An Evaluation of a Multi-target Stool DNA (Mt-sDNA) Test, Cologuard, for CRC Screening in Individuals Aged 45-49 and at Average Risk for Development of Colorectal Cancer: Act Now: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for An Evaluation of a Multi-target Stool DNA (Mt-sDNA) Test, Cologuard, for CRC Screening in Individuals Aged 45-49 and at Average Risk for Development of Colorectal Cancer: Act Now (NCT03728348), a registered phase not stated study. The primary outcome is Specificity of the Multi-target Stool DNA Test in Average Risk Population, Ages 45-49.

2. Clinical Question

This question concerns agreement with a reference method among people whose colonoscopy was negative. It is not a comparison of two treatment strategies or an evaluation of outcomes after treatment. The study title and objective identify colorectal cancer screening and test specificity as the focus, while the primary outcome definition specifies the negative-colonoscopy population used to assess that performance. Because the registry record does not give the number of completed participants with negative colonoscopy, the denominator for the specificity estimate cannot be established from the record. Nor can the test's observed specificity or precision be described numerically. The stated average timeframe was 60 days through study completion; the record does not provide individual follow-up details beyond that description.

3. Trial Design

The registry describes an evaluation of a multi-target stool DNA screening test in average-risk individuals aged 45-49. Enrollment was 983. The listed interventions are the mt-sDNA screening test and colonoscopy. Colonoscopy is identified in the primary outcome description as the reference method, rather than as a treatment comparator. The record explicitly cautions that this is not a randomized comparison and that the listed arms are not a like-for-like control comparison. Allocation is not stated.

The primary objective was to confirm specificity in the target population. Participants' test results were interpreted using a score-based algorithm, with values of 183 or more classified as positive. Colonoscopic findings were used as the reference for determining whether the test result was a true negative in the specificity analysis. Lesions were to be confirmed as malignant by histopathologic examination, and test processing was independent of colonoscopic findings.

The registry lists the study status as completed. The listed start date is 2018-11-06, and the primary completion date is 2019-06-18. The primary outcome timeframe is described as through study completion, an average of 60 days. These record entries describe the study's registration and outcome timeframe; they do not provide participant-level timing or explain the schedule of procedures.

The record lists one primary outcome measure and no analyses. It does not report a numerical specificity result, a confidence interval, or the number of participants with negative colonoscopy. The enrollment figure should therefore not be treated as the denominator for the primary specificity calculation. The outcome description limits that population to completed participants with a negative colonoscopy, and the record does not give the count meeting that criterion. The statistical plan beyond the stated endpoint definition and algorithm is not reported in the registry 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.

Before assignment: 983 participants provided written informed consent.

Overall Study

Analysis population for the primary outcome, as defined in the registry: All completed participants that had a negative colonoscopy.

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: All evaluable participants except for participants who had mt-sDNA or colonoscopy excluded.

6. Primary Endpoint

The registry lists the primary outcome as Specificity of the Multi-target Stool DNA Test in Average Risk Population, Ages 45-49, measured over Through study completion, an average of 60 days.

The primary endpoint is specificity of the multi-target stool DNA test in average-risk individuals aged 45-49. Specificity measures the proportion of people who do not have the condition according to the reference method and who also receive a negative result from the test under evaluation. In this study's definition, the reference is a negative optical colonoscopy, and the test is classified as positive when its algorithmic score is 183 or more. The endpoint is therefore based on negative test results among participants with a negative colonoscopy.

Specificity is used to describe how often a test returns a negative result in the reference-negative group. It complements measures that focus on reference-positive participants, but does not by itself describe the full performance of a screening test or its effects on health outcomes. The endpoint's interpretation depends on how the reference status is established and on the population included in its denominator. Here, the record says that lesions would be confirmed as malignant by histopathologic examination and defines the analysis population as completed participants with a negative colonoscopy.

