Colon Capsule Endoscopy (CCE) Versus Computed Tomographic Colonography (CTC) in the Identification of Colonic Polyps in a Screening Population.: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Colon Capsule Endoscopy (CCE) Versus Computed Tomographic Colonography (CTC) in the Identification of Colonic Polyps in a Screening Population. (NCT02754661), a randomized na study. The primary outcome is Number of Participants With an Actionable Lesion on CCE vs. CTC Confirmed by Optical Colonoscopy.

2. Clinical Question

This is a diagnostic-yield question, not a comparison of clinical outcomes after treatment. Randomization assigned participants to CCE or CTC, while optical colonoscopy served as the confirmation reference for the primary outcome. The registry describes the trial as multicenter, prospective, randomized, parallel, and unmasked. The posted primary counts are 34 participants with an actionable lesion among 133 in the CCE group and 11 among 128 in the CTC group. Those counts describe the per-protocol outcome groups, not necessarily every randomized participant.

3. Trial Design

The registry describes NCT02754661 as a completed, multicenter, prospective, randomized study comparing diagnostic yield in a screening population. The parallel-group allocation placed participants in either colon capsule endoscopy or computed tomographic colonography. Masking is listed as none. The record lists an enrollment of 320 and gives the start as 2016-09 and primary completion as 2018-08-07. The phase is listed as not applicable.

The primary outcome was the number of participants with an actionable lesion on CCE or CTC confirmed by optical colonoscopy. An actionable lesion was defined as any polyp or mass lesion at least 6 mm. The assessment timeframe was 5-6 weeks from the randomized procedure. Diagnostic yield was to be calculated in relation to confirmatory optical colonoscopy results. The primary analysis population was a per-protocol set, excluding participants with major protocol deviations and specified other circumstances. The complete eligibility rules for that set are not reported in the registry record.

The registry also lists sensitivity, specificity, positive predictive value, and negative predictive value for detecting polyps at least 6 mm as secondary outcomes. Results are posted for sensitivity, specificity, and positive predictive value; negative predictive value results are not reported in the registry record. The sponsor is Medtronic - MITG. Although the registry gives the primary outcome and reports a Bayesian non-inferiority analysis, the trial’s full statistical plan, original sample-size justification, non-inferiority margin, interim-analysis plan, and detailed procedures for handling missing data are not reported in the registry record. Those omissions limit how fully the design and analysis can be reconstructed from the public record.

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

Analysis population for the primary outcome, as defined in the registry: Per protocol Analysis set (PPAS) is comprised of randomized subjects without major protocol deviations (violations with significant impact on subject outcomes) and who don't meet any of the following criteria: Subject withdraws, Capsules remained in the stomach/SB during the entire procedure, colonoscopy could not be done, technical failure

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: Baseline Analysis population is based on the Full Analysis set defined as all randomized subjects who underwent colonoscopy and CTC or CCE procedures.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With an Actionable Lesion on CCE vs. CTC Confirmed by Optical Colonoscopy, measured over 5-6 weeks from randomized procedure.

The primary endpoint counts participants whose CCE or CTC examination identified an actionable lesion that was confirmed by optical colonoscopy. The registry defines an actionable lesion as any polyp or mass lesion at least 6 mm, and sets the assessment timeframe at 5-6 weeks from the randomized procedure. The outcome is therefore a participant-level diagnostic yield: it records whether a participant had a confirmed lesion, rather than counting every lesion within a participant.

Using confirmatory optical colonoscopy connects the yield measure to a reference examination. This matters because a finding on the randomized test alone would not establish that an actionable lesion was present according to the endpoint definition. The per-protocol analysis set also shapes the population represented by the primary counts. It includes randomized participants without major protocol deviations and excludes specified circumstances, including withdrawal and a capsule remaining in the stomach or small bowel throughout the procedure. The registry’s complete criteria are not reported in the registry record.

The primary posted counts are 34 of 133 participants for CCE and 11 of 128 for CTC. These are descriptive outcome counts in the per-protocol groups. They do not, by themselves, answer every question about test performance, such as how often a negative examination correctly excludes a lesion. Sensitivity, specificity, and predictive values are separate secondary measures, and each uses a denominator defined by the relevant outcome classification.

7. Results Posted in the Registry

• Computed Tomographic Colonography: 8.6% (11/128) versus 25.6% (34/133) on control; risk difference -17.0 percentage points (95% CI -25.8 to -8.1), risk ratio 0.34 (0.18–0.63); unadjusted two-proportion 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.

