EMI-EHP Weight Management and Type 2 Diabetes Pragmatic Trial: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for EMI-EHP Weight Management and Type 2 Diabetes Pragmatic Trial (NCT04531176), a randomized phase 4 study. The primary outcome is Mean Change in Body Weight Between Baseline and 12 Months.

2. Clinical Question

The clinical question is therefore about the observed average percentage change in body weight under the three approaches, rather than an outcome such as diabetes control or a direct measure of long-term health. The registry describes a pragmatic trial in an employer setting, but details about eligibility criteria, recruitment, adherence, and how the interventions were delivered are not reported in the registry record.

3. Trial Design

NCT04531176 is recorded as a Phase 4, randomized, parallel-group trial with no masking. The planned duration was 24 months, and the registry lists the study as terminated. The recorded start date is 2020-09-01 and the primary completion date is 2022-08-15. Actual enrollment was 74, while the study description planned for 300 employees or spouses covered by the employer health plan.

The three approaches were an obesity-centric program with anti-obesity medication, the same program without anti-obesity medication, and usual care. The program was described as Cleveland Clinic’s Integrated Medical Weight Management Program. The medication-supported arm was to use one of four listed medications, selected after discussion with the study doctor. The usual-care arm was described as traditional care for type 2 diabetes, hypertension, and hypercholesterolemia. The registry does not report how consistently participants received the assigned approach or whether treatment changes occurred during follow-up.

The primary outcome was mean percentage change in body weight between baseline and 12 months. Arm-level results list 23 participants for the medication-supported approach, 23 for the program without medication, and 24 for usual care. Those figures describe the posted outcome results; the registry does not explain whether they reflect all randomized participants, participants with available measurements, or a specified analysis population.

The trial’s open-label design means that masking was not used, according to the registry. The record does not describe safeguards against expectations affecting behavior, care, measurement, or outcome assessment. It also does not report why the trial was terminated, whether termination was planned, or whether the study reached its intended recruitment target. These points matter when judging how fully the planned comparison was carried out, but they do not establish a reason for the recorded status.

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:

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: Individual participants who were randomized to each arm of the study

6. Primary Endpoint

The registry lists the primary outcome as Mean Change in Body Weight Between Baseline and 12 Months, measured over 12 Months.

The primary endpoint measures the mean change in body weight from baseline to 12 months as a percentage of body weight. A negative value represents a decrease in weight over that interval, consistent with the endpoint description’s reference to percentage weight loss. The endpoint focuses on the average change in each trial group rather than on how many individuals crossed a particular threshold.

Body weight is a direct measure of the outcome targeted by the weight-management approaches, and expressing change as a percentage relates the change to a participant’s baseline body weight. This can make average changes more interpretable across participants whose initial weights differ. That is a general rationale for percentage change; the registry does not report a specific justification for choosing this endpoint or the percentage scale.

The primary time point is 12 months, although the study description planned a 24-month program. The endpoint description refers to relevant time points between baseline and 12 months, but the registry does not clarify how measurements at multiple visits were used to derive the reported mean. It also does not report the measurement protocol, whether assessors were masked, or how missing weight measurements were handled.

The posted results give an arm mean and a 90% confidence interval for each group. A confidence interval expresses uncertainty around an estimate under the statistical method used; it is not a range containing every participant’s change. The registry does not provide a separate measure of individual-level variability alongside these means and intervals.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry presents three arm means for a continuous outcome: percentage change in body weight. A mean is the arithmetic average of the measured values included in an analysis. It summarizes the group’s central result, but it does not show how individual participant results were distributed. Two groups can have similar means while differing in variability or in the pattern of individual outcomes.

Each mean is accompanied by a 90% confidence interval. In general, a confidence interval is a range generated by a statistical procedure to show uncertainty around an estimate. Its interpretation depends on the assumptions and method used. It is not a guarantee that the true value lies in the particular interval, and it does not describe the range of individual participant outcomes. The registry gives the intervals but does not name the method used to calculate them or report a separate standard deviation.

