Multicenter Endoscopic Sleeve Gastrectomy (ESG) Trial (MERIT Trial): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Multicenter Endoscopic Sleeve Gastrectomy (ESG) Trial (MERIT Trial) (NCT03406975), a randomized na study. The primary outcome is Percentage of Excess Weight Loss (EWL).

2. Clinical Question

In statistical terms, the central comparison is between the two strategies assigned by randomization: ESG plus lifestyle intervention and lifestyle intervention alone. The endpoint is a continuous measure, so the reported group means describe average outcomes, while the standard deviations describe variation among participants. The registry lists an unadjusted mean difference of 46.0 for the treatment-versus-control comparison, with an interval from 38.05 to 53.95 and p < 0.001. These values describe the posted summary data; the registry record does not report a detailed analysis plan or a sponsor-reported analysis for this endpoint.

3. Trial Design

The MERIT trial was a multicenter, randomized, unmasked, interventional trial. The registry lists a crossover model and an enrollment of 208. It names Mayo Clinic as sponsor, reports the trial status as completed, and gives the start date as 2017-12-20 and primary completion date as 2020-10-29. The phase is listed as not applicable. The brief summary describes 9 United States locations.

Participants randomized to the control group received lifestyle modification alone during the first 12 months. Participants randomized to the ESG treatment group underwent the Overstitch ESG procedure at the start of Year 1 and also received a standard moderate-intensity lifestyle intervention during the first 12 months. The primary endpoint compares these randomized groups at 12 months from randomization.

The registry also reports a crossover group described as control participants moving from lifestyle intervention to ESG. The available description says crossover eligibility depended on compliance with at least 75% of visits in Year 1, an EWL or BMI condition at the week 52 visit, and no new psychosocial contraindications as judged by the treatment team. The text describing the crossover criteria and subsequent treatment is incomplete in the registry record. The record does not report enough detail to reconstruct all crossover procedures or the timing of every crossover-group result.

Group counts differ across outcomes, and safety denominators are not identical to the primary endpoint denominators for every group. This can occur when outcome-specific data are available for different subsets; the reasons for the particular counts are not reported in the registry record. Randomization supports the treatment-versus-control comparison as the principal between-group contrast, while the crossover group reflects a selected group with a different treatment history. The registry record does not report the original sample-size justification, allocation ratio, detailed follow-up procedures, or a full statistical analysis plan.

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.

Level 1: Lifestyle Modification vs ESG

Level 2: Crossover Lifestyle to ESG

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 Percentage of Excess Weight Loss (EWL), measured over 12 months.

The primary endpoint is mean percentage of excess weight loss at 12 months from randomization. It expresses weight loss in relation to baseline excess weight rather than as a raw change in body weight. The registry defines baseline excess weight using index weight and ideal weight at a BMI reference of 25 kg/m², and defines weight loss using follow-up weight and initial weight. The reported unit is percentage of EWL.

This endpoint is continuous: each participant contributes a measured value, and the group summary is a mean with a standard deviation. The mean communicates the average endpoint value in a group. The standard deviation describes how dispersed individual values are around that group mean; it is not the uncertainty around the mean. In the posted results, the control group has a mean of 3.2 and standard deviation of 18.6, while the ESG treatment group has a mean of 49.2 and standard deviation of 32.0. The crossover group is reported separately, with a mean of 44.1 and standard deviation of 35.7.

EWL is used as the primary outcome because the research question concerns weight loss relative to excess weight over the first 12 months. The registry also lists total body weight loss as a secondary outcome. The choice of EWL does not by itself establish that it captures every aspect of benefit, safety, durability, or quality of life; those are broader aims described in the study summary. The registry record does not report further rationale for selecting this endpoint or detailed rules for handling missing weight measurements.

7. Results Posted in the Registry

• Treatment Group: Overstitch ESG Procedure: mean 49.20 versus 3.20 on control; difference 46.00 (95% CI 38.05 to 53.95), standardized effect d = 1.84; z-test p = <0.001.
• Crossover Group: Lifestyle Intervention to ESG Procedure: mean 44.10 versus 3.20 on control; difference 40.90 (95% CI 31.95 to 49.85), standardized effect d = 1.53; z-test p = <0.001.

