Stool DNA to Improve Colorectal Cancer Screening Among Alaska Native People: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Stool DNA to Improve Colorectal Cancer Screening Among Alaska Native People (NCT04336397), a randomized na study. The primary outcome is Colorectal Cancer Screening by Study Arm.

2. Clinical Question

The trial addresses implementation of screening in a particular care setting, rather than the diagnostic accuracy of MT-sDNA. The registry describes uncertainty about the feasibility and acceptability of this test in the Alaska tribal health care delivery system and whether its use would increase screening rates. The interventions therefore combine education and outreach with access to a test kit when MT-sDNA is selected. The study also includes colonoscopy as a screening option, so the primary outcome captures qualifying screening by either route rather than use of MT-sDNA alone.

The registry identifies two secondary outcomes: whether a test was requested, either MT-sDNA or colonoscopy, and MT-sDNA diagnostic follow-up. These outcomes relate to steps in the screening pathway, while the primary endpoint is a completed qualifying screening episode under its stated definition. The registry does not report the outcome counts or detailed results for those secondary measures.

3. Trial Design

This was a randomized, parallel-group study listed as completed. The registry records enrollment of 2029 and identifies the sponsor as the Alaska Native Tribal Health Consortium. The recorded start date is 2021-04-29 and the primary completion date is 2024-11-15. The study is classified as single-masked, but the identity of the masked participants or personnel is not reported in the registry record. The phase is listed as not applicable.

The three study groups were high-intensity outreach, medium-intensity outreach, and usual care. The two outreach strategies differed in the described amount of contact from a tribal health worker and in whether educational material was mailed. In both outreach groups, participants who selected MT-sDNA were sent a kit and follow-up reminders. Usual care was described as an opportunistic screening recommendation at a clinic visit. The registry does not report further details about intervention delivery, adherence, or variation among participating health organizations.

The analysis notes identify the design as cluster-randomized and caution that a simple individual-level re-analysis would be misleading. In a cluster-randomized study, assignment occurs at the level of groups or clusters rather than treating every participant as an independent randomization unit. Outcomes among people in the same cluster can be related. General statistical principle: an analysis should account for the assignment structure and potential within-cluster similarity, because treating correlated observations as independent can misstate uncertainty. The registry record does not report the cluster definition, the number of clusters, or the statistical details used to account for clustering.

The primary endpoint was assessed over 1 year after randomization and intervention. The reported statistical analyses use chi-squared methods and are labelled superiority analyses. One reports an odds ratio for a two-group comparison; another reports a p-value for a comparison across the three groups. The registry record does not describe the randomization procedure, allocation concealment, sample-size justification, or a full statistical analysis plan. It also does not identify which personnel were masked or how masking was maintained.

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

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 Colorectal Cancer Screening by Study Arm, measured over 1 year.

The primary endpoint is the count of participants with a qualifying colorectal cancer screening episode within 1 year after randomization and intervention. The registry defines an episode as at least one of the following: colonoscopy; MT-sDNA with a negative result; or MT-sDNA with a positive result followed by colonoscopy within 1 year. Thus, a positive stool test alone does not meet the stated definition unless followed by colonoscopy within the specified period.

This endpoint measures completion of screening according to the study’s pathway definition, not detection of cancer, diagnostic accuracy, or a health outcome. It includes more than one screening route, reflecting that the outreach materials described available screening options and that participants could select MT-sDNA or colonoscopy. The secondary outcome “Test Requested” records whether MT-sDNA or colonoscopy was requested, whereas the primary endpoint requires a qualifying screening episode. The separate secondary outcome of MT-sDNA diagnostic follow-up concerns a later step in the pathway.

General statistical rationale: an endpoint that records whether a participant completed a defined screening episode can assess uptake of screening after an implementation strategy. A specified follow-up window makes clear which screening activity counts toward the outcome. Interpretation still depends on complete and consistent outcome ascertainment and on the analysis respecting the cluster-randomized design. Details of ascertainment, handling of incomplete follow-up, and the analysis population are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry lists chi-squared methods for both reported superiority analyses. A chi-squared test is a general method for assessing whether categorical outcomes differ across groups. Here, the primary endpoint is a participant count for a binary event: a qualifying screening episode either occurred within the specified follow-up period or it did not. General statistical explanation: a test of this kind evaluates how compatible the observed group pattern is with a null hypothesis of no group difference, subject to the test’s assumptions and the study design. The registry does not provide enough detail to specify the exact implementation or all assumptions used.

What does the odds ratio describe?

The two-group analysis reports an odds ratio of 1.2. Odds compare the chance of an event with the chance of no event; an odds ratio compares those odds between groups. It is not the same as a percentage-point difference, and it does not directly state how many additional participants experienced screening. The group order and event coding determine which group’s odds are in the numerator and how the direction should be read. Those details are not reported in the registry record, so the estimate cannot be confidently attributed to a named arm or interpreted as a direction of benefit from the posted fields alone.

