Safety and Efficacy in Type 2 Diabetic Patients With Severe Chronic Renal Impairment, 5 mg BI 1356 (Linagliptin) vs. Placebo, Insulin Background Inclusive: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Efficacy in Type 2 Diabetic Patients With Severe Chronic Renal Impairment, 5 mg BI 1356 (Linagliptin) vs. Placebo, Insulin Background Inclusive (NCT00800683), a randomized phase 3 study. The primary outcome is HbA1c Change From Baseline at Week 12.

2. Clinical Question

The question is about a between-group contrast in change, not simply whether either group’s HbA1c changed over time. The registry describes the primary measure as Week 12 HbA1c percent minus baseline HbA1c percent. A negative change therefore denotes a lower measured HbA1c at Week 12 than at baseline; the comparison asks whether that change differed between groups.

The record also identifies outcomes at later follow-up points, including HbA1c changes and a treat-to-target response measure. Those outcomes provide additional context about the pattern over follow-up, but they are not the primary question. The registry record does not report a result for the listed response criterion of an HbA1c under 6.5% after 52 weeks.

This is a question about the trial’s reported measurements and analyses. It does not, by itself, establish how the findings should guide care for an individual or describe outcomes beyond those recorded in the trial.

3. Trial Design

The registry classifies this as a Phase 3, randomized, double-masked, parallel-group trial. Participants were assigned to a once-daily BI 1356 tablet or an identical placebo tablet. The title identifies the BI 1356 dose as 5 mg and describes the study as including an insulin background. The record does not give further details about background treatment or how it was managed.

The listed condition is type 2 diabetes, while the title specifies severe chronic renal impairment as the population of interest. Enrollment was 133. For the primary HbA1c outcome, the posted group summaries include 62 participants in the placebo group and 66 in the linagliptin group. These outcome-specific counts should be read as the numbers contributing to those posted summaries, not as a replacement for the overall enrollment figure.

The study began in 2008-12, had primary completion in 2011-01, and is marked completed. The registry describes the primary analysis population as the Full Analysis Set: treated and randomized participants with baseline and at least one on-treatment HbA1c measurement available. Last observation carried forward was used for imputation in that population.

The posted primary comparison is an ANCOVA analysis. The outcome description says the treatment means were adjusted for baseline continuous HbA1c, baseline creatinine clearance, and previous anti-diabetic medication. Details such as the randomization procedure, allocation concealment, the full protocol, the trial’s original sample-size justification, and any interim analyses are not reported in the registry record. The parallel-group structure supports a direct comparison of the assigned groups, while masking is intended to limit knowledge of assignment during the trial; the record does not provide further operational details about masking.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Full Analysis Set (FAS) included all treated and randomised participants with a baseline and at least one on-treatment HbA1c measurement available. Last observation carried forward (LOCF) was used as the imputation rule.

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.

6. Primary Endpoint

The registry lists the primary outcome as HbA1c Change From Baseline at Week 12, measured over Baseline and Week 12.

The primary endpoint is change in HbA1c from baseline to Week 12, measured in percent. HbA1c is a continuous measure, and the registry defines the change as the Week 12 value minus the baseline value. The reported arm summaries are least-squares means, with standard errors, rather than unadjusted raw averages. The outcome description says these means were adjusted for baseline HbA1c, baseline creatinine clearance, and previous anti-diabetic medication.

In general, a change-from-baseline endpoint expresses how a measured value differs from its starting value during follow-up. Comparing changes between randomized groups can help distinguish the groups’ observed trajectories from change that occurs in both groups. Here, the central statistical quantity is the between-group difference in mean change at the specified follow-up point.

HbA1c is used as an efficacy measure because it is the explicitly named primary outcome in the registry record. The endpoint is not a measure of every aspect of diabetes or of overall safety. Nor does a result at Week 12 alone describe longer-term outcomes. The registry lists later HbA1c assessments, but those are secondary outcomes rather than substitutes for the stated primary endpoint.

The Full Analysis Set requires a baseline and at least one on-treatment HbA1c measurement, and the record states that last observation carried forward was used for missing measurements. Those choices affect which participants contribute and how missing follow-up values are represented. The registry record does not report additional sensitivity analyses that would show how the primary result changes under other missing-data assumptions.