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.

Number of Participants With ≥25% EWL

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Treatment Group: Overstitch ESG Procedure: 76.6% (59/77) versus 11.8% (13/110) on control; risk difference 64.8 percentage points (95% CI 53.6 to 76.0), risk ratio 6.48 (3.83–10.96); unadjusted two-proportion z-test p = <0.001.
  • Crossover Group: Lifestyle Intervention to ESG Procedure: 70.8% (51/72) versus 11.8% (13/110) on control; risk difference 59.0 percentage points (95% CI 46.9 to 71.1), risk ratio 5.99 (3.52–10.20); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Treatment Group: Overstitch ESG Procedure than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Crossover Group: Lifestyle Intervention to ESG Procedure than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Number of Participants Off Or With Reduction in Antihypertensive Medications

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Treatment Group: Overstitch ESG Procedure: 36.1% (13/36) versus 18.8% (9/48) on control; risk difference 17.4 percentage points (95% CI -1.8 to 36.5), risk ratio 1.93 (0.93–4.00); unadjusted two-proportion z-test p = 0.073.
  • Crossover Group: Lifestyle Intervention to ESG Procedure: 48.3% (14/29) versus 18.8% (9/48) on control; risk difference 29.5 percentage points (95% CI 8.2 to 50.8), risk ratio 2.57 (1.28–5.18); unadjusted two-proportion z-test p = 0.006.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Treatment Group: Overstitch ESG Procedure and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Crossover Group: Lifestyle Intervention to ESG Procedure than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Number of Participants Off or With Reduction in Diabetes Medications

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Treatment Group: Overstitch ESG Procedure: 84.6% (11/13) versus 11.1% (3/27) on control; risk difference 73.5 percentage points (95% CI 50.6 to 96.4), risk ratio 7.62 (2.56–22.69); unadjusted two-proportion z-test p = <0.001.
  • Crossover Group: Lifestyle Intervention to ESG Procedure: 71.4% (10/14) versus 11.1% (3/27) on control; risk difference 60.3 percentage points (95% CI 33.9 to 86.8), risk ratio 6.43 (2.10–19.65); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Treatment Group: Overstitch ESG Procedure than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Crossover Group: Lifestyle Intervention to ESG Procedure than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary outcomes broadly extend the primary weight-loss question to threshold attainment, medication use, blood pressure, glycemic measures, and total body weight loss. The registry posts numerical results for some secondary outcomes and names additional ones, but the record does not provide sponsor-reported analyses for these outcomes. The pattern across the posted results is consistent with the primary endpoint in that the listed weight-related threshold and diabetes-medication comparisons favor the ESG groups in their unadjusted estimates. The antihypertensive-medication comparisons are less consistent: the randomized contrast has an interval that includes zero, while the crossover contrast does not.

These outcomes are related but not interchangeable. Meeting an EWL threshold is a categorical summary of weight loss; medication changes, blood pressure, and HgbA1c address different clinical measures. A pattern across outcomes can add context, but it does not establish that a change in one measure caused a change in another. The primary endpoint remains the prespecified central result described by the registry; secondary endpoints provide additional observations rather than independent confirmation of the same estimand.

Unadjusted secondary results carry less inferential weight when many outcomes and comparisons are examined. The registry lists 17 outcome measures and 16 posted secondary outcomes, while reporting no analyses in the registry record. Without a reported multiplicity strategy, the risk of chance findings across multiple tests cannot be assessed or controlled from the posted results. The unadjusted p-values and intervals are useful descriptions of the observed data, but they should not be read as a set of confirmatory conclusions.

The crossover comparisons require additional caution because the crossover group was formed from control participants who met registry-described conditions after the initial assignment. Its results therefore do not provide the same randomized evidence as the initial treatment-versus-control comparison. The reasons for differing endpoint denominators, the missing-data approach, and analysis populations are not reported in the registry record. These limitations apply to the secondary pattern as a whole and make the detailed outcome-by-outcome interpretation dependent on information that is not available in the record.