The registry identifies the endpoint as a percentage of participants and specifies a 95% confidence interval as its dispersion measure. However, it does not report the specificity percentage, the confidence interval limits, or the number of participants with a negative colonoscopy. Consequently, the endpoint definition can be described, but its observed value and precision cannot be assessed from the registry record. The stated timeframe is through study completion, an average of 60 days.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a binary classification summarized as a percentage: whether the stool DNA test result is negative among participants whose colonoscopy is negative. This is a diagnostic accuracy measure. In general, a binary outcome assigns each observation to one of two categories, such as positive or negative. Specificity uses the reference-negative group as its denominator and describes the fraction classified as negative by the index test. The registry supplies this definition directly; it does not report the observed fraction.

How is the test result classified?

The record describes a score generated using a logistic-regression algorithm. A score of 183 or more was treated as positive. In general, a classification threshold converts a continuous or ordered score into categories. The selected threshold determines which scores are called positive and which are called negative, and therefore forms part of the endpoint definition. The registry does not report the algorithm coefficients, the score distribution, or the number of test results on either side of the threshold. It is not possible to reconstruct the classifications from the registry information alone.

Why use colonoscopy as the reference?

For this endpoint, the negative colonoscopy defines the reference-negative group. The test's specificity is evaluated by determining whether the stool test is negative in that group. The record identifies optical colonoscopy as the reference and says lesions would be confirmed as malignant by histopathologic examination. A reference method is important because the calculation depends on the classification of participants as reference-positive or reference-negative. The registry does not give the count of negative colonoscopies or the detailed procedures used to resolve every possible finding.

What does the confidence interval add?

The registry specifies a 95% confidence interval as the measure of dispersion for the percentage endpoint. In general, an interval estimate communicates uncertainty around a point estimate. A narrower interval indicates greater statistical precision under the method used; a wider interval indicates more uncertainty. The interval is not a guarantee that an individual result will fall within its limits, nor does it establish clinical importance. Its calculation depends on the observed data and the statistical method. Since neither the endpoint estimate nor interval limits are reported, the precision of this study's estimate cannot be evaluated.

What can be inferred from the analysis description?

The outcome description says a logistic-regression algorithm generated test results. This describes the scoring process used to classify samples; it does not, by itself, establish that the study's reported specificity was estimated with a logistic-regression model. The record does not report the statistical method used to calculate the confidence interval, any adjustment, or any model used to analyze the endpoint. These elements are not reported in the registry record and should not be inferred from the algorithm description.

What is over-dispersion, in general?

For outcomes analyzed with a statistical model, over-dispersion means that variation in the data is greater than the model's assumptions expect. If it is ignored when present, standard errors can be too small and confidence intervals can be too narrow. This is a general statistical consideration, not a reported feature of this study. The registry record does not report a dispersion assessment or a model-based analysis of the endpoint, so there is no basis to say that over-dispersion was present or evaluated.

How should planning calculations be read?

The record describes sample-size planning illustrations based on the observed effect, using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The described effect to detect is a stated fraction of the observed effect. Such calculations are planning illustrations, not observed outcomes and not observed power. The record does not report numerical planning outputs in the registry record for this interpretation, and the assumptions do not establish that the study was designed or analyzed according to that scheme. In particular, the record does not report a sample-size justification or its own statistical plan.

In general, a planning calculation estimates the sample size that might be needed under specified assumptions. Its usefulness depends on whether those assumptions match the intended design, endpoint, and analysis. An illustration based on an observed effect is not equivalent to a prospective justification, and it does not show that the study had a particular probability of detecting an effect. The record gives no basis for claims about interim analyses, a stopping rule, or a prespecified statistical analysis beyond the primary endpoint description.