7. Results Posted in the Registry

• Linagliptin (BI 1356): mean -0.76 versus -0.15 on control; difference -0.61 (95% CI -1.01 to -0.21), standardized effect d = -0.53; z-test p = 0.003.

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.

HbA1c Change From Baseline at Week 52

• 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:
  • Linagliptin (BI 1356): mean -0.71 versus 0.01 on control; difference -0.72 (95% CI -1.15 to -0.29), standardized effect d = -0.58; z-test p = 0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Linagliptin (BI 1356) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

HbA1c Change From Baseline at Week 18

• 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:
  • Linagliptin (BI 1356): mean -0.57 versus 0.04 on control; difference -0.61 (95% CI -1.00 to -0.22), standardized effect d = -0.54; z-test p = 0.002.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Linagliptin (BI 1356) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

HbA1c Change From Baseline at Week 24

• 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:
  • Linagliptin (BI 1356): mean -0.64 versus 0.04 on control; difference -0.68 (95% CI -1.05 to -0.31), standardized effect d = -0.63; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Linagliptin (BI 1356) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists several later HbA1c change assessments, including Week 18, Week 24, and Week 52, as well as other follow-up times. The posted results at the displayed time points show a broadly consistent direction in the group comparisons: the estimated change in the linagliptin group is lower than the placebo estimate. This pattern is context for the primary Week 12 finding, not a replacement for it.

These secondary outcomes relate to the primary result by showing whether the direction of the HbA1c comparison is also present at later assessments. They do not establish that every secondary measure was prespecified in the same way, nor do they independently confirm the primary analysis. The registry lists a treat-to-target response outcome based on an HbA1c threshold after 52 weeks, but its result is not reported in the registry record.

Secondary findings generally carry less weight than a primary endpoint when many outcomes or time points are examined. In general, the more statistical comparisons that are considered, the greater the chance that at least one result will appear noteworthy by chance. A multiplicity procedure can address that issue, but whether the trial used such an approach is not reported in the registry record.

The secondary results are therefore best read as supportive descriptive context, with attention to the analysis method and uncertainty for each outcome. The record includes different kinds of reported summaries among the secondary analyses, and the posted analysis details do not establish a common multiplicity-adjusted interpretation across them. The tables show the individual estimates; this discussion emphasizes their overall pattern rather than treating each result as a separate definitive conclusion.

More broadly, consistency across follow-up assessments can be informative, but repeated measurements from the same trial are not independent replications. Their interpretation depends on the trial’s analysis plan, handling of repeated outcomes, missing data, and any correction for multiple testing. Those details are not reported in the registry record.

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

• HbA1c Change From Baseline at Week 52
• HbA1c Change From Baseline at Week 18
• HbA1c Change From Baseline at Week 24
• HbA1c Change From Baseline at Week 30
• HbA1c Change From Baseline at Week 36
• HbA1c Change From Baseline at Week 42
• HbA1c Change From Baseline at Week 48
• The Occurrence of Treat to Target Efficacy Response, That is an HbA1c Under Treatment of <6.5% After 52 Weeks of Treatment

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: it records a numerical change in HbA1c percent. The registry reports least-squares means for each group and a between-group analysis. In general, a least-squares mean is a model-based estimate of a group mean that accounts for covariates included in the analysis. The outcome description names baseline HbA1c, baseline creatinine clearance, and previous anti-diabetic medication as adjustment variables. The record does not give the complete model specification or additional details about how those variables were represented.

What does the ANCOVA analysis compare?

ANCOVA, or analysis of covariance, compares outcome values between groups while accounting for specified covariates. In this trial the sponsor’s posted analysis is described as ANCOVA and reports a mean difference in final values. The outcome definition separately describes the endpoint as change from baseline and says the treatment means were adjusted for the listed baseline factors. These descriptions indicate that the reported comparison is model-based; the registry does not provide enough detail to reconstruct every element of the sponsor’s model.

