Evaluation of EverLift in the Performance of Polypectomy for Polyps 4-9mm: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluation of EverLift in the Performance of Polypectomy for Polyps 4-9mm (NCT04551014), a randomized na study. The primary outcome is Number of Polyps With Complete Resection.

2. Clinical Question

This is a procedural comparison, not a comparison of cancer outcomes. The primary measure counts completely resected polyps rather than reporting a participant-level proportion or a rate per polyp. The registry reports 140 in the EverLift arm and 145 in the arm without EverLift, with participant counts of 105 and 109, respectively. Those counts should not be interpreted as resection percentages: the registry does not provide the total number of polyps assessed in each arm as a denominator for such percentages. It also does not describe how multiple polyps within a participant were handled in the statistical analysis. The central question is therefore narrower than whether the procedures differ in all aspects of colonoscopy: it concerns the reported count of completely resected polyps under the two assigned procedures.

3. Trial Design

The registry record describes a randomized, triple-masked, parallel-group controlled trial of polypectomy with or without submucosal EverLift injection. The study is marked completed and lists an enrollment of 158. It identifies Stanford University as the sponsor. The registered study period began on 2020-09-16, and primary completion is listed as 2021-05-31. The phase is recorded as not applicable.

One arm received polypectomy with EverLift, and the comparison arm received polypectomy without EverLift. The arm descriptions specify polyps 4-9mm. The primary outcome is the number of polyps with complete resection, with a timeframe of 2 weeks. The primary-outcome population is described as participants with polyps randomized to the respective procedure. The registry reports 105 participants for the EverLift primary outcome and 109 for the arm without EverLift. These outcome-group counts do not match the overall enrollment figure, and the registry record does not explain the relationship between them. It is not reported in the registry record whether the difference reflects outcome-specific eligibility, analysis populations, or another reason.

The registered secondary outcomes cover time to resection, the number of polyps requiring multiple snare attempts to achieve complete endoscopic polypectomy, and the number of participants with procedural complications. The posted secondary results include time to resection and participants with procedural complications; the registry also lists the multiple-snare-attempt outcome among its secondary outcomes, but numerical results for it are not included in the record summarized by these results. Randomization, masking procedures, concealment details, the allocation ratio as implemented, and the precise masking roles are not reported in the registry record. Nor is a full protocol or a detailed statistical analysis plan reported, so those design features should not be inferred from the brief design labels.

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: Each polyp sized 4-9 mm underwent block randomization to being removed by cold snare polypectomy (CNP) with or without submucosal injection of EverLift™.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Participants with polyps randomized the respective procedure are included in the analysis

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 Number of Polyps With Complete Resection, measured over 2 weeks.

The primary endpoint is the number of polyps with complete resection, assessed at 2 weeks. It addresses the trial’s stated procedural question: whether completeness of resection differed between polypectomy performed with submucosal EverLift injection and polypectomy performed without it. Complete removal is a direct procedural outcome relevant to the stated comparison, and a count can describe how many polyps in each arm met the registry’s definition of complete resection.

The posted results give counts of 140 and 145, alongside participant counts of 105 and 109. These are counts, not percentages or risks. The registry does not report the total number of eligible or treated polyps by arm, nor a participant-level denominator for calculating a complete-resection proportion. It also does not report the operational definition or assessment procedure used to determine complete resection, beyond naming the outcome. Consequently, the counts alone do not reveal the probability that an individual polyp was completely resected, and they should not be converted into such probabilities.

The registry records a two-sided t-test and a p-value of 0.424 for the primary comparison, classified as non-inferiority. A p-value does not state the size or precision of a treatment contrast, and the record contains no confidence interval or reported test statistic for this endpoint. The non-inferiority margin and the direction and decision rule for the test are not reported in the registry record. The endpoint can therefore be described as the registry’s chosen measure of completeness, but the record does not contain enough detail to assess the full non-inferiority framework.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods are useful only when the outcome scale, analysis population, and decision rule are clear. This record gives some of that information, but not all of it. It labels the trial randomized and triple-masked, identifies a parallel-group comparison, and lists a two-sided t-test for the primary analysis. It does not report the full analysis plan, the primary effect estimate, a confidence interval, or a non-inferiority margin. The methods can therefore be explained at the level of the posted information, while important design details remain unresolved.

What does the primary p-value mean?

A p-value describes how incompatible the observed data are with a specified statistical null model, under the assumptions of the analysis. It is not the probability that the null hypothesis is true, the probability that the result occurred by chance, or a measure of the clinical importance of a difference. The registry reports p = 0.424 from a two-sided t-test for the primary count outcome. The record does not give the test statistic, the estimated difference, its confidence interval, or sufficient detail to establish precisely how the t-test was applied to the count data. Thus the p-value should be read as a reported test result, not as a complete description of the effect.