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

• Number of Participants With ≥25% EWL
• Number of Participants Off Or With Reduction in Antihypertensive Medications
• Change in Blood Pressure
• Change in Blood Pressure for ESG Intervention From Baseline to 24 Months
• Number of Participants Off or With Reduction in Diabetes Medications
• Change in HgbA1c
• Change in HgbA1c for ESG Intervention From Baseline to 24 Months
• Percentage of Total Body Weight Loss (TBWL)

9. Statistical Interpretation

10. Statistical Methods Explained

The principal outcome is continuous, so its reported group summaries are means and standard deviations. A mean is the arithmetic average of the observed endpoint values in a group. A standard deviation describes how much individual values vary around that average. A larger standard deviation indicates more spread in the observations, but it does not by itself indicate a less precise estimate of the treatment contrast; precision also depends on the amount of information contributing to the estimate.

The reported mean difference compares the treatment-group mean with the control-group mean. For the primary comparison, the registry lists 49.2 for ESG treatment and 3.2 for lifestyle-only control, and the posted unadjusted mean difference is 46.0. The difference is expressed in percentage points of EWL, consistent with the outcome scale. It is an average contrast, not a statement about each participant’s experience. The record also includes a comparison of the crossover group with control, but that contrast should be interpreted descriptively because crossover participants were selected after initial assignment.

A confidence interval gives a range of values compatible with the observed data under the statistical method and its assumptions. A narrower interval indicates greater precision than a wider interval for the same quantity and method. For the primary mean difference, the reported interval runs from 38.05 to 53.95. The registry does not state the confidence level or provide the method used to form these intervals, so those details should not be assumed. The interval should not be read as a range containing every individual participant’s outcome.

The p-value assesses how inconsistent the observed data are with a specified null hypothesis under the calculation’s assumptions. A small p-value is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The posted treatment-versus-control primary comparison has p < 0.001. The registry lists these comparisons as unadjusted recalculations based on posted summary statistics, and reports no sponsor-reported analyses. The registry record does not describe the sponsor’s original model or a prespecified inferential procedure.

For binary outcomes, the registry reports participant counts and unadjusted risk differences. A risk difference compares the proportions experiencing the outcome in two groups. The primary secondary binary endpoint is the number of participants with at least 25% EWL. The treatment group has 59 participants meeting that threshold among 77, and control has 13 among 110. The corresponding risk difference is 0.648, with an interval from 0.536 to 0.760. The p-value is reported as p < 0.001. This comparison answers a different question from the continuous primary endpoint: it focuses on meeting a threshold rather than the average EWL value.

Intervals for a risk difference can include zero when the data are compatible with no difference as well as with differences in either direction. In the antihypertensive-medication outcome, the randomized treatment-versus-control risk-difference interval runs from -0.018 to 0.365, and p = 0.073. In contrast, the crossover-versus-control interval runs from 0.083 to 0.508, with p = 0.006. These are unadjusted results. They should not be treated as interchangeable evidence because the crossover group has a different selection and treatment history.

Over-dispersion is a general statistical concern when the observed variability in data exceeds what a model assumes. For count or binary outcomes, extra variability can make standard errors too small if a model does not account for it, producing intervals that are too narrow and p-values that are too small. The registry does not report a model-based analysis or an over-dispersion assessment for the posted secondary outcomes. This is a general explanation of why model assumptions matter, not a claim that over-dispersion occurred in this trial.

Randomization is intended to make groups comparable at baseline on average, supporting interpretation of differences between assigned groups. That logic applies most directly to the initial randomized ESG and lifestyle-only groups. It does not automatically extend to a group formed through later crossover eligibility, because entry into that group depended on post-randomization conditions described in the registry. Crossover comparisons can therefore reflect both treatment and selection. The registry record does not report an analysis accounting for those selection processes.

Multiplicity arises when many outcomes, time points, or group comparisons are examined. Even when each individual test uses a conventional threshold, testing many hypotheses increases the chance of at least one apparently strong result arising by chance. The record lists 17 outcome measures, 16 posted secondary outcomes, and no reported secondary analyses or multiplicity adjustment. The appropriate interpretation is therefore cautious: the unadjusted secondary results describe patterns in the posted data, but they do not establish a confirmatory family of findings. The registry record does not report which outcomes, if any, were prioritized in a formal testing hierarchy.

