Efficacy and Safety of Empagliflozin (BI 10773) / Linagliptin (BI 1356) Fixed Dose Combination in Treatment naïve and Metformin Treated Type 2 Diabetes Patients: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy and Safety of Empagliflozin (BI 10773) / Linagliptin (BI 1356) Fixed Dose Combination in Treatment naïve and Metformin Treated Type 2 Diabetes Patients (NCT01422876), a randomized phase 3 study. The primary outcome is Change From Baseline in Glycosylated Hemoglobin (HbA1c) for Metformin Background Patients.

2. Clinical Question

3. Trial Design

This was a phase III, randomized, double-blind, parallel-group study sponsored by Boehringer Ingelheim. The registry reports enrollment of 1405 and lists type 2 diabetes as the condition. The study compared two once-daily fixed-dose combinations of empagliflozin and linagliptin with the individual components. The trial included treatment-naïve patients and patients treated against a metformin background. Its official title specifies a 52-week study, while the primary HbA1c endpoint was assessed from baseline to week 24.

The registry records the study start as 2011-08 and primary completion as 2013-09, with status listed as completed. Randomization, double masking, and a parallel-group model are recorded. Further details about the randomization procedure, concealment, masking implementation, site structure, adherence, or discontinuations are not reported in the registry record.

The primary analysis described in the registry is for metformin-background patients in the Full Analysis Set with last observation carried forward. The registry defines this set as randomized and treated patients who had a baseline and at least one on-treatment HbA1c value. The primary endpoint summaries include the number analyzed in each group, the least-squares mean change, and its standard error. The record also posts sponsor ANCOVA comparisons for selected treatment groups. It does not report the original design assumptions or a full statistical analysis plan, so the analyses should be interpreted using only the methods and populations explicitly described in 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: Of the 1405 patients enrolled and randomized the data for 42 randomized patients were excluded from all analyses due to serious non-compliance. Therefore, 1363 patients were included in the analyses.

Only the first 6 of 10 groups are shown.

Week 24

Reasons for not completing this period:

Week 52

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full Analysis Set (FAS) with last observation carried forward (LOCF). FAS - all Metformin Background patients randomised and treated who had a baseline and at least 1 on treatment HbA1c value.

5. Baseline Characteristics

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

Population: FAS

Only the first 6 of 11 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in Glycosylated Hemoglobin (HbA1c) for Metformin Background Patients, measured over Baseline and 24 weeks.

The primary endpoint is change from baseline in glycosylated hemoglobin, or HbA1c, among metformin-background patients. HbA1c measures the percentage of hemoglobin that is glycated. The registry defines change as the week 24 HbA1c value minus the baseline value; because HbA1c itself is expressed as a percentage, the change is also expressed in percentage units. A negative change means the measured value was lower at week 24 than at baseline.

HbA1c is used as a continuous measure of glycaemic control, allowing the analysis to describe both the average change within each group and differences between groups. The registry reports least-squares means and standard errors for the primary endpoint. Least-squares means are model-based group estimates; the record’s posted sponsor comparisons use ANCOVA. Those model-based comparisons need not be identical to simple differences between the displayed group means, because adjustment and analysis details can affect the estimates.

The primary population is not all participants enrolled in the trial. It is the metformin-background Full Analysis Set defined in the registry, with last observation carried forward for the primary endpoint. Details of the timing and handling of missing values beyond that stated approach are not reported in the registry record. The outcome therefore describes a specified group and measurement time, rather than all possible aspects of diabetes control or health.

7. Results Posted in the Registry

• Metformin Background: Empagliflozin 10 mg/Linagliptin 5 mg: mean -1.08 versus -1.19 on control; difference 0.11 (95% CI -0.06 to 0.28), standardized effect d = 0.16; z-test p = 0.195.
• Metformin Background: Empagliflozin 25 mg: mean -0.62 versus -1.19 on control; difference 0.57 (95% CI 0.40 to 0.74), standardized effect d = 0.81; z-test p = <0.001.
• Metformin Background: Empagliflozin 10 mg: mean -0.66 versus -1.19 on control; difference 0.53 (95% CI 0.36 to 0.70), standardized effect d = 0.76; z-test p = <0.001.
• Metformin Background: Linagliptin 5 mg: mean -0.70 versus -1.19 on control; difference 0.49 (95% CI 0.32 to 0.66), standardized effect d = 0.71; 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.

