Effect of Duodenal Mucosal Resurfacing (DMR) Using the Revita System in the Treatment of Type 2 Diabetes (T2D): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effect of Duodenal Mucosal Resurfacing (DMR) Using the Revita System in the Treatment of Type 2 Diabetes (T2D) (NCT02879383), a randomized na study. The primary outcome is Change From Baseline at 24 Weeks in Hemoglobin A1c (HbA1c), DMR vs Sham..

2. Clinical Question

3. Trial Design

This was a randomized, double-masked, crossover trial of duodenal mucosal resurfacing using the Revita System versus a sham procedure. The registry lists enrollment of 109 and a phase of NA. The trial status is completed. Its recorded start date is 2017-03-01, and primary completion is recorded as 2019-12-20.

The design included DMR and sham groups, with results reported separately for Europe and Brazil. For the primary endpoint, the registry gives a DMR value and a sham value for each region. The stated primary endpoint timeframe is baseline and 24 Weeks post-procedure. That timing matters because the sham group crossed over to DMR at 24 weeks; the primary comparison therefore precedes the crossover. After the crossover, the former sham participants were followed for an additional 24 weeks after receiving DMR. The original DMR group was also followed for 48 weeks and was unblinded at 24 weeks.

The registry describes background medications as held constant from 24-48 weeks for the sham treatment arm. It does not report further detail on medication management during the primary comparison period. The double-masked designation applies to the trial as listed, while the arm descriptions state that participants were unblinded at 24 weeks.

The registry reports the primary endpoint for the mITT population, but does not define which participants were included in that population. It does not report a formal statistical analysis of the regional DMR-versus-sham contrasts. Safety results include the initial regional groups and separate sham crossover groups; those crossover results describe a later exposure period and should not be treated as additional randomized sham-versus-DMR comparisons.

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: All subjects participated in a 4-week oral antidiabetic drug (OAD) run-in period before the index procedure to confirm lack of blood glucose control in conjunction with medication compliance and nutritional counseling. All subjects were managed according to current diabetes standard of care.

First 24 Weeks

Reasons for not completing this period:

Second 24 Weeks (Crossover for Sham)

Analysis population for the primary outcome, as defined in the registry: mITT population

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 at 24 Weeks in Hemoglobin A1c (HbA1c), DMR vs Sham., measured over Baseline and 24 Weeks post-procedure.

The primary endpoint is change from baseline at 24 weeks in hemoglobin A1c (HbA1c), comparing DMR with sham. HbA1c is a measure used to assess average blood glucose over time, so a change from baseline describes how the measure differed from its starting value during follow-up. The registry defines the reported unit as percentage change from baseline and says the summary parameter is the median.

A median describes the middle of the observed distribution, rather than its arithmetic average. It can be useful when a distribution is skewed or contains values that would pull the mean, but the registry record does not give the underlying distribution. Although the endpoint description specifies inter-quartile range as the dispersion measure, no inter-quartile range values are reported for these groups.

The endpoint is reported separately for the Europe and Brazil groups. That presentation gives regional summaries, but it does not by itself establish the size or precision of a single overall treatment contrast. The registry record does not report a formal between-group estimate, its standard error, or a hypothesis test for this endpoint.

HbA1c is a biomarker outcome, not a direct measure of how participants felt or functioned. The endpoint therefore answers a specific laboratory-measure question at a specified follow-up time. It does not, on its own, describe every aspect of diabetes control, longer-term outcomes, or safety.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry labels the study randomized, double-masked, and crossover. These terms describe aspects of how the trial was organized, not the statistical analysis used for the primary endpoint. Random allocation is intended to make comparison groups more alike, on average, with respect to factors that could affect outcomes. Masking is intended to limit the influence of knowledge about assignment on participants or study conduct. The registry does not report who was masked, how masking was maintained, or the details of randomization. It states that participants were unblinded at 24 weeks.

What does the median describe?