The registry reports outcome-specific sample sizes, which are not identical across endpoints. Missing-data handling, analysis-population definitions, and reasons for endpoint-specific availability are not reported in the registry record. Without those details, it is not possible to assess whether all randomized participants contributed to each analysis, whether imputation was used, or how exclusions may have affected estimates. These are important elements of a full statistical analysis, especially when outcome availability differs across groups or endpoints.

11. Confidence Intervals and Precision

The primary mean difference is reported as 46.0, with an interval from 38.05 to 53.95. The interval is centered around the estimated contrast in a broad sense and excludes zero, consistent with the reported p < 0.001 under the unadjusted calculation. The interval gives more information than the p-value alone because it shows a range of plausible values for the group-average contrast under the method’s assumptions. The registry does not state the confidence level or interval-construction method, so neither should be inferred.

Precision is different from the observed size of an effect. A large estimate can still be imprecise if its interval is wide, and a small estimate can be precise if its interval is narrow. In this record, the standard deviations show substantial spread in individual EWL values, while the interval around the mean difference expresses uncertainty in the estimated group contrast. These quantities answer different questions and should not be conflated.

The binary outcome intervals also convey precision. For the at-least-25%-EWL comparison, the randomized treatment-versus-control risk difference interval runs from 0.536 to 0.760. For reduction in antihypertensive medications, the randomized comparison interval runs from -0.018 to 0.365 and includes zero. The diabetes-medication comparison has an interval from 0.506 to 0.964. These intervals are unadjusted; the registry record does not report a multiplicity adjustment or a stated confidence level.

Interval width depends on the amount and distribution of information available, along with the statistical method. Outcome-specific denominators vary, and some secondary outcomes have relatively limited numbers of participants in the listed groups. The registry does not report whether missing data, exclusions, or other features affected precision. A confidence interval describes uncertainty for an estimated group contrast; it does not show the range of individual responses, prove a mechanism, or establish that a result will be replicated in another study.

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 figures are planning illustrations based on fractions of the observed effects, not the trial’s original design calculations. All figures in these comparisons are participants per arm. The planning assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is the stated fraction of the observed effect. These figures are not observed power.

The 80% power column gives the estimated participants per arm for a future study to have an 80% chance of detecting the selected effect under those assumptions, if that effect and the model assumptions hold. The 90% power column gives the corresponding planning count for a 90% chance. Greater planned power requires more participants for a given effect and assumptions. The figures should be read as illustrations rather than as a recommendation or a reconstruction of the MERIT trial’s design.

Rows describe planning for 100%, 75%, or 50% of the observed effect. The smaller the effect chosen for detection, the more participants are generally required to distinguish it from chance variation with the same significance level and power target. This pattern is visible across the listed calculations. For the primary continuous comparison, planning for 100% of the observed effect gives 5 participants per arm for 80% power and 7 for 90% power; planning for 50% gives 19 and 25. The crossover comparison uses its own observed estimate and has different planning figures.

These calculations do not account for every practical design issue, such as loss to follow-up, the crossover structure, or any adjustment strategy that may be used in a future study. The registry’s original sample-size justification and assumptions are not reported in the registry record. The plotted curve should be read as a visual expression of the same planning relationship: as the effect targeted for detection becomes smaller, the required sample size rises under the stated assumptions.

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 = 46.000; pooled standard deviation σ = 24.990.
2. n = 2 × (1.96 + 0.842)² × 24.990² / 46.000² = 4.6, rounded up to 5 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 <- 46                      # observed difference in means
sigma <- 24.9901212914656                     # 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(3406975)
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 = 5 (80% power), 7 (90% power)
 75% of observed effect: n per arm = 9 (80% power), 12 (90% power)
 50% of observed effect: n per arm = 19 (80% power), 25 (90% power)
power.t.test: 5.8 
Simulated power at n = 5 per arm: 0.720 (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. Multiplicity, Missing Data and Analysis Populations

The registry lists 17 outcome measures and 16 posted secondary outcomes, including measures of EWL, medication use, blood pressure, HgbA1c, and total body weight loss. Considering many outcomes creates a multiplicity problem: the more statistical comparisons examined, the greater the possibility that some results will appear notable by chance even if there is no underlying difference for those outcomes. An adjustment or prespecified testing hierarchy can address this issue, but the registry record does not report one.

