Real World Glycemic Effectiveness of Linagliptin: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Real World Glycemic Effectiveness of Linagliptin (NCT03338803), a registered phase not stated study. The primary outcome is Change in Glycosylated Hemoglobin (HbA1c) Across Age Categories.

2. Clinical Question

The registry frames the study as assessing whether linagliptin’s glycemic effectiveness is comparable across age and renal function groups. The stated primary outcome focuses on HbA1c change across age categories. The record lists the age categories as 40 to 54 years, 40 to 54 years, 65 to 74 years and 75+ years; the repeated category is part of the registry wording, and the record does not clarify it. Renal-function comparisons are described among secondary outcomes, but their category definitions are not reported in the registry record.

This question is addressed using existing clinical records rather than treatment assignment. As a result, the registry description supports a question about observed HbA1c changes in patients prescribed linagliptin, not a randomized estimate of the effect of linagliptin compared with another treatment or with no treatment. The record does not provide endpoint estimates, so it does not answer whether the observed changes differed across age groups.

3. Trial Design

This is a completed, non-interventional cohort study sponsored by Boehringer Ingelheim. The registry reports enrollment of 11001 and describes use of existing data from the Optum Clinical Database, which contains electronic health record data from providers across the United States. The study start date is 2017-11-21, and the primary completion date is 2018-02-02. The patient identification period is described separately as 01 January 2012 through 30 September 2016.

The described cohort consists of patients with type 2 diabetes who had a written prescription for linagliptin. No randomized assignment is described, and no like-for-like control arm is specified. Thus the registry describes an observational cohort of linagliptin initiators, not a randomized comparison. The record does not report how patients were selected from the database beyond the stated prescription and identification period, nor does it give a flow of patients from identification to analysis.

The primary outcome is mean change in HbA1c across age categories, measured from baseline to 60 to 180 days. Two secondary outcomes concern the percentage of adults achieving HbA1c < 7.0% across age categories and across renal-function categories. The record lists 4 outcome measures but lists 0 analyses. No result estimates, analysis population counts, or group-specific summaries are reported in the registry record. The completed status therefore describes the study’s registry status; it does not itself provide evidence about the size or direction of HbA1c change.

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: Type 2 diabetes mellitus (T2DM) patients with written prescription for initiating Linagliptin (Tradjenta®, Jentadueto®, or Jentadueto XR®) were identified during identification period starting on 01 January 2012 and ending on 30 September 2016. Aim was to describe real world glycemic-lowering effectiveness across range of ages and renal function.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Patients with a written prescription for linagliptin were identified during an identification period starting on 01 January 2012 and ending on 30 September 2016.

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: Patients with a written prescription for linagliptin were identified during an identification period starting on 01 January 2012 and ending on 30 September 2016.

6. Primary Endpoint

The registry lists the primary outcome as Change in Glycosylated Hemoglobin (HbA1c) Across Age Categories, measured over Baseline and 60 to 180 days.

The primary endpoint is change in glycosylated hemoglobin (HbA1c), reported as a mean in percentage units, across age categories. The timeframe is baseline and 60 to 180 days. The registry defines each patient’s change by subtracting the last (most recent) HbA1c value during the pre-index period, including the index date, from the last (most recent) HbA1c value 60 to 180 days after the index date. Under that definition, a negative change would indicate a lower follow-up value than baseline; this is an interpretation of the stated subtraction order, not a reported study result.

In general, HbA1c is used as a measure of glycemic status, and a change endpoint expresses how a continuous measurement differs between two periods for the same person. Reporting a mean describes the average change, while a standard deviation describes the spread of individual changes around that average. The registry specifies standard deviation as the dispersion measure. Neither the mean change nor its standard deviation is reported in the registry record, so the magnitude, direction and variability of the primary endpoint cannot be assessed from the posted results.

The listed age categories include a repeated 40 to 54 years entry. The record does not clarify the intended category structure, and the primary outcome description does not provide subgroup estimates. It therefore does not support conclusions about differences in mean HbA1c change among the intended age groups.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the mean endpoint represent?

The primary outcome is a continuous measure: each patient’s HbA1c change, summarized by a mean across patients in an age category. In general, a mean provides a measure of the average value in the analyzed group. It is useful for describing the average change, but it does not show whether individual patients had similar changes or widely varying ones. The registry names standard deviation as the measure of dispersion, which would describe the spread of patient-level changes around the mean. Neither quantity is posted for this outcome.

The endpoint definition uses the most recent pre-index HbA1c, including the index date, and the most recent HbA1c during 60 to 180 days after the index date. The follow-up measurement is compared with the pre-index measurement by subtracting the latter from the former. In general, using a change score makes the contrast within a patient explicit. Its interpretation still depends on which patients have both measurements and on how the measurement times vary. The record does not report the number with paired measurements or the distribution of the timing of those measurements.