In general, covariate adjustment can improve precision when the covariates are related to the outcome and can account for measured baseline differences. It does not adjust for variables that were not included or measured. Randomization is the design feature that supports comparing assigned groups, while the model specifies how the reported estimate is calculated. Neither feature removes uncertainty from the estimate.

How should the change score be read?

The registry defines the change as the Week 12 HbA1c percent minus the baseline HbA1c percent. Each group’s least-squares mean summarizes its adjusted average change. The between-group estimate then compares those averages. A negative estimate for linagliptin versus placebo means the estimated change in the linagliptin group is lower than the estimated change in the placebo group. It is a contrast between group averages, not a description of each participant’s response.

The posted standard errors describe uncertainty in the arm-level least-squares means. In general, a standard error reflects how much an estimate would vary across repeated samples under the model and its assumptions. It is not the standard deviation of individual participant measurements. This distinction matters: a small standard error can accompany substantial variation among individuals because the two quantities describe different features of the data.

What is the role of the recalculation?

The recalculated comparison uses the posted group summaries to form a simple, unadjusted estimate and interval. The calculation uses summary statistics, including standard errors, and an approximate pooled standard deviation. It is not a reconstruction of the sponsor’s adjusted ANCOVA. The recalculated p value is also unadjusted. In general, a summary-statistic recalculation is useful as a transparent check on direction and approximate magnitude, but it cannot reproduce an analysis that depends on participant-level data or model details that are not posted.

The sponsor’s reported result is the trial’s own analysis. Because the registry does not flag a discrepancy, the close direction and magnitude of the two estimates can be described without treating the recalculation as a correction. The sponsor analysis remains the appropriate source for the reported ANCOVA result; the recalculation serves as a separate, simplified comparison.

What do confidence intervals and p values convey?

In general, a confidence interval gives a range of values compatible with the observed data and the statistical procedure at a stated confidence level. Its width reflects precision: wider intervals allow a broader range of effects, while narrower intervals indicate greater precision under the same assumptions. The posted primary sponsor interval is a 95% interval, as is the interval shown for the recalculation.

In general, a p value describes how unusual data at least as incompatible with a specified null model would be if that model and the analysis assumptions were correct. A small p value is not the probability that the null hypothesis is true. Nor does it show whether an observed difference is important in practice. The reported sponsor and recalculated p values come from different analyses, so they should be interpreted in their own contexts.

How were participants and missing measurements handled?

The Full Analysis Set includes treated and randomized participants with a baseline and at least one on-treatment HbA1c measurement. The record states that last observation carried forward was the imputation rule. In general, this method uses a participant’s last available measurement to fill a later missing measurement. Its assumptions may not hold in every setting, particularly if the outcome would otherwise change after the last observed value. The registry does not report sensitivity analyses using alternative missing-data methods, so the effect of that choice cannot be assessed.

The posted outcome counts are not identical to the total enrollment figure. That is not by itself a contradiction: enrollment and endpoint-specific analysis counts refer to different stages or requirements. The registry record does not give a full participant flow in the registry record, so reasons for missing endpoint measurements or differences in analysis denominators are not reported in the registry record.

What cannot be learned from the posted methods?

The registry identifies randomization, masking, parallel groups, the primary analysis population, the imputation rule, and the ANCOVA method. It does not report the full statistical analysis plan, original sample-size justification, allocation-concealment details, interim-analysis plan, multiplicity adjustment, or detailed diagnostics. Those omissions limit the ability to evaluate all modeling choices and the prespecification of every analysis. General statistical explanations on this page describe common concepts; they do not fill in those unreported design details.

11. Confidence Intervals and Precision

The sponsor’s primary estimate is -0.60, with a 95% confidence interval from -0.89 to -0.31. The interval is entirely below zero, so the posted ANCOVA estimate is compatible with a negative between-group difference under the analysis assumptions. The recalculated interval, from -1.012 to -0.208, also lies below zero. Its wider span reflects the uncertainty calculated by that separate summary-statistic method; it should not be read as the sponsor model’s interval.

In general, a confidence interval communicates both the estimated direction and the precision of an effect estimate. It is not a range containing a fixed proportion of individual participants’ responses. Nor does a 95% interval mean there is a 95% probability that the true effect lies inside this particular interval. That interpretation would require a different statistical framework.