The posted secondary comparisons are unadjusted, and the record reports no sponsor-reported analyses for them. Consequently, their p-values should be read as individual, unadjusted summaries rather than as multiplicity-controlled evidence. The registry record does not report whether any outcomes were designated as key secondary outcomes, whether testing was hierarchical, or whether any correction was planned or performed.

Missing-data procedures are also not reported in the registry record. The group counts vary across outcomes, and the safety denominators differ from some efficacy denominators. The record does not explain whether these differences reflect outcome-specific eligibility, incomplete assessments, exclusions, or other reasons. It also does not specify whether missing observations were imputed, whether complete-case analyses were used, or whether sensitivity analyses assessed the effect of missing data.

Analysis populations are not reported in sufficient detail to establish whether the posted analyses follow intention-to-treat, per-protocol, or another approach. This matters because excluding participants or analyzing only those with available outcome data can affect the balance created by randomization. It is particularly relevant to the crossover group, where entry depended on post-randomization criteria described in the registry. The record does not report methods to account for that selection.

The original statistical analysis plan, model specifications, covariate adjustments, interim analyses, and sample-size justification are not reported in the registry record. These omissions do not invalidate the posted descriptive data, but they limit assessment of whether the reported analyses match prespecified procedures and how robust the results are to alternative assumptions.

15. 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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: abdominal pain (3.9%), nausea (2.6%), vomiting (2.6%), esophageal mucosal tear due to device malfunction (1.4%), malnutrition (1.4%).

The safety record reports serious adverse events in 1 participant among 110 at risk in the lifestyle-only control group, 12 among 77 at risk in the ESG treatment group, and 4 among 73 at risk in the crossover group. Deaths are reported as 0 in each group. The registry also reports other, non-serious adverse events affecting 46 participants in control, 77 in the treatment group, and 64 in the crossover group, using the corresponding listed at-risk denominators.

The serious-event terms with the highest reported percentages include abdominal pain at 3.9%, nausea at 2.6%, vomiting at 2.6%, esophageal mucosal tear due to device malfunction at 1.4%, malnutrition at 1.4%, and post ESG hospital stay at 1.4%. These percentages are quoted as listed in the registry. The record does not give enough detail to establish event duration, severity grading beyond the serious-event classification, timing, relatedness assessments, or whether participants experienced more than one event.

These counts describe reported events within the registry’s safety reporting. They are not adjusted comparisons and do not, on their own, establish that an intervention caused an event or that one strategy has a particular overall safety profile. Group follow-up and safety-analysis methods are not reported in the registry record.

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

The primary result describes a higher observed mean percentage of EWL in the randomized ESG treatment group than in the lifestyle-only control group at 12 months. The reported means are 49.2 and 3.2, with an unadjusted mean difference of 46.0 and an interval from 38.05 to 53.95. The interval and p < 0.001 indicate that the posted data are not closely compatible with a null mean difference under the unadjusted calculation. The result is an average contrast and does not describe every participant’s outcome.

The result does not establish that all participants experienced the same degree of weight loss, that the procedure is suitable for every person with obesity, or that outcomes persist beyond the reported follow-up. The registry summary states that long-term weight loss, safety, durability, and quality of life were aims of the research, but the primary endpoint is at 12 months. Further conclusions about those broader aims require results and methods not reported in the registry record.

The crossover-group estimates should not be interpreted as if that group were randomized at the outset to ESG. These participants came from the control group and crossed over under conditions described in the registry. Their outcome can reflect eligibility, selection, and treatment history as well as the procedure. The randomized treatment-versus-control comparison is the clearest contrast for the initial assigned strategies.