Change From Baseline in Fasting Plasma Glucose at Week 24 for Metformin Background Patients

• 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:
  • Metformin Background: Empagliflozin 10 mg/Linagliptin 5 mg: mean -32.18 versus -35.25 on control; difference 3.07 (95% CI -3.93 to 10.07), standardized effect d = 0.11; z-test p = 0.390.
  • Metformin Background: Empagliflozin 25 mg: mean -18.83 versus -35.25 on control; difference 16.42 (95% CI 9.49 to 23.35), standardized effect d = 0.56; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Metformin Background: Empagliflozin 10 mg/Linagliptin 5 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Metformin Background: Empagliflozin 25 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Fasting Plasma Glucose at Week 24 for Treatment Naive Patients

• 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:
  • Treament Naive: Empagliflozin 10 mg/Linagliptin 5 mg: mean -28.21 versus -29.55 on control; difference 1.34 (95% CI -6.05 to 8.73), standardized effect d = 0.04; z-test p = 0.722.
  • Treatment Naive: Empagliflozin 25 mg: mean -24.24 versus -29.55 on control; difference 5.31 (95% CI -2.10 to 12.72), standardized effect d = 0.17; z-test p = 0.160.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Treament Naive: Empagliflozin 10 mg/Linagliptin 5 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Treatment Naive: Empagliflozin 25 mg and the control. The interval gives the range of effects that remain compatible with the data.

Change From Baseline in Body Weight for Metformin Background Patients

• 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:
  • Metformin Background: Empagliflozin 10 mg/Linagliptin 5 mg: mean -2.60 versus -2.99 on control; difference 0.39 (95% CI -0.46 to 1.24), standardized effect d = 0.11; z-test p = 0.366.
  • Metformin Background: Empagliflozin 25 mg: mean -3.18 versus -2.99 on control; difference -0.19 (95% CI -1.04 to 0.66), standardized effect d = -0.05; z-test p = 0.660.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Metformin Background: Empagliflozin 10 mg/Linagliptin 5 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Metformin Background: Empagliflozin 25 mg and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists secondary outcomes covering fasting plasma glucose, body weight, and treat-to-target efficacy response, with measures for metformin-background and treatment-naïve patients. Posted numerical analyses are available for fasting plasma glucose in both treatment backgrounds and body weight in metformin-background patients. The registry lists six secondary outcomes, but numerical results for treat-to-target efficacy response are not reported in the registry record. Secondary findings therefore add context only where a numerical result is posted.

Across the posted continuous secondary outcomes, group changes are not uniformly aligned in their between-group comparisons. The registry includes sponsor ANCOVA analyses for selected comparisons and unadjusted recalculations for some of those outcomes. The fasting plasma glucose results include sponsor-reported contrasts in metformin-background patients, while the posted treatment-naïve sponsor comparisons have confidence intervals that include zero. For body weight in metformin-background patients, sponsor results differ by comparison: one reported contrast has an interval that includes zero, and another does not. These patterns do not amount to a single overall conclusion across all secondary measures.

The secondary outcomes relate to the primary HbA1c result by describing other measures at the same general assessment period, but they are distinct endpoints and should not be treated as interchangeable confirmation. The registry does not report a multiplicity procedure for the secondary outcome family. General explanation: when multiple outcomes and comparisons are examined, an isolated small p-value is less persuasive if the analysis does not account for the number of opportunities to find a statistically notable result. Secondary analyses are also less definitive when their prespecification, hierarchy, and missing-data handling are unclear. The primary result should therefore remain the main focus, with secondary findings interpreted as supporting context rather than independent proof.