The reported enrollment of 983 is a registry enrollment figure, not a substitute for endpoint-specific information. The primary specificity calculation is defined for completed participants with a negative colonoscopy. Without the count in that group and the test classification counts, neither the percentage nor its uncertainty can be reproduced. The statistical methods can therefore be explained from the endpoint definition, but the results and their precision remain not reported in the registry record.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Multi-target DNA Test Results")
n      <- c(792)
events <- c(755)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Multi-target DNA Test Results  792  755  95.3  93.6  96.7  93.6  96.6  93.6  96.6

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists one primary outcome measure and no analyses. It does not provide a statistical analysis plan describing how outcomes were selected, how many comparisons were intended, or whether any adjustment for multiple testing was planned or used. The record also does not report interim analyses. These matters are not reported in the registry record; the absence of an analysis entry should not be interpreted as evidence that a particular procedure was or was not performed.

Multiplicity arises when several statistical hypotheses or outcomes are assessed. In general, repeated testing can increase the chance of obtaining at least one apparently notable result by chance, depending on the testing strategy. Adjustment methods can address that issue, but the appropriate approach depends on the planned analyses and their purpose. Because the record does not describe a broader set of tests or an adjustment strategy, no conclusion about multiplicity control can be drawn.

The primary endpoint's stated analysis population consists of all completed participants who had a negative colonoscopy. The record does not report the number meeting this definition, the number with missing test or colonoscopy information, or how incomplete observations were handled. A missing-data method, including any imputation or exclusion rules, is not reported in the registry record. It is therefore unclear from the registration whether missing observations affected the endpoint calculation.

The record does not describe a separate intention-to-treat or per-protocol population. Since this is not described as a randomized comparison, those labels should not be assumed to apply. Enrollment of 983 does not establish the primary outcome denominator, and the registry does not report how many participants completed the relevant procedures. These gaps limit assessment of analysis-population definitions, missingness, and the relationship between enrollment and the endpoint estimate.

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 0% reporting threshold.

The registry safety entry lists the group as Average Risk Patients. It reports 983 at risk for serious adverse events and 0 affected; for other adverse events, it reports 983 at risk and 0 affected. Deaths are reported as 0. The registry lists no top serious-event terms.

These figures are registry-reported counts, not a description of the types, timing, severity, or causes of events. The record does not explain further how the at-risk entries were defined or how safety information was collected. The figures should therefore be read as the counts recorded in the registry, not as evidence that no event of any kind occurred outside the reported categories or observation process.

The safety counts do not provide a comparison of harms between randomized treatment groups. The record states that this is not a randomized comparison and that the listed arms are not like-for-like control comparisons. The safety entry also does not supply additional details about exposure or participant-level follow-up. These limitations matter when interpreting zero affected entries: they describe what is recorded for the stated categories, but the registry information alone does not establish broader safety conclusions.

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

The registry record establishes the study's aim and the way its primary diagnostic endpoint was defined. It identifies an average-risk population aged 45-49, a multi-target stool DNA test, a score threshold of 183 or more for a positive result, and optical colonoscopy as the reference method. It defines specificity using completed participants with a negative colonoscopy and names a 95% confidence interval as the dispersion measure.

What the record does not establish is the observed specificity. It does not report the percentage, confidence interval limits, or endpoint denominator. Thus, it cannot show from the registry entry whether the test's specificity was high or low, how precise the estimate was, or whether a prespecified confirmation criterion was met. No such criterion is stated in the record. The enrollment figure of 983 does not resolve this issue, because the relevant endpoint group is defined by completion and negative colonoscopy rather than enrollment alone.

The study is an evaluation of a screening test against a reference method, not a randomized comparison of treatments. The listed intervention and procedure should not be interpreted as comparable treatment arms. The record notes that allocation is not stated and that the arms are not a like-for-like control comparison. Accordingly, the registry information does not support causal claims about treatment effects or health outcomes.

Specificity is one component of test performance. By itself, it does not give a complete account of the screening process, diagnostic accuracy across all reference categories, downstream procedures, or participant outcomes. The record does not report the other measures needed to make those assessments. It also does not provide enough information to assess missing observations, subgroup performance, or possible sources of bias.