The reported test is a one-sided t-test and is labeled as a non-inferiority analysis. In general, a t-test assesses a mean-based comparison while accounting for uncertainty in the estimate. A one-sided test evaluates evidence in a specified direction, unlike a two-sided test, which considers departures in either direction. To interpret a one-sided result, the direction and hypothesis need to be clear. The registry does not report the direction of the test, the null and alternative hypotheses, the non-inferiority margin, or which specific group comparison was tested.

What does non-inferiority mean?

In general, a non-inferiority design asks whether an intervention is not worse than a comparator by more than a prespecified margin. That margin defines the largest difference treated as acceptable for the question being studied. A p-value alone does not provide the margin or identify the comparison. Because the registry does not state the margin or comparison-specific estimates, the non-inferiority label is not enough to reconstruct the intended claim. It is also not appropriate to infer a margin from the arm means or their confidence intervals.

The registry lists all three groups in the reported test. A conventional t-test is usually associated with a comparison of two means, but the record does not explain how this one-sided test was applied across the three listed groups. It could reflect an analysis or contrast not fully described in the posted record; the details are not reported in the registry record. The recorded p = 0.004 should therefore be read as the value attached to the registry’s listed analysis, not as a fully interpretable test of a named pairwise contrast.

Why can over-dispersion matter?

In general, over-dispersion means that observed variability is larger than a statistical model expects. If a model assumes less variability than the data contain, uncertainty can be understated and tests can appear more decisive than warranted. For continuous body-weight change, variability is commonly expressed through a dispersion measure such as a standard deviation. The registry does not report a separate dispersion statistic for these arm means, and it does not describe a model that would permit assessment of over-dispersion. The confidence intervals offer some information about uncertainty in the means, but cannot replace a complete description of the analysis method.

Random allocation is intended to make groups comparable on average at the start of a study, reducing systematic differences in assignment. The registry describes randomization and a planned 1:1:1 allocation. It does not describe the random sequence method, concealment of allocation, stratification, or any adjustments used in the outcome analysis. The observed outcome group sizes are 23, 23, and 24; the record does not explain whether all randomized participants contributed to the primary analysis.

The trial was open-label. In general, lack of masking can affect behavior, expectations, care decisions, or outcome assessment, depending on the outcome and study conduct. Body weight is a measurable outcome, but the record does not state who measured it, how it was measured, or whether outcome assessors were masked. These are uncertainties about reporting, not evidence that any particular bias occurred.

The posted confidence intervals are arm-specific. For a comparative question, the relevant estimate is usually a between-group contrast, accompanied by its uncertainty. Comparing separate arm intervals informally does not give the same result as estimating the contrast directly, because the uncertainty of a contrast depends on the uncertainty and relationship between the group estimates. No comparison-specific estimates or confidence intervals are reported in the registry record.

The sample-size figures associated with planning illustrations are not observed power and do not document the original design rationale. In general, a planning calculation depends on choices such as the target effect, significance level, allocation, and statistical assumptions. The registry does not report the trial’s own sample-size justification. Any planning illustration based on an observed effect is separate from the evidence generated by the trial and should not be interpreted as showing the probability that the trial detected an effect.

10. Analyses Reported by the Sponsor

The registry reports a one-sided t-test labeled as non-inferiority, with p = 0.004. It does not report a comparison-specific effect estimate, confidence interval, non-inferiority margin, or the hypotheses for that test. The p-value therefore cannot be connected from the record to a clearly named contrast or used to reconstruct the sponsor’s intended non-inferiority conclusion.

The arm-level means and 90% confidence intervals describe changes within each group. They are not substitutes for a direct estimate of the difference between groups. No recalculated comparison results are reported in the registry record, so a numerical comparison between the registry’s test and an unadjusted recalculation cannot be made from the posted details.