The two intervals offer a useful comparison of the reported model and the recalculation, but they are not identical because the analyses differ. The sponsor reports ANCOVA, whereas the recalculation is an unadjusted approximation based on posted summaries. The interval endpoints should therefore be kept with their respective estimates and methods.

The group standard errors also help describe precision of the mean changes, but they do not reveal the full distribution of participant-level outcomes. A confidence interval can be relatively informative about an average while leaving substantial uncertainty about individual responses. The registry record does not provide additional precision measures for all individual-level outcomes or a detailed distribution of HbA1c changes.

Precision is distinct from clinical importance. An interval can indicate a difference from zero without establishing how consequential that difference is. The registry record does not provide a threshold for interpreting the size of the change, so this page describes the numerical estimate and its uncertainty rather than assigning a clinical category to it.

12. Planning a Similar Trial: Sample Size

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

The sample-size display is a planning illustration based on the observed effect, not a record of the trial’s original design calculation. The figures are participants per arm, not total participants and not total events. The assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect.

The table presents planning scenarios for detecting 100%, 75%, or 50% of the observed effect. For each scenario, the 80% power column indicates the number of participants per arm in a hypothetical study designed to have an 80% chance of detecting that effect under the stated assumptions. The 90% power column gives the corresponding planning number for a 90% chance. These are probabilities attached to a hypothetical design and analysis, not guarantees about a future study’s result.

Smaller effects are harder to distinguish from random variation with the same precision, so the planning illustration requires more participants per arm when the target is a smaller fraction of the observed effect. The values should be read as conditional on the stated assumptions, not as a recommendation or as evidence that the original trial was adequately or inadequately sized.

This is not observed power. It is a planning exercise using an effect estimate observed in the posted results; the trial’s original sample-size assumptions and justification are not reported in the registry record. The display also does not incorporate any adjustments for missing data, multiple outcomes, or model features beyond the assumptions. The curve visualizes the same planning relationship and should be interpreted on those terms.

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 = -0.610; pooled standard deviation σ = 1.159.
2. n = 2 × (1.96 + 0.842)² × 1.159² / -0.610² = 56.6, rounded up to 57 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 <- -0.61                      # observed difference in means
sigma <- 1.15874521625355                     # 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(800683)
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 = 57 (80% power), 76 (90% power)
 75% of observed effect: n per arm = 101 (80% power), 135 (90% power)
 50% of observed effect: n per arm = 227 (80% power), 304 (90% power)
power.t.test: 57.6 
Simulated power at n = 57 per arm: 0.799 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The sponsor’s reported primary analysis is an ANCOVA comparison of final values. It gives a mean difference of -0.60, a 95% confidence interval from -0.89 to -0.31, and p < 0.001. This is the trial’s own reported analysis and should be distinguished from the simple recalculation displayed alongside it.

The recalculation estimates a difference of -0.61, with a 95% confidence interval from -1.012 to -0.208 and an unadjusted p value of 0.003. Its estimate is close in direction and magnitude to the sponsor estimate, but the methods are not the same. The recalculation is based on posted group summaries and is unadjusted; it cannot reproduce the sponsor’s ANCOVA using the full analysis and adjustment structure.

The primary outcome description says the means were adjusted for baseline HbA1c, baseline creatinine clearance, and previous anti-diabetic medication. The registry does not provide enough detail to reproduce the full sponsor model. Differences between the recalculated and sponsor-reported intervals or p values should therefore be understood as differences between analyses, not as evidence that one number is a correction of the other.

The registry’s discrepancy indicator is false. The posted sponsor result and the recalculation agree on the direction of the estimated contrast. For interpretation of the trial’s planned analysis, the sponsor-reported ANCOVA result is the relevant reported result; the recalculation is a transparent, simplified check using the available summaries.

15. Multiplicity, Missing Data and Analysis Populations

The record identifies a primary endpoint and multiple secondary outcomes, including HbA1c changes at several follow-up times and a treat-to-target response measure. It also reports multiple outcome measures and analyses overall. When a trial evaluates multiple endpoints or time points, the chance of at least one apparently notable result can increase. In general, multiplicity procedures are used to control or describe that risk, depending on the study’s objectives and analysis plan.

