statistical result also does not determine clinical importance on its own. The registry does not report a threshold for a clinically meaningful HbA1c difference. Nor does the primary outcome settle the comparison for body weight, hypoglycaemia, or other outcomes. Those measures require their own estimates, uncertainty intervals, and analysis context.
The result is specific to the registered comparison, the Week 52 timeframe, and the described endpoint population. The registry’s missing-data approach uses last observation carried forward for missing Week 52 values, and details about missingness and sensitivity analyses are not reported. The estimate should therefore be read in light of that analysis rule rather than assumed to represent all enrolled participants under every possible missing-data approach.
Finally, these results are not a general statement about every patient or every use of either treatment combination. They describe trial-group statistics in the recorded study. They do not provide individualized medical guidance, establish what happened during the full extension period, or answer questions not represented by the reported outcomes.
14. Why This Trial Matters Statistically
This trial offers a clear example of why the comparison estimate, confidence interval, and analysis label need to be read together. The arm-level adjusted mean changes are both -0.52 percent, while the ANCOVA comparison reports a difference of 0.00 with a 95% confidence interval from -0.11 to 0.11. The similarity of the rounded estimates is informative, but it is not equivalent to proof of exact equality.
The registry also illustrates how a p-value can be misunderstood if its testing framework is overlooked. The reported p < 0.001 appears alongside a non-inferiority-or-equivalence classification. Because those frameworks use margins and hypotheses that differ from a conventional test for any difference, the p-value cannot be interpreted in isolation. The registry record’s omission of the margin and full hypothesis details makes that limitation especially visible.
There is also a practical distinction between an outcome population and a safety population. The primary endpoint counts and safety at-risk counts differ, and each is relevant to a different analysis. Keeping those denominators distinct avoids treating all reported numbers as if they came from the same set of participants.
Finally, the endpoint’s missing-data rule matters to the meaning of the estimate. Last observation carried forward is straightforward to state, but its suitability depends on how outcomes evolve and why follow-up values are missing. The registry does not provide the details needed to assess those issues. Together, these features demonstrate why a careful statistical reading includes the estimate, uncertainty, hypothesis framework, analysis population, and missing-data approach—not just a single number.
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 registry record does not report the non-inferiority or equivalence margin, the precise null hypothesis, or enough testing details to determine exactly how the reported p < 0.001 relates to the stated analysis classification. This limits independent interpretation of the formal claim. The record also does not provide the covariates used in the ANCOVA, so the adjustment behind the reported comparison cannot be fully assessed.
The primary endpoint population is described as participants with non-missing baseline and Week 52 values, with last observation carried forward for missing Week 52 values. Counts and patterns of missing endpoint data, reasons for missingness, timing of carried-forward observations, and sensitivity analyses are not reported in the registry record. Those omissions prevent assessment of how dependent the estimate may be on the missing-data rule.
The registry lists three secondary outcomes, but a multiplicity plan and the confirmatory status of those outcomes are not reported. The record also does not provide the original sample-size justification. Planning illustrations based on the observed effect are not a substitute for that justification and are not observed power.
Although the official title includes a 156-week extension period, the primary endpoint is baseline to Week 52. Extension-period results and their analysis details are not reported in the registry record discussed for this comparison. The registry does not explain why endpoint analysis counts differ from safety at-risk counts. The listed serious adverse-event terms are not assigned to a treatment arm in the record, and deaths are not reported. These limitations restrict what can be concluded from the posted summary statistics.
16. Related Tutorials
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: Efficacy and Safety of Dapagliflozin in Combination With Metformin in Type 2 Diabetes Patients — NCT00660907.
- Registry record last updated 2015-03-30; status last verified 2015-03.
- Results first posted 2013-08-23.
- 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
The registry describes a completed Phase III randomized, parallel-group, double-blind, active-controlled study, with quadruple masking recorded. It compares dapagliflozin plus metformin with glipizide plus metformin in adults with type 2 diabetes and inadequate glycaemic control on metformin alone. Enrollment was 1217, and the official title describes a 52-week study with a 156-week extension period.
The primary outcome is adjusted mean HbA1c change from baseline to Week 52. Both arms have a posted adjusted mean change of -0.52 percent; the reported ANCOVA mean difference in final values is 0.00, with a 95% confidence interval from -0.11 to 0.11 and p < 0.001. The analysis is classified as non-inferiority or equivalence, but the margin and full hypothesis details are not reported in the registry record.
Interpretation is limited by missing details about multiplicity, model covariates, missingness, sensitivity analyses, and the extension-period results. The reported endpoint estimate should be read within its stated population and last-observation-carried-forward approach.