A non-inferiority analysis asks whether the data rule out a loss larger than a prespecified acceptable margin. That question requires the margin, a clearly specified direction, and an inferential procedure aligned with the hypothesis. The registry labels the primary analysis non-inferiority, but the margin and decision rule are not reported in the registry record. A p-value alone, especially without the estimate and interval, cannot establish that the non-inferiority criterion was met. Nor does a non-significant difference from a conventional superiority test demonstrate equivalence.

Why do counts and denominators matter?

The primary endpoint is recorded as a count of polyps, while the arm sample sizes are expressed as participants. The registry gives the number of completely resected polyps, but not the total number of polyps assessed in each arm. A count can be meaningful as an outcome, but a rate or proportion requires a suitable denominator. Without the total polyp counts, readers cannot recover a complete-resection proportion from the posted information. If participants contributed more than one polyp, observations within a participant could also be statistically related. Whether that issue applies, and whether the analysis accounted for it, is not reported in the registry record. This is a general consideration for outcomes measured on multiple lesions per participant, not a claim about what occurred in this trial.

What does a confidence interval add?

A confidence interval gives a range of values compatible with the data under a specified model and procedure. It helps distinguish a precise estimate from one that is too uncertain to support a narrow conclusion. The primary outcome has no reported confidence interval or effect estimate in the registry record. The secondary continuous outcome has an unadjusted re-calculated mean difference and interval, while the complication comparison has an unadjusted re-calculated risk difference and interval. Those calculations describe their respective secondary outcomes; they do not supply the missing primary estimate or interval.

How should the continuous secondary analysis be read?

Time to resection is reported as a mean and standard deviation for each arm. A t-test compares group means. In general, this approach relies on assumptions about the sampling distribution of the estimated mean difference and the independence of the observations; the registry record does not report diagnostic checks or a detailed model specification. The posted means are 109.8 seconds with EverLift and 38.8 seconds without EverLift. The reported analysis gives p < 0.001. A simple, unadjusted re-calculation from the posted summary statistics gives a mean difference of -71.0 seconds for without EverLift relative to with EverLift, with a 95% interval from -85.81 to -56.19 and p < 0.001. The reported and re-calculated p-values use different displayed precision; the record does not indicate a discrepancy flag. The re-calculation is not a replacement for an analysis adjusted for any design features that are not reported.

Why does over-dispersion matter for counts?

General explanation: count outcomes are often more variable than a simple model assumes, especially when counts differ substantially across units or when events cluster. Over-dispersion means that observed variability exceeds the model’s expected variability; ignoring it can make standard errors too small and intervals too narrow. The registry lists a count endpoint but does not report a count model, a dispersion estimate, or a check for over-dispersion. It is therefore not possible to determine from the record whether this issue was assessed or whether it affected the primary analysis. The point is a general methodological consideration, not a finding about these data.

How should the planning illustrations be understood?

The sample-size illustrations are computed from observed secondary effects, not from the trial’s own design assumptions. They use a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation, with the target effect set at 100%, 75%, or 50% of the observed effect. For the time outcome, the illustrations give participants per arm for 80% and 90% power at each of those target effects. The complication illustrations do the same for that binary outcome. These are planning scenarios, not observed power, and they do not show what power the completed trial had for its primary outcome. Estimates based on observed effects can be unstable, particularly when effects are imprecise or event counts are small. The record does not report the trial’s original sample-size justification or the assumptions used to plan enrollment.

Finally, the secondary p-values are unadjusted. General explanation: when multiple outcomes or comparisons are tested, the chance of at least one apparently unusual result can increase. Adjustment procedures or a prespecified hierarchy can address that issue, but the registry record does not report a multiplicity strategy. This does not invalidate a result by itself; it means that isolated secondary p-values warrant more caution than a single prespecified primary test interpreted under a complete analysis plan.

10. Analyses Reported by the Sponsor

The registry reports a two-sided t-test for the primary outcome with p = 0.424 and classifies the analysis as non-inferiority. It does not report a primary effect estimate, confidence interval, or non-inferiority margin. The available record therefore does not permit a direct numerical comparison between a primary re-calculation and the sponsor’s analysis: no primary re-calculated contrast is reported because the endpoint is a count outcome without the information needed for that comparison. The reported p-value should be retained as the registry’s result, while its non-inferiority interpretation remains difficult to assess without the missing margin and decision details.