The primary outcome is summarized by a median in each regional group. In general, the median is the central value when observations are ordered. It is less affected than the mean by unusually high or low observations, which can make it a useful summary when data are asymmetric. A median alone does not describe the full spread of participant values. The endpoint description names the inter-quartile range as the dispersion measure, but the registry does not report those values. The sample sizes listed with the primary endpoint provide context for the summaries but do not replace information about variability.

What are the posted bounds?

The record gives a lower and upper bound for each regional median, but does not label the bounds or describe how they were calculated. A range of bounds could represent different kinds of uncertainty or summary, depending on the method. Without a label and method, it is not appropriate to call them confidence intervals, assume a particular confidence level, or use them as a formal test. General statistical interpretation: an interval estimate communicates uncertainty only when its construction and intended coverage are known. Those details are not reported in the registry record.

Where is the treatment contrast?

A comparative analysis usually estimates a contrast between groups and describes its uncertainty. That allows readers to distinguish the estimated size of a difference from how precisely it is known. Here the registry posts separate regional medians and bounds but does not report the DMR-versus-sham contrast, a standard error, a p-value, or a formal test. The difference between two posted medians is not itself a substitute for the trial's statistical analysis. No new contrast should be inferred from the displayed arm summaries.

The record identifies the primary population as mITT, but does not provide its definition or state how participants with missing endpoint data were handled. General statistical interpretation: an intention-to-treat approach generally aims to retain participants according to randomized assignment, while modified versions can apply additional inclusion rules. The label alone does not reveal those rules. The registry also does not describe the analysis model, covariate adjustment, handling of the regional groups, or the method used to calculate the posted bounds.

How should the crossover be considered?

In a crossover design, participants can receive different treatments in sequence. Such designs can use within-participant comparisons, but interpretation may depend on the timing of treatment, persistence of effects, and possible sequence or period effects. General statistical interpretation: if treatment effects carry into a later period, outcomes after switching may not represent a clean comparison with ongoing control. The registry's primary endpoint is at 24 weeks, when sham participants were scheduled to cross over. The primary regional summaries are therefore distinct from the later crossover follow-up. The registry does not report a crossover-period efficacy analysis or methods addressing sequence, period, or carryover effects.

The sponsor's sample-size illustration is described as planning based on the observed effect, with a two-sided significance level 0.05, equal allocation, and a normal approximation; its target effect is a stated fraction of the observed effect. These are planning assumptions, not the trial's reported design justification. The record does not report the trial's own sample-size justification or numerical planning estimates in the registry record for this analysis. General statistical interpretation: a planning calculation estimates a sample size for a future study under stated assumptions; it is not observed power and does not establish the precision of the current trial's estimates.

The registry lists 0 analyses. It does not provide a detailed statistical analysis plan or explain whether analyses were conducted but not entered in the registry. General statistical interpretation: a prespecified analysis plan can clarify outcome definitions, model choices, missing-data handling, and multiplicity procedures before results are examined. Those procedures are not reported in the registry record, so the posted summaries cannot be used to reconstruct them.

10. Multiplicity, Missing Data and Analysis Populations

The primary endpoint is identified as the change from baseline in HbA1c at 24 weeks. The registry lists 2 outcome measures, but it does not provide a detailed account of the other measure or any posted secondary outcomes; the record lists 0 secondary outcomes posted. It also lists 0 analyses. The registry does not report whether the primary endpoint was part of a broader family of planned statistical tests or how any multiplicity was handled.

Multiplicity is a general statistical concern when multiple outcomes, time points, groups, or analyses are tested. If many tests are considered, the chance of at least one apparently notable result can increase. Adjustments or a prespecified hierarchy can address that concern, but no approach is reported in the registry record. The regional presentation gives separate summaries for Europe and Brazil, but the record does not state whether these were planned as independent confirmatory comparisons or how regional results were intended to be interpreted.

