A Study of Omarigliptin (MK-3102) in Participants With Type 2 Diabetes Mellitus With Chronic Kidney Disease or Kidney Failure on Dialysis (MK-3102-019): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Omarigliptin (MK-3102) in Participants With Type 2 Diabetes Mellitus With Chronic Kidney Disease or Kidney Failure on Dialysis (MK-3102-019) (NCT01698775), a randomized phase 3 study. The primary outcome is Change From Baseline in Glycosylated Hemoglobin (A1C) at Week 24.

2. Clinical Question

3. Trial Design

This was a Phase III, multicenter, randomized, double-blind study with a parallel-group model and triple masking recorded in the registry. It evaluated omarigliptin versus placebo in people with type 2 diabetes and moderate or severe chronic kidney disease or kidney failure on dialysis who had inadequate glycemic control. The registry records enrollment of 213 and a completed study. The recorded start date is 2012-10-02, and the primary completion date is 2016-01-19.

The primary endpoint was assessed at Week 24, following a Phase A treatment period of 24 weeks. The arm descriptions identify omarigliptin and matching placebo during this period. They also refer to Phase B and later treatment, but the primary result is specifically the Week 24 Phase A comparison. The registry record does not provide enough detail to resolve all of the later-phase arm-description wording, and those later descriptions should not be used to reinterpret the primary contrast.

Random assignment is intended to make treatment groups comparable, on average, with respect to measured and unmeasured characteristics. Masking is intended to limit the influence of treatment knowledge on trial conduct and outcome assessment. The registry records the design features but does not identify which people or roles were masked. It also does not report a detailed account of the randomization procedure or allocation concealment.

The posted primary analysis includes 106 participants in each treatment arm. The analysis population is defined in the endpoint description, rather than simply as everyone enrolled. Therefore, the enrollment figure and the analyzed arm counts refer to different registry summaries and should not be treated as interchangeable.

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.

Phase A

Reasons for not completing this period:

Phase B

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full analysis set (FAS) population included all randomized participants who received at least 1 dose of study medication and had a baseline measurement or a post-randomization measurement in Phase A for the analysis endpoint subsequent to at least 1 dose of study medication.

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 Change From Baseline in Glycosylated Hemoglobin (A1C) at Week 24, measured over Baseline and Week 24.

The primary endpoint was change from baseline in glycosylated hemoglobin, or A1C, at Week 24, measured in percent. A1C is a continuous measure used to represent glycemic status over time. A change from baseline expresses how the measure shifted during the study period, while comparing changes between randomized groups helps assess the difference associated with the assigned study treatments under the analysis model.

The registry reports least-squares mean changes with 95% confidence intervals. Least-squares means are model-based estimates, rather than simple arithmetic averages of the observed values. The model description specifies constrained longitudinal data analysis, with the same baseline mean imposed across treatment groups. It includes treatment, renal insufficiency stratum, baseline insulin-treatment stratum, time, and specified interactions. This approach uses repeated measurements and the randomized comparison while accounting for the listed factors.

A1C was selected as the primary measure because the study’s stated hypothesis concerned whether omarigliptin produced greater A1C reduction than placebo after 24 weeks. The registry does not report a threshold for what size of A1C change would be clinically important. Statistical evidence about a difference and judgments about its clinical relevance are therefore separate questions.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes the primary endpoint as a continuous outcome and reports least-squares mean change from baseline. A continuous measure can take values across a scale rather than being limited to event categories. A mean summarizes the central tendency of those measurements, while an estimated difference between treatment groups addresses the trial’s comparison. Model-based means can account for features of the design and data that a simple average does not address.

What does the constrained longitudinal analysis do?

The endpoint description states that change in A1C was analyzed using constrained longitudinal data analysis, or cLDA. The model imposed the restriction that the baseline mean was the same across treatment groups. In a randomized study, baseline measurements are collected before treatment can cause a change; modeling them jointly with follow-up measurements can support an efficient comparison. The constraint reflects the expectation that randomized groups share a common baseline mean in the model. This is a modeling assumption, not a claim that every participant had the same baseline A1C.

The listed cLDA terms include treatment, renal insufficiency stratum, baseline treatment with insulin stratum, time, and interactions of time with treatment and with the two strata. In general, including stratification factors and interactions allows estimated changes to account for those design features and to represent how change may vary across time or strata. The registry does not provide coefficient estimates or a fuller specification that would permit independent reconstruction of the model.

Why are least-squares means reported?

Least-squares means are estimates generated from a statistical model. They represent treatment-group means after accounting for terms included in that model, rather than unadjusted averages calculated only from the displayed arm summaries. The registry reports the omarigliptin estimate as -0.77 and the placebo estimate as -0.44. It also reports a separate adjusted difference in least-squares means. The contrast is the central inferential comparison; the confidence interval and p-value quantify uncertainty and evidence under the stated analysis.

How should the confidence interval and p-value be read?