Sensitivity of CCE Versus CTC in the Detection of Polyps ≥6 mm

• 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:
  • Computed Tomographic Colonography: 26.8% (11/41) versus 64.2% (34/53) on control; risk difference -37.3 percentage points (95% CI -56.0 to -18.6), risk ratio 0.42 (0.24–0.72); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Computed Tomographic Colonography than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Specificity of CCE Versus CTC in the Detection of Polyps ≥6 mm

• 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:
  • Computed Tomographic Colonography: 98.9% (86/87) versus 96.2% (77/80) on control; risk difference 2.6 percentage points (95% CI -2.1 to 7.3), risk ratio 1.03 (0.98–1.08); unadjusted two-proportion z-test p = 0.272.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Computed Tomographic Colonography and the control. The interval gives the range of effects that remain compatible with the data.

Predictive Value of a Positive Test (PPV) of CCE Versus CTC in the Detection of Polyps ≥6 mm

• 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:
  • Computed Tomographic Colonography: 91.7% (11/12) versus 91.9% (34/37) on control; risk difference -0.2 percentage points (95% CI -18.2 to 17.7), risk ratio 1.00 (0.82–1.21); unadjusted two-proportion z-test p = 0.980.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Computed Tomographic Colonography and the control. The interval gives the range of effects that remain compatible with the data.

The secondary results address complementary aspects of test performance rather than repeating the primary diagnostic-yield question. Collectively, the posted findings show that the relative pattern is not uniform across all measures: some contrasts differ in direction and precision, and the predictive-value analyses use smaller groups than the primary comparison. The negative predictive value is listed as a secondary outcome, but its result is not reported in the registry record.

These measures relate to the primary result because all concern detection of polyps at least 6 mm, but they are not equivalent endpoints. Sensitivity and specificity condition on different reference classifications, while predictive values condition on the test result. Their denominators and interpretation therefore differ from the primary participant-level count of confirmed actionable lesions. Agreement or divergence across these measures cannot be reduced to a single summary without considering those distinctions.

The secondary analyses are reported as non-inferiority analyses, including Farrington-Manning tests for the posted sensitivity, specificity, and positive predictive value results. The non-inferiority margin and full analytic details are not reported in the registry record. In addition, the registry does not report a multiplicity plan addressing the several secondary questions. When multiple outcomes are examined, isolated unadjusted results carry less weight because the chance of finding an apparently notable result across a collection of tests can increase. The secondary findings are best regarded as supportive and descriptive in relation to the primary outcome, not as independent confirmation of a single overall conclusion.

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

• Sensitivity of CCE Versus CTC in the Detection of Polyps ≥6 mm
• Specificity of CCE Versus CTC in the Detection of Polyps ≥6 mm
• Predictive Value of a Positive Test (PPV) of CCE Versus CTC in the Detection of Polyps ≥6 mm
• Predictive Value of a Negative (NPV) of CCE Versus CTC in the Detection of Polyps ≥6 mm

9. Statistical Interpretation

10. Statistical Methods Explained

What does the risk difference describe?

General explanation: a risk difference subtracts the event proportion in one group from the event proportion in another. In this comparison, the stated direction is CTC minus CCE. The primary estimate of -0.170 is therefore a contrast between the proportions with a confirmed actionable lesion in the two per-protocol groups. The posted group proportions are 0.086 for CTC and 0.256 for CCE. The risk difference keeps the result on an absolute scale: it describes the separation between the proportions, rather than their relative size.

The estimate is unadjusted, meaning it is calculated from the posted group counts without an adjustment for additional covariates. The registry does not report a covariate-adjusted primary estimate. The comparison should consequently be understood as a simple summary of the observed binary outcomes in the analysis groups, not as a model-based adjustment for participant characteristics.

How should the p-value be read?

General explanation: a p-value describes how unusual data at least as incompatible with a specified null hypothesis would be if that hypothesis and the model assumptions were true. It is not the probability that the null hypothesis is true, nor does it measure the size or importance of an effect. The recalculated primary comparison has an unadjusted p-value of <0.001. The calculation is not a substitute for the sponsor’s reported Bayesian non-inferiority analysis, which has a different reported framework and lacks key details in the registry record.

What does the confidence interval add?