The registry record does not indicate a discrepancy between its reported analysis and a recalculation. That does not resolve the missing information about the test’s contrast or non-inferiority margin. Where a sponsor’s own analysis and a simple recalculation differ, the sponsor’s own analysis takes precedence; in this record, the ingredients needed to assess such a numerical comparison are not reported.

The most defensible reading is to distinguish the reported p-value from the descriptive arm means. The means show average decreases in all groups, while the analysis description is insufficiently detailed to determine what inferential claim was tested.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome and reports a non-inferiority analysis involving all three groups. It does not state how many comparisons were planned, which contrasts were primary, or whether any adjustment for multiple comparisons was used. When several groups or outcomes are compared, the chance of at least one apparently notable result can increase; in general, multiplicity procedures are used to address that issue. The record does not report a multiplicity plan, so it is not possible to determine how the reported p-value fits into a wider set of tests.

The record lists two outcome measures but reports zero posted secondary outcomes. It does not provide secondary outcome estimates in the registry record. The primary outcome is specified as mean body-weight change at 12 months, but details about other time points and any additional analyses are not reported.

Missing-data handling is not reported in the registry record. The outcome group sizes are 23, 23, and 24, while total enrollment is 74. These figures do not establish whether every randomized participant had a 12-month measurement or whether the analysis used a particular population. The record does not describe exclusions, imputation, sensitivity analyses, or reasons measurements may have been unavailable.

The analysis population is also not reported. It is not clear whether results were based on all randomized participants, participants who followed assigned treatment, or another defined group. Those choices can affect interpretation, especially when participants discontinue or switch approaches. The record does not report interim analyses, stopping rules, or a prespecified statistical analysis plan. Their absence from the registry description does not establish whether they were or were not used; the details are not reported in the registry record.

12. 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: nausea, vomiting, pain (4.3%).

The registry’s serious adverse event counts are 0 among 24 participants in the medication-supported group, 1 among 23 participants in the program-without-medication group, and 0 among 23 participants in usual care. Deaths are reported as 0 in all three groups. These are the counts and denominators listed in the adverse-event record; they should not be treated as estimates of comparative safety without further information about event definitions, follow-up, and analysis.

Other adverse events are reported for 3 of 24 participants in the medication-supported group, 0 of 23 in the program-without-medication group, and 2 of 23 in usual care. The record lists “Nausea, Vomiting, Pain” as a serious-event term with 4.3%. It does not explain in the posted information how that term relates to the group-level serious-event counts, whether it represents one event or a category, or how events were attributed.

In general, adverse-event counts depend on the period of observation, definitions, ascertainment, and reporting practices. The registry does not provide a detailed safety analysis or comparative confidence intervals. The small posted group counts and the terminated status are context for reading the figures, not a basis for drawing conclusions about a treatment’s safety.

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

The posted outcome data describe average percentage decreases in body weight from baseline to 12 months in each of the three groups. The mean is most negative for the obesity-centric approach with anti-obesity medication, followed by the obesity-centric program without medication and usual care. The associated 90% confidence intervals show uncertainty around each arm mean, not the range of individual responses.

These descriptive results do not establish the size or certainty of a treatment effect relative to another approach. For that, the relevant information would include clearly defined between-group contrasts, uncertainty intervals, and the analysis population. The registry’s one-sided t-test is labeled non-inferiority, but its specific contrast and margin are not reported. The reported p = 0.004 cannot fill in those missing details or establish the meaning of the test on its own.

The results do not show whether the approaches affected outcomes beyond body weight, such as diabetes management or other health measures. The record lists two outcome measures but no posted secondary outcome results. Nor do the 12-month results establish what happened over the approximately two-year intervention period described in the study plan.

The trial was randomized, which is an important design feature for comparisons, but actual enrollment was 74 rather than the planned 300. The registry does not report why the trial was terminated, how missing outcome data were handled, or whether the participants with posted outcomes represent all randomized participants. These reporting gaps limit how strongly the observed arm pattern can be attributed to the assigned approaches.