A confidence interval gives a range of effect values compatible with the data and model under the stated procedure. It conveys precision as well as direction: a narrower interval indicates greater precision than a wider interval, all else equal. It does not show the spread of individual responses, and it does not give the probability that the true effect lies within this particular interval. The reported interval for the between-group difference is -0.63 to -0.02. The registry’s p-value is 0.035. A p-value is calculated under a null hypothesis and model assumptions; it is not the probability that the null hypothesis is true and does not measure clinical importance.

Why might another calculation differ?

A simple unadjusted calculation from posted summaries may not match a model-adjusted comparison. The groups’ least-squares means are rounded, and the reported analysis uses an ANCOVA comparison of least-squares means. In general, adjusted analyses can use participant-level observations and covariates that are not recoverable from rounded group estimates. The registry does not report enough model detail or participant-level data to recreate the sponsor’s analysis independently. No separate recalculated contrast is reported in the registry record.

What role does over-dispersion play?

Over-dispersion concerns outcomes whose observed variability exceeds what a model’s assumed distribution allows, often relevant for counts or event rates. The primary endpoint is continuous A1C, and the registry describes a longitudinal model rather than a count model. The record does not report an over-dispersion assessment for this endpoint. More generally, checking whether a model’s assumptions fit the data matters because standard errors and confidence intervals rely on those assumptions.

The registry states that the analysis population included randomized participants who received at least 1 dose and had a baseline measurement or a qualifying post-randomization measurement for the endpoint. This is a defined analysis set, not necessarily every randomized participant. The record does not give participant-level missingness patterns, the missing-data assumptions used in the model, or sensitivity analyses addressing alternative assumptions. Consequently, readers can identify the stated endpoint method and analysis population, but cannot fully assess robustness to missing data or independently reproduce the inferential analysis.

10. Analyses Reported by the Sponsor

The registry’s inferential result is the reported ANCOVA difference in least-squares means: -0.33, with a 95% confidence interval from -0.63 to -0.02 and p = 0.035. The arm-level least-squares means are descriptive components of the same general comparison, but they are not a substitute for the adjusted contrast. The primary endpoint description also specifies cLDA, including a shared baseline-mean constraint and listed stratification terms.

The registry record does not report a separate independently recalculated contrast. A recalculation from posted arm summaries, if made, would be an unadjusted reconstruction and could differ from the sponsor’s model-based analysis because the summaries are rounded and do not contain all participant-level information or model details. No numeric recalculated result can be compared with the sponsor-reported result from the registry record. The reported sponsor analysis should therefore be read as the available inferential analysis, while recognizing that the registry does not include enough detail to reproduce it independently.

The sponsor’s result is a statistical estimate for the stated Week 24 comparison. It does not resolve clinical importance, multiplicity across outcomes, or sensitivity to missing-data assumptions, which are not reported in the registry record.

11. Multiplicity, Missing Data and Analysis Populations

The primary endpoint is change from baseline in A1C at Week 24. The registry also lists secondary outcomes, including fasting plasma glucose at Week 24, A1C and fasting plasma glucose at Week 54, and estimated glomerular filtration rate at Week 24 and Week 54. It records 10 outcome measures and 7 analyses. These entries indicate that the trial reported more than the primary comparison, but the registry record does not describe how the statistical testing was organized across outcomes or analyses.

Multiplicity arises when several hypotheses are tested. Without an adjustment or a prespecified testing hierarchy, interpreting each p-value as if it were the only test can overstate the strength of evidence across the collection of results. The registry record does not report a multiplicity adjustment, a gatekeeping strategy, or which secondary outcomes were confirmatory versus exploratory. The primary p-value should therefore be identified as the result for the stated primary comparison; the record does not establish how it fits into a broader family of tests.

The registry defines the full analysis set as randomized participants who received at least 1 dose of study medication and had a baseline measurement or a post-randomization measurement in Phase A for the endpoint after at least 1 dose. This definition clarifies who could contribute to the primary analysis. It is not identical to a statement that every randomized participant was included regardless of treatment received or endpoint availability.

The registry record does not report the number or pattern of missing primary endpoint measurements, the method used to handle missing values, or sensitivity analyses under alternative missing-data assumptions. It also does not report interim analyses, stopping rules, or the detailed statistical analysis plan. These omissions limit assessment of whether missingness or repeated examination of the data could have affected the reported inference. They do not, on their own, show that the analysis was biased; they mark information that cannot be evaluated from the registry entry.

12. 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: chronic kidney disease (2.8%), peritonitis (1.9%), pneumonia (1.9%), hyperkalaemia (1.9%), benign prostatic hyperplasia (1.9%).

The registry posts serious adverse event counts for the Phase A groups: 10 of 106 participants in the omarigliptin group and 13 of 106 in the placebo group. It also posts other adverse event counts of 24 of 106 and 25 of 106, respectively. These are separate categories in the record; they should not be combined or treated as interchangeable. The listed serious-event terms include chronic kidney disease at 2.8%, and peritonitis, pneumonia, hyperkalaemia, benign prostatic hyperplasia, and hypoglycaemia at 1.9% each.

The adverse-event record also includes later-phase group labels and counts. Those groups reflect the registry’s Phase A and Phase B descriptions and are not the same presentation as the primary Week 24 efficacy comparison. The death counts are not reported in the registry record. The entries do not provide enough information to infer causation or to compare risks while accounting for exposure time, event timing, or differences in follow-up. They are descriptive safety counts, not proof that an event was caused by or prevented by a study treatment.