For time to resection, the registry’s reported two-sided t-test has p < 0.001. A simple unadjusted re-calculation from the posted group means and standard deviations gives p < 0.001 and an estimated mean difference of -71.0 seconds for the arm without EverLift relative to the arm with EverLift. These p-values are consistent at their displayed precision in indicating a strong separation under the respective tests, but the registry does not provide the sponsor’s test statistic or full calculation details. The difference in printed thresholds should not be treated as an exact numerical contradiction.

For procedural complications, the registry does not report a sponsor analysis. The posted participant counts support an unadjusted re-calculation, but that is not a sponsor result. Re-calculations based on summary data are descriptive checks and may not reproduce an analysis using participant-level data, covariate adjustment, clustering, or other methods. The record does not describe such methods for these outcomes. No discrepancy flag is recorded for this trial, so there is no stated conflict requiring one analysis to take precedence over another.

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists one primary outcome and three secondary outcomes, but it does not report how the testing of these outcomes was coordinated. No multiplicity adjustment, testing hierarchy, or division between confirmatory and exploratory outcomes is described. This matters because each additional statistical test creates another opportunity for a result to appear notable even when there is no underlying effect. General statistical methods can address this through prespecified procedures, but readers cannot determine whether any such approach was used. The posted secondary p-values are therefore best regarded as unadjusted evidence for their individual outcomes, rather than as a family of tests with a documented overall error rate.

The primary outcome is classified as non-inferiority, yet the margin, direction, and decision rule are not reported in the registry record. Those details are central to judging a non-inferiority result. The record also omits the primary effect estimate and confidence interval, so the result cannot be checked against a proposed margin. The reported p-value of 0.424 does not fill those gaps. It cannot by itself establish that a clinically acceptable loss was excluded, nor can it establish equivalence.

Missing-data methods are also not reported in the registry record. The participant counts attached to the primary outcome, 105 and 109, differ from the overall enrollment of 158, but the record does not explain why or identify the analysis populations in enough detail to reconcile them. The primary population is described as participants with polyps randomized to the respective procedure. It is not reported whether all randomized participants contributed data, how missing or unevaluable outcome assessments were handled, or whether the analyses followed intention-to-treat, per-protocol, or another population definition. These omissions matter for a non-inferiority question because different analysis populations can affect interpretation.

The endpoint counts are numbers of polyps, but group sample sizes are stated as participants. The registry does not report the number of polyps assessed or the treatment of multiple polyps from one participant. It is not reported in the registry record whether within-participant dependence was accounted for, or whether any participants contributed more than one polyp. No assumptions about that data structure should be made from the counts alone. Without these details, readers cannot fully assess the analysis population, denominator, or precision of the primary comparison.

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

Serious events with the highest arm-level rate: splenic artery thrombosis (1.0%), post-polypectomy bleeding (0.9%).

The safety record lists 105 participants at risk in the EverLift group and 109 in the group without EverLift. It reports serious adverse events affecting 1 participant with EverLift and 2 without EverLift. Other adverse events are recorded as affecting 0 participants in each group, and deaths are recorded as 0 in both groups. These are small event counts, so they provide limited precision for comparing uncommon harms. A difference in observed counts should not be interpreted as proof that the underlying risks differ, or as proof that they are the same.

The registry lists splenic artery thrombosis at 1.0% and post-polypectomy bleeding at 0.9% among the top serious adverse event terms. The record does not give a mapping from each listed term to a treatment arm, and the terms should not be assigned to an arm based on the group totals. The record also does not provide further case descriptions, severity details beyond the serious-event classification, or an adjudication method. It is therefore not possible to determine from these entries whether an event was caused by a procedure or device.

These safety figures describe what the registry reports for the listed follow-up; they do not establish a comprehensive comparative safety profile. The complication outcome is listed with a timeframe of 4 weeks, from procedure through 30 day follow-up. Interpretation should remain limited to these reported counts and terms, with no causal conclusion drawn from the sparse events.

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

The result establishes that the registry contains a randomized comparison of polypectomy with and without EverLift injection and reports counts of completely resected polyps for each arm. It also reports a primary two-sided t-test p-value and labels the analysis non-inferiority. Those details describe the registered comparison and its posted outcome; they do not, on their own, establish a complete-resection rate, a clinically important difference, or a non-inferiority conclusion.

The primary counts are 140 and 145, with participant counts of 105 and 109. Because the total number of polyps assessed in each arm is not reported, the counts cannot be transformed into polyp-level rates from the registry record. The missing primary estimate and confidence interval also prevent an assessment of the range of effect sizes compatible with the data. Most importantly for the non-inferiority label, the margin and decision rule are not reported. A p-value of 0.424 is not evidence of equivalence and is not sufficient to show that a prespecified non-inferiority criterion was met.