Whether this trial prespecified a multiplicity strategy, adjusted secondary p values, or used a testing hierarchy is not reported in the registry record. The posted p values should therefore be read as the values attached to their individual reported analyses, not as evidence that the full family of secondary comparisons has been controlled for multiplicity. The primary result should remain distinct from the collection of secondary results.

The primary population is described as the Full Analysis Set: treated and randomized participants with a baseline and at least one on-treatment HbA1c measurement. The registry states that last observation carried forward was used as the imputation rule. In general, carrying the last observed value forward assumes that this value is an acceptable stand-in for the missing later measurement. If the underlying outcome would have changed, that assumption can affect estimated group means and uncertainty.

The registry record does not report the amount or pattern of missing HbA1c data, reasons for missingness, or sensitivity analyses using different assumptions. It also does not provide enough information to assess whether missingness differed between groups. Those details matter because the imputation method alone does not establish how robust the result is to missing data.

The record reports outcome-specific group counts and overall enrollment, but a full participant flow and the reasons participants were not included in particular summaries are not reported in the registry record. The original sample-size justification, interim-analysis plan, full statistical analysis plan, and other analysis-population definitions are also not reported in the registry record. These absences limit assessment of prespecification and robustness; they do not, by themselves, show that a particular analysis was inappropriate.

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: renal failure acute (7.4%), angina pectoris (4.6%), hyperkalaemia (4.6%), pulmonary congestion (4.6%), acute myocardial infarction (4.4%).

The adverse-event summary reports serious adverse events affecting 27 of 65 participants at risk in the placebo group and 25 of 68 participants at risk in the linagliptin group. The other adverse-event summary reports affected counts of 50 of 65 and 61 of 68, respectively. These are group-level counts as posted; the registry entries do not give additional definitions or adjudication details for these categories.

The listed serious-event terms include renal failure acute, angina pectoris, hyperkalaemia, pulmonary congestion, acute myocardial infarction, and pneumonia, each with a percentage shown in the registry. The list does not provide enough detail to assign each listed percentage to a specific treatment group or to infer comparative risk for an individual event.

In general, an adverse event recorded during a trial is not necessarily caused by the assigned treatment. Counts also do not describe event severity, duration, recurrence, or the clinical context unless those details are reported. The registry has no death counts posted for these groups, so deaths are not reported in the registry record. The summaries should be read as recorded safety data, not as proof of cause or a complete safety assessment.

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

The primary result shows a lower estimated mean HbA1c change in the linagliptin group than in the placebo group at Week 12. The sponsor’s ANCOVA estimate is -0.60, and its 95% confidence interval is from -0.89 to -0.31. The recalculation from the posted summaries points in the same direction. These results describe the comparison in the analyzed trial population under the stated methods.

They do not show that every participant experienced the same change, and they do not establish the outcome for an individual. Group means summarize average measurements; standard errors and confidence intervals describe uncertainty around those averages rather than the full spread of participant responses. The registry record does not provide a threshold for judging the clinical importance of the estimated difference.

The primary endpoint is a Week 12 HbA1c change, not a measure of all health outcomes or all dimensions of safety. Later HbA1c outcomes provide additional follow-up context, but the listed treat-to-target response result is not reported in the registry record. The reported analysis also uses a defined Full Analysis Set and last observation carried forward for imputation; the robustness of the findings to other missing-data assumptions is not reported in the registry record.

The trial title identifies people with type 2 diabetes and severe chronic renal impairment, and the posted findings should be interpreted within that study context. The record does not provide enough information to extend the estimates to populations with different characteristics, treatment backgrounds, or follow-up. It also does not report the detailed protocol or full analysis plan needed to assess every design and analysis decision.

Finally, statistical evidence for a difference is not itself a treatment recommendation. The results describe the posted trial estimates and their precision. They do not establish that a treatment works or does not work in every setting, nor do they replace an assessment of benefits, harms, and individual circumstances.