The secondary outcomes provide context but involve multiple measures, unadjusted comparisons, and outcome-specific denominators. The registry does not report a multiplicity plan, missing-data approach, analysis populations, confidence-interval level, or full statistical analysis plan. These limits affect how strongly secondary findings and precision claims can be interpreted. The statistics describe the posted record; they do not substitute for a complete clinical evaluation or a causal analysis with fully documented methods.

17. Why This Trial Matters Statistically

This trial is statistically interesting because it combines an initial randomized comparison with a crossover structure. The primary randomized contrast compares ESG plus lifestyle intervention with lifestyle modification alone at 12 months. That design addresses a direct comparative question while preserving a distinction between assignment at the start and treatment received later. The crossover group illustrates why treatment history and selection matter when interpreting later group summaries.

The primary endpoint is continuous and reported with both means and standard deviations, while a secondary weight-loss outcome converts the measure into a threshold count. These different summaries answer different questions. Means retain information about the scale of EWL, whereas a threshold outcome focuses on whether a specified level was reached. Reading them together can show complementary aspects of the observed results, without treating them as independent replications.

The record also highlights the importance of precision and multiplicity. The primary contrast has a reported interval and an unadjusted p-value, while multiple secondary outcomes and comparisons are listed without reported sponsor analyses or a multiplicity strategy. A statistically compelling primary estimate does not automatically make every secondary finding confirmatory. Confidence intervals, outcome definitions, analysis populations, and missing-data handling all affect what conclusions can reasonably be drawn.

Finally, the planning illustrations show how assumptions about the effect targeted for detection influence required sample sizes. They are explicitly based on observed effects and stated planning assumptions, not on the trial’s original design. Keeping those planning calculations separate from observed results avoids confusing a hypothetical future-study calculation with the evidence collected in this trial.

18. 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 does not include a full statistical analysis plan, the original sample-size justification, the prespecified primary analysis model, or the confidence level and construction method for the posted intervals. It also does not report interim analyses, covariate adjustments, analysis-population definitions, or missing-data procedures. These gaps limit independent assessment of whether the posted calculations match the trial’s planned analyses.

The crossover-group description is incomplete, and the group was formed from control participants meeting post-randomization conditions. This selection makes its comparison with control less straightforward than the initial randomized comparison. The record does not report a method to account for selection or provide enough detail to reconstruct all crossover timing and follow-up.

Outcome-specific sample sizes vary, and the reasons for the differences are not reported. This makes it difficult to determine whether variation in denominators reflects missing measurements, eligibility differences, or other analysis decisions. The registry reports numerous outcomes and unadjusted secondary comparisons but does not report a multiplicity strategy. Secondary p-values should therefore be interpreted cautiously.

The primary endpoint is reported at 12 months. Although long-term durability and other outcomes are described as study aims, the record does not provide detailed long-term primary results in the data summarized. Safety counts are also descriptive; the registry record does not provide event timing, relationship assessments, or complete details of severity and recurrence. These limitations concern the information available in the registry record and do not, by themselves, determine the quality of trial conduct.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Multicenter Endoscopic Sleeve Gastrectomy (ESG) Trial (MERIT Trial) — NCT03406975.
• Registry record last updated 2022-10-18; status last verified 2022-09.
• Results first posted 2022-10-18.
• 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.

22. Record Summary

The MERIT trial was a completed, randomized, unmasked, multicenter study with a crossover model and registered enrollment of 208. It compared ESG plus lifestyle intervention with lifestyle modification alone for the primary outcome of percentage of EWL at 12 months. The posted randomized-group means were 49.2 for ESG and 3.2 for control; the unadjusted mean difference was 46.0, with an interval from 38.05 to 53.95 and p < 0.001.

The registry also reports a crossover group, secondary outcomes, and adverse-event counts. Crossover findings are descriptive because the group was selected after initial randomization. Secondary comparisons are unadjusted, and the registry record does not report a full statistical analysis plan, multiplicity adjustment, missing-data procedures, or analysis populations. The results therefore describe the posted estimates and their stated limitations rather than resolving every question about long-term outcomes or safety.