General explanation: a confidence interval expresses uncertainty around an estimate under a specified method. Repeated use of the same method in comparable studies would yield intervals that cover the target value at the stated rate. For the primary re-calculation, the 95% interval runs from -0.258 to -0.081 on the CTC-minus-CCE risk-difference scale. Its negative values indicate that the estimated direction remains the same across the interval. The interval does not account for uncertainty about the analysis population, missing data, the endpoint’s operational details, or the model and assumptions not captured by the calculation.

How does non-inferiority differ from this re-calculation?

General explanation: a non-inferiority analysis assesses whether a result is compatible with a prespecified limit on how much one group may be worse than another for a chosen outcome. That limit, often called the margin, is central to interpretation and must be set before examining the results. The registry reports a Bayesian posterior probability of 1.000 for the primary non-inferiority analysis, but does not report the margin, model specification, or supporting interval. The registry’s secondary non-inferiority analyses report Farrington-Manning tests for some measures, but their interpretation also depends on the margin. Those details are not fully available for the primary analysis in the registry record.

Why are the secondary measures distinct?

General explanation: sensitivity concerns detection among participants classified as having the target condition, while specificity concerns classification among those without it. Positive and negative predictive values describe the probability of the condition given a positive or negative test, respectively, and depend on the population being evaluated. These measures use different denominators and answer different questions. Their results should not be collapsed into a single judgment about diagnostic performance. The registry lists four secondary measures; the negative predictive value result is not reported in the registry record.

General explanation: binary outcomes record whether an event occurred for each participant. The primary outcome and the listed secondary measures are participant counts or proportions, rather than continuous measurements. The record does not report the primary statistical method beyond its Bayesian non-inferiority label, so the assumptions and model details for that sponsor analysis cannot be evaluated from the registry entry alone.

11. Confidence Intervals and Precision

The primary re-calculated risk difference is -0.170, with a 95% confidence interval from -0.258 to -0.081. It compares the CTC proportion with the CCE proportion, in that order. The interval is entirely below zero, so the uncertainty represented by this calculation retains the direction of a lower observed yield in the CTC group. The interval also shows that the precise magnitude is not determined by the point estimate alone.

General explanation: a confidence interval is a range produced by a statistical method to represent uncertainty around an estimate. Its width reflects both the data and the method’s assumptions. It is not a range guaranteed to contain the true value in this particular trial, and it does not cover every source of uncertainty. In this case, the recalculated interval reflects uncertainty in the unadjusted comparison of the posted counts. It does not resolve uncertainty arising from the per-protocol selection, unreported missing-data handling, or absent details of the sponsor’s Bayesian model.

The sponsor’s primary non-inferiority analysis is reported as a posterior probability of 1.000, but no confidence interval or margin is given for that result. Its precision and substantive interpretation therefore cannot be assessed in the same way as the re-calculated risk-difference interval. The secondary estimates have their own intervals, some of which are broad, and they answer distinct questions with different analysis denominators. The secondary results should not be used to narrow the uncertainty in the primary estimate.

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 expressed as participants per arm, not total events. The planning calculation uses a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. It considers effects set at the stated fractions of the observed effect: 100%, 75%, and 50%. The original design assumptions used by the trial are not reported in the registry record; these planning assumptions describe the illustration, not the trial’s original design.

The 80% power column gives the estimated participants per arm for a hypothetical study designed to detect the specified effect with 80% power under those assumptions. The 90% power column gives the corresponding per-arm planning figure for 90% power. Power is the probability of detecting an effect of the specified size under the assumed model and test if that effect is present. These figures are not observed power.

The observed effect is only one possible basis for planning a future comparison, and using it directly can be unstable. A smaller target effect requires more participants to distinguish it from random variation at the same power and significance level. The figures for 75% and 50% of the observed effect therefore illustrate how planning needs can grow when a more modest difference is the target. They are not evidence that a future study will achieve those power levels if its actual effect, allocation, or analysis differs from the assumptions.