Accordingly, the record supports a descriptive account of the reported group means and intervals. It does not, by itself, support a definitive comparative conclusion about benefit, non-inferiority, or clinical importance. No medical conclusion should be inferred from the p-value or the ordering of the arm means alone.

14. Why This Trial Matters Statistically

This trial illustrates the difference between reporting group-level outcomes and reporting a fully interpretable comparison. Three arm means and their confidence intervals give readers a clear description of the average 12-month body-weight change in each group. A comparative clinical question, however, requires the contrast being tested and the uncertainty around that contrast. Those details are especially important when a reported test is labeled non-inferiority.

The record also illustrates why a study’s planned enrollment and actual enrollment should be kept distinct. It planned enrollment of 300 and reports actual enrollment of 74. The posted group results are correspondingly limited in size, but the registry does not provide enough information to determine the reason for the difference or its effect on the analysis. The planned figure is not evidence about the precision actually achieved.

The open-label, pragmatic design is relevant to interpretation because the interventions differ in their intensity and content. The registry describes usual care and two weight-management approaches, one with medication and one without. It does not describe adherence, treatment switching, measurement procedures, or the exact handling of data. Those details would help readers understand how closely observed outcomes reflect the intended approaches.

Finally, the registry’s non-inferiority label and p-value demonstrate the importance of reporting the margin, direction, and named comparison. Without them, a statistical test cannot be translated confidently into a clinical claim. Careful separation of descriptive findings from incompletely specified inference helps avoid overstating what the trial record establishes.

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 limitations specific to this record constrain interpretation. The study is listed as terminated, and its actual enrollment of 74 is below the planned enrollment of 300. The registry does not state why the trial was terminated or whether recruitment stopped before the planned target for a particular reason. The reason should not be inferred from the status alone.

The outcome results include arm means and 90% confidence intervals, but no separate dispersion values or comparison-specific effect estimates. The one-sided t-test is labeled non-inferiority, yet the record does not identify the tested contrast, the non-inferiority margin, the test direction, or its hypotheses. This prevents a complete assessment of the reported inferential result.

The registry does not report the primary analysis population, missing-data methods, exclusions, adherence, or treatment changes. The arm counts associated with the outcome do not clarify whether all randomized participants were included or how unavailable measurements were treated. The method for combining measurements between baseline and 12 months is also not described beyond the endpoint wording.

The open-label design and lack of reported detail on outcome measurement leave uncertainty about how weight was assessed and whether assessors were masked. The record also gives no posted secondary outcome results, although it lists two outcome measures. There is no reported sample-size justification, interim-analysis plan, multiplicity strategy, or detailed analysis plan. These items are not reported in the registry record; their absence should not be taken as proof that they were not considered outside the record.

The safety information is limited to event counts and a listed serious-event term. It does not provide detailed definitions, exposure time, attribution, or comparative analyses. Collectively, these limitations mean the record is most informative as a description of the posted arm-level weight changes, rather than as a complete account of comparative efficacy or safety.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: EMI-EHP Weight Management and Type 2 Diabetes Pragmatic Trial — NCT04531176.
• Registry record last updated 2024-09-19; status last verified 2024-08.
• Results first posted 2024-07-22.
• 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

NCT04531176 is a Phase 4 randomized, open-label, parallel-group trial recorded as terminated, with actual enrollment of 74. It compared a medically supervised weight management program with anti-obesity medication, the program without medication, and usual care for employees or spouses covered by an employer health plan with obesity and type 2 diabetes. The study description planned 300 participants and an approximately two-year intervention.

The primary outcome was mean percentage change in body weight between baseline and 12 months. The reported means were -8.74%, -6.68%, and -4.48%, respectively, with 90% confidence intervals of -10.74% to -6.74%, -8.71% to -4.65%, and -6.52% to -2.45%. The registry reports a one-sided t-test labeled non-inferiority, p = 0.004, but does not identify the comparison or margin. Missing-data methods, analysis population, and reason for termination are not reported in the registry record.