The appropriate conclusion is limited: the registry describes a completed evaluation intended to confirm specificity in average-risk individuals aged 45-49, but the numerical primary result and its precision are not reported in the registry record. The design and endpoint can be interpreted, while the performance result itself cannot be judged from the registered information.

14. Why This Trial Matters Statistically

This study illustrates why a diagnostic endpoint must be read together with its reference method and analysis population. Specificity is not simply a property of a test in isolation: the calculation depends on who is classified as reference-negative and on which test results are considered negative. Here, the registry identifies colonoscopy as the reference method, specifies a threshold for a positive stool test, and defines the endpoint population as completed participants with a negative colonoscopy. Those elements clarify the intended estimand.

At the same time, the record demonstrates the difference between a registered objective and an interpretable result. The objective is to confirm specificity, and the registry names a 95% confidence interval, but it does not report the estimate, its interval, or its denominator. Without those items, readers cannot assess either the observed performance or its precision. An enrollment figure alone cannot replace endpoint-specific counts.

The record also makes an important design distinction: this is not a randomized comparison, and the listed arms are not a like-for-like control comparison. Colonoscopy serves as a reference procedure in the specificity definition. Treating it as a randomized control arm would confuse a diagnostic accuracy evaluation with an intervention comparison and could lead to unsupported causal interpretations.

Statistically, the study is a reminder to separate the test's score-generating algorithm from the analysis of its performance. The logistic-regression algorithm and cutoff describe how test results were classified. The registry does not provide the separate analysis details needed to reproduce the specificity estimate or its confidence interval. Clear registration of both the endpoint definition and the resulting counts would be necessary for an independent assessment of the result.

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.

The primary limitation for interpreting this trial from the registry record is that its central numerical outcome is not reported. The specificity estimate, its 95% confidence interval, and the number of completed participants with a negative colonoscopy are all absent. As a result, neither the magnitude nor the precision of the primary outcome can be assessed.

The outcome population is defined more narrowly than the enrollment entry. The registry reports enrollment of 983, but does not report how many participants completed the relevant procedures, had negative colonoscopy, or had usable test results. The relation between enrollment and the endpoint denominator is therefore not available. The handling of missing test results, incomplete procedures, or other exclusions is not reported in the registry record.

The record describes a logistic-regression algorithm and a threshold of 183 or more for a positive test, but does not provide the algorithm coefficients, score distributions, or classification counts. It also does not report the statistical method used to calculate the confidence interval or a full analysis plan. The endpoint description alone is not sufficient to reproduce the result.

Finally, this is not a randomized comparison, allocation is not stated, and the listed arms are not like-for-like control groups. The record therefore does not support treatment-effect interpretations. The registration also does not report subgroup estimates, additional diagnostic performance results, or details that would allow assessment of the potential impact of missing data or other sources of bias. These limitations are specific to the information available in the registry record and do not establish what may have been reported elsewhere.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: An Evaluation of a Multi-target Stool DNA (Mt-sDNA) Test, Cologuard, for CRC Screening in Individuals Aged 45-49 and at Average Risk for Development of Colorectal Cancer: Act Now — NCT03728348.
• Registry record last updated 2021-06-04; status last verified 2021-06.
• Results first posted 2021-06-04.
• 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

The registry record describes a completed evaluation of the multi-target stool DNA test Cologuard in average-risk individuals aged 45-49. Enrollment was 983. The study's primary objective was to confirm test specificity, using optical colonoscopy as the reference method. The test result was generated by a logistic-regression algorithm, with scores of 183 or more considered positive. The primary endpoint was defined for completed participants with a negative colonoscopy, with a 95% confidence interval specified as the dispersion measure.

The record lists the study start date as 2018-11-06 and primary completion date as 2019-06-18. It describes the outcome timeframe as through study completion, an average of 60 days. It also states that this was not a randomized comparison and that the listed arms are not like-for-like control groups. The numerical specificity result, confidence interval, endpoint denominator, and missing-data approach are not reported in the registry record. Safety entries report 0 serious adverse events, 0 other adverse events, and 0 deaths in the listed group.