What would uncertainty estimates tell a reader?

In general, a confidence interval expresses the range of estimates reasonably compatible with the data and the statistical procedure used. A narrow interval indicates greater precision than a wide interval, though neither alone resolves design limitations or bias. A p-value, when reported, describes how compatible the data are with a specified null model under the assumptions of the test; it is not a measure of clinical importance. The registry does not post confidence intervals or p-values for the primary endpoint, and it does not describe a test or model. No inference about statistical significance or precision is available.

Standard deviation and confidence interval answer different questions. Standard deviation describes variation among observations; a confidence interval describes uncertainty in an estimated quantity. The registry specifies standard deviation for dispersion but gives no observed value. It also does not report a standard error or an interval estimate. As a result, readers cannot judge whether an average change, if calculated, was estimated precisely or whether individual changes were concentrated around that average.

How does the cohort design affect interpretation?

The study uses existing electronic health record data and identifies patients with written prescriptions for linagliptin. In general, observational data describe care as it occurred, rather than outcomes under randomized assignment. Differences in health, treatment history, follow-up, or measurement availability can be associated with both cohort membership and outcomes. These are general concerns for observational analyses; the record does not report whether or how any such factors were measured or addressed in this study.

No comparison cohort is specified in the registry description. Without a described control group, the posted question is about change among linagliptin initiators, rather than a between-group contrast. A before-and-after change can be described if results are reported, but by itself it does not establish that treatment caused the change. In this record, even the descriptive estimate is absent.

What is and is not documented about the analysis?

The registry identifies the primary endpoint, its unit, the mean as the summary parameter, standard deviation as dispersion, and the follow-up timeframe. It does not report the statistical model, covariates, subgroup estimation procedure, assumptions, or approach to repeated or unevenly timed measurements. It also lists 0 analyses, so there are no posted analysis results to interpret. The specific statistical analysis plan is not reported in the registry record.

Over-dispersion is a general concern when observed variation is greater than a model expects. If a model assumes too little variability, its uncertainty estimates may be too small. The registry identifies a continuous outcome with standard deviation, but does not report a model or data from which over-dispersion could be assessed. It would not be appropriate to assert that over-dispersion occurred or that a particular adjustment was used.

The record includes planning assumptions for a sample-size illustration: a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect to detect defined as a stated fraction of the observed effect. It describes participant counts per arm for non-time-to-event illustrations and total event counts for time-to-event illustrations. These are planning assumptions, not evidence that the study used those methods or had those design features. The study description does not identify two randomized arms, and no calculated planning counts or observed-effect value are reported in the registry record. Such illustrations are not observed power and cannot substitute for endpoint results.

What does the age-category wording imply?

The primary description lists 40 to 54 years twice, followed by 65 to 74 years and 75+ years. Because the registry does not explain the repeated entry, an intended category cannot be inferred or silently corrected. This ambiguity matters for interpretation: a reader cannot determine from the record alone how the planned age groupings were meant to be represented. The secondary outcome refers to renal-function categories, but their definitions and analysis are not reported either.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 4 outcome measures, including a primary HbA1c-change outcome and secondary outcomes based on achieving HbA1c < 7.0% across age and renal-function categories. It does not report a multiplicity strategy, such as whether testing across several outcomes or subgroups was adjusted or interpreted under a prespecified hierarchy. The statistical plan is not reported in the registry record. Without posted analyses and a stated approach, the possibility of multiple comparisons cannot be evaluated from the record.

Missing-data handling is also not described. The primary endpoint requires a pre-index measurement and a later measurement in the 60 to 180 days after the index date. The registry does not state how patients without either measurement were handled, whether incomplete records were excluded, or whether any missing-data method was used. It does not report the number with available measurements or the analytic denominator for any outcome. The enrollment figure of 11001 should not be treated as the endpoint-specific analysis population.

The record does not define an analysis population beyond the description of patients with a written prescription for linagliptin. It does not report whether analyses included all identified patients, only patients with follow-up data, or another subset. No adjusted analysis population or subgroup counts are listed, and the record lists 0 analyses. These omissions prevent evaluation of attrition, representativeness, missingness patterns, and the consequences of any exclusions. The registry record also does not report interim analyses or stopping rules; their presence or absence should not be inferred from the completed status.

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

The record establishes that the study asked about HbA1c change after linagliptin initiation and whether that change varied across age categories. It also identifies secondary outcomes concerning the percentage achieving HbA1c < 7.0% across age and renal-function categories. These descriptions define the intended clinical measurements, but they are not outcome findings. The registry does not report numerical results for these measures.

Accordingly, the posted information does not show whether mean HbA1c increased or decreased, how large any change was, how variable individual changes were, or whether the data were compatible with different changes across subgroups. It also does not show what proportion achieved the stated HbA1c threshold. No precision estimates or statistical tests are reported. The absence of results in the registry record is not evidence that the outcome was unchanged or that a comparison was not statistically significant.