These calculations should be read as conditional planning scenarios rather than a retrospective judgment of whether the completed trial was adequately powered. The trial’s original sample-size justification is not reported in the registry record.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.256; treatment proportion p₁ = 0.086; average p̄ = 0.171.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5322 = 1.0431.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5185 = 0.4364.
4. n = (1.0431 + 0.4364)² / (-0.1697)² = 76.0, rounded up to 77 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
p0 <- 0.255639097744361                          # control proportion
p1 <- 0.0859375                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(2754661)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 77 (80% power), 102 (90% power)
 75% of observed effect: n per arm = 150 (80% power), 200 (90% power)
 50% of observed effect: n per arm = 365 (80% power), 488 (90% power)
power.prop.test: 76 
Simulated power at n = 77 per arm: 0.826 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary result as a Bayesian posterior probability of 1.000 in a non-inferiority analysis. It does not provide the non-inferiority margin, the model specification, a confidence interval, or further details needed to reconstruct that analysis. The posterior probability should therefore be reported as the registry states it, without assigning it a more specific interpretation than the record supports.

The re-calculated comparison uses the posted primary counts and denominators to estimate an unadjusted risk difference, with a 95% confidence interval and an unadjusted p-value. That calculation answers a direct comparison question on the risk-difference scale. It is not a recreation of the Bayesian analysis, and the two results are not interchangeable: they use different stated statistical frameworks, and the registry does not give enough detail to determine whether they use the same decision criterion.

The registry record does not flag a discrepancy between the posted primary counts and the reported analysis. The re-calculation is consequently a supplementary descriptive comparison, while the sponsor’s analysis remains the reported non-inferiority analysis. Neither should be extended beyond its stated outcome, analysis population, and available methodological details.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary endpoint and several secondary diagnostic measures. It reports five outcome measures and five analyses, but a plan for controlling the family-wise error rate or another multiplicity strategy is not reported in the registry record. Without that information, the unadjusted secondary p-values should be interpreted cautiously, particularly when several related outcomes are considered together. The primary comparison is more central to the stated study objective than individual secondary tests, but the registry does not describe a formal hierarchy or gatekeeping procedure.

The primary analysis population is a per-protocol analysis set. It consists of randomized participants without major protocol deviations and excludes participants who withdrew, whose capsules remained in the stomach or small bowel throughout the procedure, and other specified circumstances. The full set of exclusion criteria is not reported in the registry record. This population definition matters because per-protocol results describe participants meeting the analysis criteria, rather than automatically representing all randomized participants.

The registry does not report the number of participants excluded from the primary analysis for each reason, the handling of missing outcome information, or whether any imputation was used. It also does not report a complete participant flow that would allow the posted enrollment figure to be reconciled with each outcome denominator. The secondary analyses have different group denominators, which is consistent with the distinct classifications each measure requires; the detailed derivation of those groups is not reported in the registry record.

The registry reports a Bayesian non-inferiority analysis for the primary endpoint and Farrington-Manning non-inferiority tests for some secondary outcomes. The primary margin and full statistical plan are not reported. Interim analyses, if any, are not reported in the registry record. These gaps limit assessment of the prespecification, missing-data sensitivity, and overall error control.

16. 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: rectal haemorrage (0.7%), haematochezia (0.7%), post procedural haemorrahage (0.7%).

The registry reports serious adverse events in 2 participants among 145 at risk in the CCE group and 1 participant among 141 at risk in the CTC group. Other adverse events were reported for 4 participants among 145 at risk in the CCE group and 8 among 141 at risk in the CTC group. Deaths are recorded as 0 in both groups. These are event counts with the registry’s stated at-risk denominators.

The serious-event terms listed among the most frequent include Rectal Haemorrage, Haematochezia, and Post procedural haemorrahage, each reported as 0.7%. The registry’s terminology is retained as posted. These figures describe reported events; they do not establish that a procedure caused an event or that the groups differ in safety. The record does not provide enough detail to assess event severity beyond the serious-event classification, timing, relatedness, or completeness of adverse-event reporting. Counts should be interpreted in the context of the recorded denominators and the trial’s follow-up and reporting procedures, which are not fully described in the registry record.

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

The posted primary data show that confirmed actionable lesions were recorded for 34 of 133 participants in the CCE group and 11 of 128 in the CTC group. The unadjusted CTC-minus-CCE risk difference is -0.170, with a 95% confidence interval from -0.258 to -0.081. This describes diagnostic yield in the registry’s per-protocol analysis set and under its endpoint definition. It is not a result about later cancer outcomes, survival, or the consequences of choosing one examination over another.

The registry also reports a Bayesian posterior probability of 1.000 for a primary non-inferiority analysis. Because the non-inferiority margin and model details are not reported, that value cannot be fully interpreted as a decision against a defined threshold from the registry record alone. The direct re-calculation and the sponsor analysis answer different statistical formulations and should be read separately.