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

• Change From Baseline in Fasting Plasma Glucose at Week 24 for Metformin Background Patients
• Change From Baseline in Fasting Plasma Glucose at Week 24 for Treatment Naive Patients
• Change From Baseline in Body Weight for Metformin Background Patients
• Change From Baseline in Body Weight for Treatment Naive Patients
• Occurrence of Treat to Target Efficacy Response for Metformin Background Patients
• Occurrence of Treat to Target Efficacy Response for Treatment Naive Patients

9. Statistical Interpretation

10. Statistical Methods Explained

Statistical methods are useful only when the estimand, population, and comparison are clear. The primary measure is a continuous outcome: change in HbA1c from baseline to week 24. A group mean summarizes the average change, while a between-group contrast addresses how the estimated changes differ. Neither is a count of patients who reached a target. The registry reports least-squares means for the group summaries and ANCOVA for the sponsor’s selected comparisons.

What does ANCOVA contribute?

General explanation: analysis of covariance, or ANCOVA, models an outcome while accounting for one or more covariates. In a randomized study, such adjustment can improve precision or account for baseline measurements, depending on the model specified. It produces an adjusted contrast rather than simply subtracting two displayed group averages. The registry identifies the sponsor’s comparison method as ANCOVA, but the full model specification, included covariates, and handling of model assumptions are not reported in the registry record. The sponsor’s estimates therefore should be read as the reported adjusted analyses, not reconstructed from group means.

What is the unadjusted recalculation?

The recalculated comparisons use the posted arm summaries to form simple, unadjusted comparisons. The posted group values are least-squares means and their reported dispersion is standard error. The recalculation records an approximate pooled standard deviation, as the underlying participant-level observations are not part of the posted summary statistics. It is not a substitute for the sponsor’s ANCOVA, which may account for covariates and use details that cannot be recovered from the group summaries. The unadjusted p-values and confidence intervals are consequently best treated as checks on the posted summaries, not as the primary trial analysis.

General explanation: a standard error describes the uncertainty in an estimated mean, whereas a standard deviation describes variation among individual observations. They are not interchangeable. In this registry record, standard errors accompany least-squares mean changes. Because the recalculated pooled standard deviation is approximate, calculations based on it inherit that approximation. This matters particularly when interpreting standardized effect estimates or planning illustrations derived from the posted summary data.

Why can analyses differ?

Analyses can differ because they use different estimands, adjustment variables, populations, missing-data procedures, and variance estimates. The sponsor’s analyses are ANCOVA contrasts; the recalculations are unadjusted and use posted aggregate summaries. The signs of the contrasts also follow different conventions: a difference calculated in one group order can have the opposite sign from a sponsor contrast using the reverse order. The sign alone should not be read as a contradiction without checking which group is subtracted from which. The record identifies a discrepancy, so the sponsor’s own analysis takes precedence for the trial’s reported inferential comparisons.

General explanation: a p-value measures how compatible the observed data are with a specified null model, under the assumptions of the analysis. It does not measure the probability that a treatment works, nor does it describe the size or clinical importance of a difference. A confidence interval adds an estimate of uncertainty and the range of effect values reasonably compatible with the analysis. Both depend on the model and the data used. The unadjusted recalculated p-values should therefore not be conflated with sponsor-reported ANCOVA p-values.

What does the analysis population mean?

The registry specifies a Full Analysis Set for the primary outcome: metformin-background patients who were randomized and treated and had baseline and at least one on-treatment HbA1c value. The primary endpoint uses last observation carried forward, abbreviated LOCF. General explanation: LOCF fills a missing later measurement with the last available on-treatment observation. This approach can be sensitive to assumptions about how outcomes change after the last observed measurement. The registry does not give further detail about missingness patterns, reasons for missing outcomes, or sensitivity analyses, so the robustness of the primary result to alternative missing-data methods cannot be assessed from this record.

The record lists a primary endpoint and secondary outcomes, and it includes multiple sponsor comparisons. General explanation: when many outcomes or comparisons are tested, the chance of at least one apparently strong result can rise unless multiplicity is addressed. Whether and how multiplicity was controlled is not reported in the registry record. Accordingly, unadjusted secondary comparisons and exploratory recalculations carry less inferential weight than prespecified, appropriately controlled analyses. The original statistical plan, including the hierarchy of hypotheses and any interim analyses, is not reported in the registry record.

11. Confidence Intervals and Precision

Confidence intervals describe uncertainty around an estimated contrast under the analysis used. A relatively narrow interval indicates greater statistical precision than a wider interval, but precision alone does not establish clinical importance. Intervals that include zero are compatible with no difference under the specified contrast model; intervals that exclude zero indicate that the data are less compatible with a zero contrast under that model. These interpretations depend on the model assumptions and do not replace consideration of the effect estimate itself.