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

The primary result is evidence of a difference in Week 24 A1C change between the randomized omarigliptin and placebo groups under the registry’s reported analysis. The estimated between-group difference is -0.33, and the 95% confidence interval ranges from -0.63 to -0.02. The direction is consistent with a greater reduction in the omarigliptin group, given the reported contrast orientation. The p-value is 0.035. These statistics describe the estimated group contrast and its uncertainty; they do not describe every participant’s response.

The result does not establish how important the estimated difference is in clinical terms. The registry record does not provide a threshold for a clinically meaningful A1C difference or enough contextual information to judge that question. A statistically detectable contrast and a clinically important contrast are related but distinct concepts.

The result also does not answer every question about the study. It is for the primary A1C outcome at Week 24, not for the listed Week 54 outcomes, kidney-function outcomes, or all safety outcomes. It does not establish the effect for people outside the recorded study population. Nor does the registry entry provide sufficient information to assess the impact of missing measurements, multiplicity, or all model assumptions.

Confidence intervals express uncertainty in the estimated average contrast under the analysis. They do not predict the range of individual changes, guarantee that a future study will produce the same estimate, or show the probability that a particular treatment effect is true. The posted analysis should be read as one randomized-trial comparison for the defined population and time point, with the limitations in reporting kept in view.

14. Why This Trial Matters Statistically

This trial raises a statistical question in a population defined by type 2 diabetes and moderate or severe chronic kidney disease or kidney failure on dialysis. Its primary outcome is a continuous measure, and its analysis uses model-based estimates rather than relying only on unadjusted group averages. The registry’s description of cLDA makes visible how baseline measurements, time, treatment, and stratification factors enter the analysis.

The comparison also illustrates why the arm summaries and the inferential contrast need to be read together but not confused. Least-squares means describe model-based group changes; the ANCOVA contrast is the reported direct comparison, accompanied by a confidence interval and p-value. The contrast indicates direction and precision, while the absence of a clinical-importance threshold means statistical evidence alone cannot settle practical importance.

There are multiple posted outcomes and analyses, including measures at different time points. That makes clear reporting of endpoint hierarchy and multiplicity procedures important for interpreting the collection of results. The registry record does not describe those procedures. It also does not document missing-data patterns or sensitivity analyses, so readers cannot fully assess the robustness of the primary estimate to incomplete endpoint data.

The trial is therefore useful statistically as an example of interpreting a randomized, masked, parallel-group comparison with an adjusted continuous endpoint in a specific clinical population. Its value for inference depends not only on the reported point estimate, but also on the confidence interval, analysis population, model description, and the limits of registry reporting.

15. Limitations

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

Several trial-specific details needed for a fuller assessment are not reported in the registry record. The record does not provide the sample-size justification, the planned statistical analysis document, or a detailed account of randomization and allocation concealment. The enrollment figure is 213, while the primary arm summaries each include 106 participants; the record does not explain the relationship between enrollment and the endpoint analysis counts in further detail.

The primary endpoint description identifies cLDA and its principal terms, while the reported comparison is described as ANCOVA. The registry does not provide sufficient model specifications, covariance details, or participant-level data to reproduce the estimate and standard error. It also does not report diagnostic checks or sensitivity analyses that would help evaluate the model assumptions.

Information about missing primary measurements, reasons for missingness, and the method for handling missing data is not reported in the registry record. Nor is there a reported multiplicity strategy for the several outcome measures and analyses. Interim analyses and stopping rules are not reported in the registry record. These gaps limit the ability to judge how robust the nominal p-value is to missing data, model choices, or testing across outcomes.

The adverse-event entries are counts and selected event terms. The registry does not provide the detail needed to assess timing, exposure-adjusted event rates, attribution, or the completeness of follow-up by treatment group. The arm descriptions for Phase B are also incomplete in wording and should not be used to infer details beyond what is stated. These limitations qualify interpretation of the posted results but do not alter the reported Week 24 estimates.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of Omarigliptin (MK-3102) in Participants With Type 2 Diabetes Mellitus With Chronic Kidney Disease or Kidney Failure on Dialysis (MK-3102-019) — NCT01698775.
• Registry record last updated 2018-08-09; status last verified 2017-09.
• Results first posted 2017-02-23.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

The registry records a completed Phase III, multicenter, randomized, double-blind, parallel-group study with triple masking, comparing omarigliptin with placebo in participants with type 2 diabetes and moderate or severe chronic kidney disease or kidney failure on dialysis. The primary outcome was change from baseline in A1C at Week 24.

The sponsor-reported between-group difference in least-squares means was -0.33, with a 95% confidence interval from -0.63 to -0.02 and p = 0.035. The estimate indicates a greater A1C reduction with omarigliptin under the reported analysis. The registry does not report a separate recalculated contrast, a multiplicity procedure, or detailed missing-data sensitivity analyses. Interpretation should remain specific to the defined analysis population, primary endpoint, and Week 24 comparison.