The primary endpoint is reported for the mITT population. The registry does not define this population, explain exclusions, or state how missing baseline or follow-up HbA1c values were handled. The number of participants contributing to each endpoint summary is listed by arm, but reasons for any differences between those counts and other trial counts are not reported in the registry record. No imputation method, complete-case rule, sensitivity analysis, or missing-data assumption is described.

General statistical interpretation: missing-data methods matter because participants without measured outcomes may differ from participants with observed outcomes. The consequences depend on why data are missing and on the assumptions used to analyze them. Without information about missingness and the analysis method, readers cannot assess how sensitive the endpoint summaries might be to unobserved outcomes.

11. Safety Results

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

Deaths

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: jejunal perforation (5.9%), duodenal ulcer (5.9%), intervertebral disc protrusion (5.9%), ear disorder (2.7%), acute myocardial infarction (2.6%).

The registry reports serious adverse events and other adverse events separately for the regional treatment groups and for sham crossover groups. For serious events, it lists 5 of 39 in DMR Procedure (Europe), 1 of 37 in Sham Procedure (Europe), 3 of 17 in DMR Procedure (Brazil), and 0 of 16 in Sham Procedure (Brazil). The crossover groups are listed separately: 0 of 33 in Sham Crossover (Europe) and 0 of 4 in Sham Crossover (Brazil). Deaths are listed as 0 for each group.

These are counts of participants affected as recorded in the registry, with the corresponding at-risk group counts. Serious adverse events are not the same as all adverse events. The registry also reports other adverse events for each group, but those counts should be read as a separate category and not added to serious-event counts. Because participants assigned to sham later crossed over, crossover safety counts describe a later treatment period and should not be interpreted as a concurrent sham control.

The registry lists the top serious-event terms with percentages: Jejunal perforation, 5.9%; Duodenal ulcer, 5.9%; Intervertebral disc protrusion, 5.9%; Ear Disorder, 2.7%; Acute myocardial infarction, 2.6%; and back pain, 2.6%. The record does not link these terms to specific treatment groups in the information summarized, nor does it state that the listed events were caused by treatment. Safety counts alone do not establish comparative risk.

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

The posted results describe median percentage changes in HbA1c from baseline at 24 weeks for DMR and sham, reported separately for Europe and Brazil. They also describe serious and other adverse-event counts, including separate later crossover groups. This gives readers a view of the observed summaries and the trial's follow-up structure.

The results do not establish a formal treatment effect. Although randomization creates a basis for comparison, the registry record does not report a direct DMR-versus-sham estimate, a test statistic, a p-value, or a clearly labeled interval estimate. Separate arm medians do not answer precisely how large the between-group difference is or how uncertain that difference is. In particular, the posted bounds are not identified as confidence intervals, so they cannot support a conventional confidence-interval interpretation.

The values are regional summaries, not a pooled estimate. They do not show whether the regional results differ beyond chance, and they do not explain any difference between regions. The record does not report a model combining regions or assessing whether treatment effects vary by region. Readers should not treat the Brazil or Europe result as necessarily representative of all participants in the trial.

The endpoint concerns HbA1c, a laboratory measure. It does not by itself capture symptoms, daily functioning, complications, or other outcomes. The safety counts provide descriptive event information, but do not establish causation or a comparative safety profile. The top serious-event terms are listed with percentages, but the record does not connect each term to a particular group or provide the context needed to interpret those percentages as group-specific risks.

Finally, the crossover changes what can be learned after 24 weeks. Once sham participants received DMR, their later follow-up was no longer a continuing sham comparison. The registry does not report a crossover-period efficacy analysis. The posted results therefore support a limited descriptive reading of the primary 24-week regional summaries and the recorded safety counts, rather than a broader conclusion about long-term comparative outcomes.