The sponsor-reported primary-outcome comparisons have 95% confidence intervals. For the combination-versus-component contrasts reported in the registry, the intervals are entirely on the negative side of zero under the sponsor’s contrast convention. In contrast, the recalculated comparison between the two combination doses has an estimate of 0.11 and an interval from -0.056 to 0.276, which includes zero. That unadjusted result is less precise about the direction and size of a difference between those groups. The two analyses answer different statistical questions and should not be treated as interchangeable.

The arm-level primary summaries report standard errors of 0.06, but they do not include arm-level confidence limits in the registry record. A standard error expresses uncertainty in the estimated mean, not the range of individual participant responses. The precision of any estimate also depends on the analysis population, model, and missing-data handling. The registry does not report participant-level data or enough model detail to independently reconstruct the sponsor’s ANCOVA intervals.

General explanation: a confidence interval is not a guarantee that the true effect lies within its endpoints, and it is not a range containing a stated proportion of individual outcomes. It is a way to communicate uncertainty under a statistical procedure. A narrow interval may still leave uncertainty about whether an effect is important, while a wide interval may leave a broad range of plausible effects. The posted results support interpretation of the specific endpoints and comparisons reported, not conclusions about unreported outcomes.

12. Planning a Similar Trial: Sample Size

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

The planning illustrations express sample size as participants per arm, not as total events. They are based on the observed effect in the posted data, with the effect to detect set at the stated fraction of that observed effect. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The original trial’s own sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the illustrative participants per arm for a design with an 80% chance of detecting the specified effect under those assumptions, if the assumed effect and model are appropriate. The 90% power column gives the corresponding planning figure for a 90% chance. These are prospective planning concepts, not descriptions of the actual trial’s power after its results are known. The displayed figures are not observed power.

Rows that plan for 75% or 50% of the observed effect assume smaller effects than a row planning for 100% of the observed effect. Smaller effects are harder to distinguish from random variation, so they require more participants per arm to achieve the same nominal power under the stated assumptions. This relationship is an illustration, not a claim that the effect in a future study will match any particular fraction. The planning figures are based on simple unadjusted recalculations from posted summaries, and therefore should not be mistaken for the sponsor’s original design calculations or a replacement for design work using a prespecified analysis model.

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 = 0.110; pooled standard deviation σ = 0.696.
2. n = 2 × (1.96 + 0.842)² × 0.696² / 0.110² = 628.2, rounded up to 629 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 <- 0.11                      # observed difference in means
sigma <- 0.695849654431928                     # 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(1422876)
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 = 629 (80% power), 841 (90% power)
 75% of observed effect: n per arm = 1,117 (80% power), 1,496 (90% power)
 50% of observed effect: n per arm = 2,513 (80% power), 3,364 (90% power)
power.t.test: 629.1 
Simulated power at n = 629 per arm: 0.800 (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 reports sponsor analyses using ANCOVA for selected comparisons among metformin-background groups. Those model-based results are the trial’s own reported inferential analyses. The recalculated comparisons are simple, unadjusted re-analyses based on aggregate arm summaries, with approximate pooled standard deviations derived from the posted standard errors. They may differ from the sponsor’s results because the methods and possibly the contrast conventions differ.

The sponsor reports comparisons of each combination with selected individual components. For these contrasts, the posted ANCOVA estimates, confidence intervals, and p-values should take precedence over the recalculated values. The record explicitly flags a discrepancy between sponsor results and recalculations. The recalculation remains useful as a transparent check of what can be recovered from the posted summaries, but it cannot reproduce an adjusted analysis without the sponsor’s model details and underlying data.

Not every possible pairwise comparison is reported by the sponsor. A recalculated comparison should not be described as a sponsor finding where no corresponding sponsor analysis appears in the registry record. Nor should the difference between two methods automatically be interpreted as evidence that either is erroneous: they may estimate different contrasts or use different assumptions. The registry does not report enough detail to resolve the discrepancy beyond giving the sponsor’s reported analyses, which take precedence.