18. Why This Trial Matters Statistically

This trial is statistically informative because it makes a randomized, masked comparison of a continuous outcome in a population identified as having severe chronic renal impairment. The primary question is defined in terms of change from baseline, and the record reports both group-level estimates and a between-group analysis. That structure makes it possible to distinguish a comparison of average changes from a description of change within either group alone.

The record also illustrates why analysis methods matter. The sponsor reports an adjusted ANCOVA result, while the recalculation uses posted summaries to produce a simpler, unadjusted estimate. Their directions are consistent, but their confidence intervals and p values are not interchangeable. A reader needs to keep the model-based trial result separate from a summary-statistic check.

The Full Analysis Set definition and the last-observation-carried-forward rule are also important to interpretation. These specify who contributes to the primary analysis and how missing measurements were handled. Without a report of missing-data patterns or sensitivity analyses, readers cannot fully evaluate how much the imputation choice may have affected the result.

Multiple HbA1c outcomes and follow-up times make the secondary pattern relevant, while also raising the need to understand how multiplicity was handled. The registry record does not report that plan. This is a useful reminder that a collection of individually reported results should not automatically be treated as a single multiplicity-controlled body of evidence.

Finally, the displayed planning figures show how sample-size requirements depend on the effect a future study aims to detect and the chosen power target. They are planning illustrations based on an observed effect, not a reconstruction of the original trial design. Together, these features show how endpoint definition, analysis population, uncertainty, missing data, and multiplicity shape the statistical reading of a clinical trial.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The registry record gives an outline of the design and selected results, but not the complete protocol or statistical analysis plan. Details not reported in the registry record include the original sample-size justification, the randomization and allocation-concealment procedures, interim analyses, and a multiplicity strategy. These omissions limit evaluation of prespecification and the full conduct of the analysis.

The primary analysis uses a defined Full Analysis Set and last observation carried forward. The record does not report the number or pattern of missing measurements, reasons for missingness, or sensitivity analyses with alternative approaches. Consequently, the impact of the missing-data rule cannot be assessed from the posted information alone.

The sponsor’s analysis is reported as ANCOVA, but the full model specification and participant-level data are not available in the registry entries. The summary-statistic recalculation is approximate and unadjusted; it cannot recreate the sponsor’s analysis. The two estimates should therefore be interpreted according to their methods rather than combined as if they were the same calculation.

There are also limits to the safety summaries. The record lists affected counts and several serious-event terms, but does not provide enough detail to evaluate event timing, duration, causality, or the group assignment for each listed percentage. Death counts are not reported in the registry record. These summaries alone cannot establish a complete comparative safety profile.

The trial’s title describes severe chronic renal impairment and an insulin-inclusive background, but the record gives limited detail about participant characteristics and background treatment. Generalizability beyond the trial population and treatment setting cannot be assessed fully. The registry also does not report the result for the listed Week 52 treat-to-target response measure.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Safety and Efficacy in Type 2 Diabetic Patients With Severe Chronic Renal Impairment, 5 mg BI 1356 (Linagliptin) vs. Placebo, Insulin Background Inclusive — NCT00800683.
• Registry record last updated 2014-05-20; status last verified 2014-05.
• Results first posted 2012-02-01.
• 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

This completed Phase 3 trial randomized participants to once-daily BI 1356, identified as linagliptin, or an identical placebo tablet. The title specifies 5 mg and describes the population as people with type 2 diabetes and severe chronic renal impairment. The trial was double-masked and used a parallel-group design.

The primary endpoint was HbA1c change from baseline to Week 12. The sponsor reported an ANCOVA mean difference of -0.60, with a 95% confidence interval from -0.89 to -0.31 and p < 0.001. A separate unadjusted recalculation from the posted summaries estimated a difference of -0.61, with a 95% confidence interval from -1.012 to -0.208.

The record defines the Full Analysis Set and states that last observation carried forward was used for missing measurements. It lists later HbA1c outcomes and adverse-event summaries, but key details about multiplicity, missing-data sensitivity analyses, original sample-size planning, and the full statistical plan are not reported in the registry record. The posted results describe this trial’s measurements and analyses; they do not by themselves establish clinical importance or individual outcomes.