The results do not establish that either examination is suitable for every screening population, nor do they determine how unreported negative predictive value results would compare. The analysis is based on a per-protocol group, and the record does not fully document exclusions, missing-data handling, or the complete analysis plan. Secondary measures add context but involve different definitions and denominators, and multiplicity handling is not reported.

In general, a diagnostic-yield difference is evidence about how often a test identifies a finding under a particular study design and reference standard. It does not, by itself, establish clinical benefit, harm, or the best choice for an individual. The posted data should be interpreted as a registry-record comparison of diagnostic outcomes, with uncertainty and reporting limitations kept in view.

18. Why This Trial Matters Statistically

This trial illustrates why diagnostic comparisons need a clearly defined outcome and reference standard. The primary endpoint is not simply a finding on either examination; it counts an actionable lesion only when confirmed by optical colonoscopy. That distinction defines what the reported yield means and makes the confirmation process central to interpretation.

It also demonstrates how different statistical analyses can describe the same trial from different angles. The posted counts support an unadjusted risk-difference calculation, while the registry separately reports a Bayesian non-inferiority analysis. Without the primary non-inferiority margin and model details, the posterior probability cannot be fully connected to a decision rule. A numerical result should not be interpreted independently of its estimand and method.

The secondary outcomes further show that diagnostic performance is multidimensional. Sensitivity, specificity, and predictive values concern different conditional probabilities, so one primary yield comparison does not replace them. Missing results, multiplicity planning, analysis-population details, and the handling of incomplete data are also important to understanding how much confidence to place in a set of related findings.

For biostatisticians and clinical researchers, the record is a useful example of separating descriptive group results, re-calculated inference, and sponsor-reported non-inferiority analysis. It also shows why confidence intervals and protocol details matter alongside p-values and posterior probabilities. The conclusions that can be drawn are bounded by the outcomes, populations, and statistical details actually documented in the registry record.

19. 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 primary comparison is reported for a per-protocol analysis set, and the full criteria and participant flow are not reported in the registry record. The primary outcome denominators therefore do not, on their own, explain how all enrolled or randomized participants contributed to the analysis. The record also does not describe how missing outcomes were handled or whether sensitivity analyses assessed the effect of exclusions.

The sponsor’s primary non-inferiority analysis is reported as a posterior probability, but the margin, model specification, and supporting interval are not reported. This prevents a full appraisal of the hypothesis and decision threshold. The unadjusted risk-difference calculation provides a separate comparison from posted counts, but it cannot fill those gaps or account for unreported covariates and analysis choices.

The registry lists several secondary outcomes, while negative predictive value results are not reported in the registry record. A multiplicity strategy is also not reported, making individual unadjusted secondary findings less definitive when considered as a group. The record does not report interim analyses or the original sample-size justification.

The primary endpoint is diagnostic yield confirmed by optical colonoscopy, not a clinical outcome beyond detection. The registry record does not establish whether findings translate into differences in subsequent care or health outcomes. Safety reporting gives counts and denominators, but the available record does not fully describe timing, relatedness, or event-reporting procedures. These limitations are specific to the available trial record and constrain the scope of interpretation.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Colon Capsule Endoscopy (CCE) Versus Computed Tomographic Colonography (CTC) in the Identification of Colonic Polyps in a Screening Population. — NCT02754661.
• Registry record last updated 2019-10-01; status last verified 2019-08.
• Results first posted 2019-10-01.
• 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.

23. Record Summary

NCT02754661 is a completed, multicenter, prospective, randomized, parallel, unmasked comparison of colon capsule endoscopy and computed tomographic colonography in a screening population. The primary outcome was a participant with an actionable lesion at least 6 mm, confirmed by optical colonoscopy 5-6 weeks from the randomized procedure. The primary per-protocol counts were 34 of 133 for CCE and 11 of 128 for CTC.

The unadjusted CTC-minus-CCE risk difference is -0.170, with a 95% confidence interval from -0.258 to -0.081 and p < 0.001. The registry reports a Bayesian posterior probability of 1.000 for the primary non-inferiority analysis, but its margin and model details are not reported. Results for several secondary measures are posted; negative predictive value results are not reported in the registry record. Interpretation is limited by incomplete reporting of analysis-population details, missing-data handling, multiplicity, and the statistical plan.