15. Multiplicity, Missing Data and Analysis Populations

The registry record describes the primary endpoint and lists secondary outcomes, but it does not report whether a multiplicity adjustment was used. The multiplicity strategy, if any, for the several outcomes and selected between-group comparisons is not reported in the registry record. It is also not reported whether the comparisons were organized in a formal testing hierarchy, whether any tests were exploratory, or whether an interim analysis was planned or conducted. These omissions limit how strongly individual p-values can be interpreted across the full set of analyses.

General explanation: testing several hypotheses can increase the chance of at least one small p-value even when no real effects exist. Procedures such as an ordered testing strategy or adjustment of significance thresholds can address that issue, but the appropriate approach depends on the protocol’s prespecified questions. Without the statistical plan, the registry results cannot establish whether the posted analyses were protected against multiplicity. Unadjusted p-values for secondary or recalculated comparisons should consequently be viewed cautiously.

The primary endpoint population is described as the Full Analysis Set for metformin-background patients, with randomized and treated patients required to have a baseline and at least one on-treatment HbA1c value. The primary endpoint uses LOCF. The registry does not report the amount or pattern of missing data, reasons for missingness, or sensitivity analyses testing alternative assumptions. The handling of missing data for the secondary outcomes is not reported in the registry record. It also does not report whether participants were analyzed according to randomized assignment after treatment discontinuation or describe the analysis populations used for safety outcomes in enough detail to assess all exclusions.

General explanation: missingness can affect estimates when people with unavailable outcomes differ from those whose outcomes are observed. LOCF is one imputation approach and can be sensitive to the assumption that the last observed value remains an appropriate substitute for a later measurement. Without missingness summaries and sensitivity analyses, the degree to which the results depend on that assumption cannot be determined from the registry record.

16. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: chest pain (0.7%), transient ischaemic attack (0.7%), cholelithiasis (0.7%), bacterial infection (0.4%), pyelonephritis acute (0.4%).

The registry reports serious adverse event counts and other adverse event counts by treatment group. For serious adverse events, the listed affected counts are 12 among 273 at risk for empagliflozin 25 mg/linagliptin 5 mg, 16 among 272 for empagliflozin 10 mg/linagliptin 5 mg, 19 among 276 for empagliflozin 25 mg, 16 among 275 for empagliflozin 10 mg, and 10 among 267 for linagliptin 5 mg. The listed other adverse event counts are 80 among 273, 86 among 272, 70 among 276, 83 among 275, and 94 among 267, respectively. These are group-level counts, not adjusted comparisons.

The registry lists several serious adverse event terms with percentages: chest pain, transient ischaemic attack, and cholelithiasis each at 0.7%; bacterial infection, acute pyelonephritis, and benign parathyroid tumour each at 0.4%. The group attribution and full event-level detail are not reported alongside these terms in the record. Death counts are not reported in the registry record.

These counts do not by themselves establish that one treatment has a different safety profile from another. The record does not provide a comparative statistical analysis of adverse events, complete event timing, severity categories beyond the serious-event classification, or enough context to assess exposure time. Rare or infrequent events may also be difficult to compare using counts alone. Safety interpretation should remain limited to the events and group counts posted in the registry.

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

The posted primary outcome shows negative average changes in HbA1c from baseline to week 24 across all five metformin-background groups. The sponsor’s ANCOVA analyses report selected contrasts between the fixed-dose combinations and individual components, with confidence intervals and p-values as recorded. Those results describe the specified population, measurement, and comparisons; they do not establish every possible pairwise difference or describe all participants in the trial.

A statistically notable comparison is not automatically a measure of clinical importance. The registry does not provide a threshold for a clinically meaningful HbA1c difference, nor does it report enough broader clinical outcome information to assess whether the observed changes translated into effects on complications, symptoms, quality of life, or other outcomes. The primary endpoint is a laboratory measure, not a direct count of patient-important events.

The findings also do not establish equivalence between treatments when a comparison is not statistically significant. A confidence interval that includes zero can still contain effects in either direction; the interval’s width indicates how much uncertainty remains. Likewise, the absence of a reported difference is not proof that treatments have identical effects. An equivalence or noninferiority conclusion would require a prespecified margin and an appropriate analysis, neither of which is reported in the registry record.