The observational design also limits what a result could establish. The described cohort consists of patients prescribed linagliptin, with no separate control group specified. In general, changes over time in such a cohort can reflect influences beyond the treatment exposure. The record does not describe methods for adjusting for other factors, so it does not support a causal interpretation of any change.

Enrollment of 11001 describes the study’s registered enrollment, not the number necessarily contributing to each outcome. The number with the required HbA1c measurements, the subgroup denominators, and the handling of missing data are not reported in the registry record. Those details would be necessary to understand how broadly an observed estimate applied to the enrolled cohort.

The findings that can be stated are therefore limited to the study design and outcome definitions: an existing-data cohort study of adults with type 2 diabetes prescribed linagliptin, with mean HbA1c change specified across age categories. The posted record does not establish the size, precision, or causal meaning of an outcome.

12. Why This Trial Matters Statistically

This study raises an important statistical distinction between a well-defined research question and an interpretable result. The registry specifies the HbA1c-change endpoint, the baseline and follow-up periods, and the intended age-based framing. Yet it lists 0 analyses and supplies no endpoint estimates. A registered outcome definition alone cannot tell readers what the data showed or how uncertain an estimate might be.

The record also illustrates why cohort enrollment and analytic sample size are not interchangeable. Enrollment is 11001, but the primary endpoint requires paired HbA1c measurements, and no endpoint-specific count is reported. In general, the number contributing to an analysis determines how much direct information is available for that estimate; subgroup counts and measurement completeness also matter. Without those counts, the enrollment figure cannot establish precision.

The observational design is another central statistical feature. Existing clinical records can describe outcomes in routine-care cohorts, but comparisons over time do not automatically isolate the contribution of a treatment. A comparator and clear analytic methods would be important for estimating between-group differences, while the registry description specifies neither a control cohort nor adjustment methods. This does not make the descriptive question meaningless; it defines what kind of interpretation the design can support.

Finally, the repeated age-category entry and unspecified renal-function categories show how registry wording can affect interpretability. Clear category definitions, analysis populations, missing-data methods, and results would be needed to assess subgroup estimates. The record’s completion status does not fill those gaps. Statistically, the most defensible conclusion is that the intended outcome is identifiable, but its empirical result and uncertainty are not available in the registry record.

13. 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 main limitation of the available record is the absence of posted outcome results. The primary mean change and standard deviation are not reported, and the record lists 0 analyses. This prevents assessment of the size, direction, variability, or precision of the primary outcome. The secondary outcome percentages and their denominators are also not reported.

The study description does not specify a comparator group, although its broad objective concerns comparable effectiveness. The identified patients are those with a written prescription for linagliptin. The record does not explain how other treatments or untreated patients were handled, and it does not report an adjusted comparison. The results, even if available, would need to be interpreted within the observational cohort design rather than as a randomized treatment contrast.

The endpoint requires measurements before and after initiation, but the record does not provide the number of patients with both measurements, the amount of missing outcome data, or its handling. Enrollment of 11001 does not reveal the endpoint-specific sample size. No analysis population definition or patient-flow information is reported in the registry record.

There are also details that limit clarity about subgroup interpretation. The primary description repeats the 40 to 54 years category and does not explain the intended age grouping. Renal-function categories are named for a secondary outcome but not defined. The registry does not report methods for multiple outcomes or subgroup comparisons, or the statistical model and covariates. These are limitations of the information recorded; they do not establish that the full study lacked such methods, only that they are not reported in the registry record.

The adverse-event group lists zero for the reported serious and other event counts and deaths. Without additional definitions, denominators, or context in the registry record, those entries should not be interpreted as establishing that no adverse events occurred among all patients in the study.

14. Related Tutorials

15. Related Statistical Calculators

16. Sources

• ClinicalTrials.gov: Real World Glycemic Effectiveness of Linagliptin — NCT03338803.
• Registry record last updated 2019-05-06; status last verified 2019-02.
• Results first posted 2019-05-06.
• 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.

17. Record Summary

NCT03338803, titled “Real World Glycemic Effectiveness of Linagliptin,” is a completed, non-interventional cohort study sponsored by Boehringer Ingelheim. The registry reports enrollment of 11001 and use of existing data from the Optum Clinical Database. The population is described as adults with type 2 diabetes who had a written prescription for linagliptin.

The primary outcome is mean HbA1c change across age categories, from baseline to 60 to 180 days, with standard deviation specified as the dispersion measure. Secondary outcomes concern achieving HbA1c < 7.0% across age and renal-function categories. The record lists 4 outcome measures and 0 analyses. Numerical outcome results, analysis populations, missing-data methods, and the statistical analysis plan are not reported in the registry record. The reported enrollment and completed status do not supply those missing endpoint results.