13. Why This Trial Matters Statistically

This trial is statistically notable because its central question is comparative, while the registry reports the primary outcome as separate regional summaries. The distinction matters: an arm-specific median describes the observed outcome distribution in one group, whereas a treatment contrast directly estimates how outcomes differ between randomized groups. Without a reported contrast and its uncertainty, readers cannot assess the precision of the comparative effect from the registry record alone.

The regional breakdown also highlights the importance of analysis choices. Europe and Brazil have separate DMR and sham summaries, and the registry does not report a combined estimate or a formal assessment of regional variation. Such a presentation can be informative, but it requires careful interpretation: variation in observed values is not, by itself, evidence that the treatment effect differs by region. The analysis methods and intended role of these regional estimates are not reported.

The crossover makes timing central to interpretation. The primary endpoint occurs at 24 weeks, just as the sham group crossed over to DMR. This preserves the initial DMR-versus-sham comparison at the primary assessment, while later follow-up answers a different question because the groups no longer have the same treatment assignments. Crossover safety data should therefore be distinguished from the randomized comparison period.

The record also illustrates why endpoint definitions and analysis details matter. It specifies a median and inter-quartile range, yet does not report the inter-quartile range values or label the posted bounds. It identifies an mITT population but does not define it or explain missing-data handling. These omissions limit assessment of both variability and analytic robustness, even when the arm summaries themselves are available.

For clinical researchers, the key statistical lesson is to separate descriptive results from inferential conclusions. The posted values are useful observations, but a complete interpretation of a randomized trial also depends on the prespecified estimand, analysis method, uncertainty estimates, missing-data approach, and treatment of multiple comparisons. Those details are not reported in the registry record.

14. 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 leaves several trial-specific interpretive details unreported. It does not define the mITT population, describe eligibility criteria in the information summarized, or explain the analysis population for each regional result. It names the inter-quartile range as the dispersion measure but provides no values for it. The lower and upper bounds accompanying the medians are not labeled, and their calculation method is not reported.

No formal between-group analysis is posted for the primary endpoint. There is no reported treatment contrast, test, p-value, standard error, or model description. The regional summaries are not accompanied by a reported pooled analysis or a test of regional differences. The record does not say whether regional analyses were prespecified, or what role they were intended to play in the study's conclusions.

The registry also does not report a statistical analysis plan, missing-data procedures, sensitivity analyses, or multiplicity handling. Although the primary endpoint and population are named, the rules for endpoint inclusion and analysis are not available in the registry record. This limits the ability to evaluate potential bias from missing measurements or analytic choices.

The crossover adds a further limitation for interpreting later follow-up. The sham group received DMR after 24 weeks, so later observations cannot be read as a sustained randomized comparison of DMR versus sham. The registry does not report a crossover-period efficacy analysis or statistical methods addressing sequence, period, or carryover effects. Safety data for the crossover groups are presented separately, but the record does not attribute listed serious-event terms to specific groups or establish treatment causation.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Effect of Duodenal Mucosal Resurfacing (DMR) Using the Revita System in the Treatment of Type 2 Diabetes (T2D) — NCT02879383.
• Registry record last updated 2024-02-14; status last verified 2024-02.
• Results first posted 2021-05-25.
• 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.

18. Record Summary

The registry identifies NCT02879383 as a completed randomized, double-masked crossover trial of DMR using the Revita System versus a sham procedure for type 2 diabetes. Enrollment is listed as 109, and the primary endpoint is change from baseline in HbA1c at 24 weeks in the mITT population. Sham participants crossed over to DMR at 24 weeks.

Primary endpoint results are reported as regional median percentage changes from baseline for Europe and Brazil. The registry does not report a formal between-group analysis, a labeled interval method, or a pooled treatment estimate. It lists serious and other adverse-event counts for initial regional groups and sham crossover groups, as well as deaths recorded as 0 in each group.

The record does not report the definition of the mITT population, missing-data methods, multiplicity procedures, or a detailed statistical analysis plan. It lists 0 analyses and 0 secondary outcomes posted.