The trial included treatment-naïve and metformin-background patients, but the primary HbA1c results pertain to the metformin-background analysis population. Findings should not automatically be generalized to the treatment-naïve group, other patient populations, other treatment durations, or settings not described in the record. The registry does not report enough information to assess all aspects of adherence, concomitant care, or the representativeness of enrolled patients. The results are statistical descriptions of the posted trial outcomes, not medical advice or a recommendation about treatment.

18. Why This Trial Matters Statistically

This trial illustrates why a clinical result must be read alongside its population, endpoint definition, analysis method, and contrast. The primary outcome is a continuous change measure in a specified metformin-background Full Analysis Set, while the broader study also includes treatment-naïve patients and several secondary outcomes. Keeping those populations and endpoints distinct prevents a result from being generalized beyond the analysis that produced it.

It also demonstrates the difference between adjusted sponsor analyses and unadjusted recalculations from aggregate summaries. The sponsor’s ANCOVA comparisons and the recalculated contrasts are not interchangeable; their estimates can reflect different adjustment and contrast conventions. The registry flags a discrepancy, making it especially important to distinguish the trial’s reported analysis from a reconstruction based on summary statistics. The sponsor’s own analysis takes precedence for the reported trial comparisons.

Finally, the record underscores the role of uncertainty and design assumptions. Confidence intervals show the range of estimates compatible with a particular analysis, while sample-size illustrations depend on assumptions about the effect to detect, allocation, significance level, and statistical approximation. Those planning figures are not observed power and do not reveal the trial’s original design justification. Missing multiplicity, missing-data, and full model details also define the limits of what can be concluded from the registry record. Careful statistical reading makes clear both what the results indicate and what remains unresolved.

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.

Several limitations specific to this trial record affect interpretation. The primary endpoint is reported for metformin-background patients rather than for the entire enrolled population. The primary analysis uses a Full Analysis Set definition requiring treatment and both baseline and at least one on-treatment HbA1c value; the extent to which this population differs from all randomized patients cannot be determined from the registry record.

The registry reports LOCF for the primary endpoint but does not give missingness counts, reasons for missing measurements, or sensitivity analyses. The detailed ANCOVA model, including its covariates and other specifications, is also not reported. As a result, the sponsor analysis cannot be independently reproduced from the record, and its relation to the unadjusted recalculations cannot be fully resolved. The record flags a discrepancy, so the sponsor’s reported results take precedence.

The registry does not report whether multiplicity was controlled across the primary and secondary analyses, nor does it provide the original statistical analysis plan or original sample-size assumptions. Some secondary outcome categories are listed without numerical analyses. Safety information is presented as counts and selected terms, without a complete comparative analysis, event timing, or death counts. These gaps prevent a comprehensive assessment of robustness, secondary hypotheses, and comparative safety from the registry record alone.

The trial title describes a 52-week study, but the primary endpoint discussed is the baseline-to-week-24 HbA1c change. The information in the record does not establish whether the week 24 results capture later outcomes or the full duration of treatment. These are limitations of the posted record and analyses, not evidence that an unreported result is absent from other sources.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Efficacy and Safety of Empagliflozin (BI 10773) / Linagliptin (BI 1356) Fixed Dose Combination in Treatment naïve and Metformin Treated Type 2 Diabetes Patients — NCT01422876.
• Registry record last updated 2015-04-02; status last verified 2015-03.
• Results first posted 2015-04-02.
• 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 record describes a completed phase III randomized, double-blind, parallel-group study sponsored by Boehringer Ingelheim, with enrollment of 1405. It evaluated once-daily empagliflozin/linagliptin fixed-dose combinations against individual components in treatment-naïve and metformin-background patients with type 2 diabetes. The primary endpoint was week 24 change from baseline in HbA1c among metformin-background patients in a Full Analysis Set using LOCF.

The sponsor reports ANCOVA comparisons for selected primary and secondary outcomes. Unadjusted recalculations from posted summaries differ in method and are not a substitute for the sponsor analysis; the record flags a discrepancy, so sponsor results take precedence. The record also posts secondary outcome summaries and adverse event counts, but leaves important details about multiplicity, missingness, model specification, and original design assumptions unreported.