The secondary outcomes provide further observations, including procedure time and participant-level complications, but they measure different things. The time analysis has a reported p-value below < 0.001, and the complication comparison has a broad re-calculated interval that includes zero. These results do not determine the primary endpoint, and unadjusted secondary tests should be interpreted cautiously when the registry does not report how multiple outcomes were handled. The sparse safety counts also leave substantial uncertainty about uncommon events.

Several key methods details are not reported in the registry record: the original sample-size justification, detailed statistical plan, handling of missing outcomes, analysis population definitions, and the treatment of multiple polyps within participants. The record therefore supports a description of the trial and its posted outcomes, but not a definitive assessment of the primary non-inferiority claim or an overall clinical conclusion. It does not show that one procedure should be chosen over the other.

14. Why This Trial Matters Statistically

This trial raises a useful statistical issue: the unit named in an outcome may differ from the unit used to describe the sample. The primary outcome counts polyps, while the arm sizes are expressed as participants. When a participant can contribute multiple lesions, the analysis needs to make clear how those observations are related and how the denominator is defined. The registry record does not give the total polyps assessed in each arm or describe how any within-participant dependence was handled. This makes it difficult to interpret the count as a rate or assess the precision of the comparison.

The non-inferiority label also illustrates why a test name is not enough to interpret a result. Non-inferiority depends on a justified margin and a clear comparison of the estimated effect and its uncertainty with that margin. The record supplies neither the margin nor a primary interval. The reported p-value of 0.424 cannot answer the full non-inferiority question by itself. This is a general lesson for reading clinical trial registries: the conclusion depends on the prespecified estimand and decision rule, not just on whether a p-value crosses a conventional threshold.

Finally, the trial has a notable contrast between the detailed secondary time result and the limited primary statistical reporting. The time outcome can be described through means, standard deviations, and an unadjusted re-calculation, whereas the primary count comparison lacks an effect estimate and interval. These outcomes are not interchangeable. The contrast underscores the value of complete reporting for the primary estimand, including denominators, uncertainty, missing-data handling, and analysis populations. Those details are necessary for readers to distinguish a clearly measured effect from a result whose interpretation remains limited by the registry record.

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 trial constrain interpretation of the posted results. The primary endpoint is a count of completely resected polyps, but the record does not provide the total number of polyps assessed in either arm. The reported participant counts therefore cannot serve as the denominator for a polyp-level resection rate. The record also does not specify how multiple polyps within a participant were handled or whether the analysis accounted for possible dependence between lesions from the same participant.

The primary result is reported as a two-sided t-test with p = 0.424 and a non-inferiority classification, but no primary effect estimate or confidence interval is given. The non-inferiority margin, direction, and decision rule are not reported in the registry record. Without those details, the basis for interpreting the analysis as non-inferiority cannot be assessed. The relation between overall enrollment of 158 and the primary outcome counts of 105 and 109 is also not explained. Analysis-population definitions and missing-outcome handling are not reported.

The record does not report the trial’s original sample-size justification, detailed statistical analysis plan, interim analyses, or the procedures used to maintain masking and conceal allocation. The phase is recorded as not applicable, but no further design rationale is given. Numerical results are not reported for the listed multiple-snare-attempt secondary outcome. No multiplicity strategy is reported for the secondary analyses, and the safety record provides sparse event counts without further case-level details or an arm assignment for the listed serious event terms.

These limitations do not establish that the trial was conducted improperly; they identify information absent from the registry record that is needed for a more complete statistical assessment. In particular, they limit conclusions about the magnitude and precision of the primary comparison and about the non-inferiority claim.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Evaluation of EverLift in the Performance of Polypectomy for Polyps 4-9mm — NCT04551014.
• Registry record last updated 2022-09-01; status last verified 2022-08.
• Results first posted 2022-09-01.
• 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

NCT04551014 is a completed randomized, triple-masked, parallel-group trial sponsored by Stanford University. It compared polypectomy with submucosal EverLift injection against polypectomy without injection for polyps 4-9mm. The primary outcome was the number of polyps with complete resection at 2 weeks. The registry reports counts of 140 and 145, with participant counts of 105 and 109, and a two-sided t-test p-value of 0.424 classified as non-inferiority.

Posted secondary results include time to resection and participants with procedural complications. The registry also lists multiple snare attempts as a secondary outcome, but numerical results for that measure are not reported in the registry record. The non-inferiority margin, primary effect estimate and interval, detailed analysis plan, analysis-population definitions, and missing-data methods are not reported. The primary counts therefore cannot be interpreted as rates, and the non-inferiority classification cannot be fully evaluated from the posted statistical details.