The reported 95% confidence interval is 1.0 to 1.5. General statistical explanation: a confidence interval conveys the range of effect estimates that are reasonably compatible with the observed data under the analysis procedure, while also reflecting sampling uncertainty. It is not a guarantee that the true value lies within the interval, nor does it account for every possible source of bias or design limitation. The interval’s lower endpoint and the p-value are both reported as rounded values. The registry does not state whether the interval and test were calculated using identical methods, so no more detailed reconciliation is possible from the record.

How should the p-values be read?

The two-group analysis reports p = 0.04, and the three-group analysis reports p < 0.01. General statistical explanation: a p-value measures how unusual the observed data, or data more extreme, would be under a specified null hypothesis and the assumptions of the analysis. It is not the probability that the null hypothesis is true, the probability that chance alone caused the result, or a measure of effect size. A p-value also does not supply the missing arm-specific screening rates. The three-group result is an overall comparison; without further contrasts, it does not identify which groups differ.

The analyses are labelled superiority. This label indicates that the analyses address whether a group difference is present in the direction defined by the analysis, rather than testing equivalence or non-inferiority. The registry does not provide a formal hypothesis statement, a margin, or a detailed testing plan. The direction and clinical interpretation of the pairwise odds ratio remain constrained by the unreported group ordering and event coding.

Why does clustering matter?

The analysis notes identify the trial as cluster-randomized and explicitly state that a simple individual-level re-analysis would be misleading. General statistical explanation: individuals in the same cluster may share features or care processes, making their outcomes more alike than outcomes from unrelated participants. Analyses that ignore such dependence can understate or otherwise misrepresent uncertainty. Suitable methods account for the assignment unit and the correlation structure, but the specific method, cluster definition, cluster count, and any assumptions about within-cluster variation are not reported in the registry record. For that reason, no independent participant-level recalculation is presented.

What information is unavailable?

The posted analysis does not include the primary endpoint counts by arm, arm-specific event rates, or the group labels attached to the two-group odds ratio. It does not report the exact analysis population, outcome-missingness rules, model details, or adjustments beyond identifying chi-squared methods. The registry record also does not report whether the tests were adjusted for multiple comparisons. These omissions limit the ability to reproduce or extend the sponsor-reported analyses. The results should therefore be read as the reported inferential findings, not as a complete statistical account of the trial.

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("High Intensity", "Medium Intensity", "Usual Care")
n      <- c(746, 696, 587)
events <- c(205, 165, 32)
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  High Intensity  746  205  27.5  24.3  30.8  24.4  30.8  24.4  30.8
ARM  Medium Intensity  696  165  23.7  20.6  27.0  20.7  27.0  20.7  27.0
ARM  Usual Care  587  32  5.5  3.8  7.6  3.9  7.6  3.9  7.6

11. Analyses Reported by the Sponsor

The available inferential results are the sponsor-reported analyses: an odds ratio of 1.2 with a 95% confidence interval of 1.0 to 1.5 and p = 0.04 for a two-group comparison, and p < 0.01 for a three-group comparison. The registry notes that this was a cluster-randomized design and that a simple individual-level re-analysis would be misleading. It therefore presents only the sponsor’s own analysis rather than an unadjusted recalculation from summary statistics.

The posted information does not map the two-group comparison to named arms, state the event coding, or provide group-specific endpoint counts. It is not possible from the registry record to reproduce the odds ratio or explain the detailed adjustment for clustering. The three-group p-value is also not accompanied by group-specific estimates or follow-up contrasts. The sponsor’s tests are consequently the relevant reported inferential results, while their interpretation remains limited by the details not recorded.

The analysis notes do not report a discrepancy between a sponsor estimate and a recalculation. No recalculated effect is available to compare with the sponsor’s results. These findings should be described as the registry’s posted analyses, not as an independent confirmation of the numerical estimates.

12. Multiplicity, Missing Data and Analysis Populations

The registry reports two primary-outcome analyses: a two-group odds-ratio analysis and an overall comparison involving three groups. It does not report whether these tests were prespecified as a hierarchy, whether the two-group analysis was the principal contrast, or whether any adjustment for multiple comparisons was planned or applied. The relationship between the pairwise and three-group analyses is therefore not fully described. Without that information, the reported p-values should be treated as the individual results shown in the registry, rather than as evidence that a particular family-wise error rate was controlled.

Missing outcome data and the handling of participants without an observed qualifying screening episode are not described. The record does not state whether missing observations were treated as no screening, excluded, imputed, or handled by another method. The analysis population is also not reported: the registry does not specify whether the results use all randomized participants, participants with available follow-up, or another defined population. These details matter because the primary endpoint depends on recording screening and any required follow-up within the stated time window.

The record does not report participant flow by arm, withdrawals, or the amount and reasons for incomplete follow-up. It also does not provide the screening counts by arm needed to inspect the outcome distribution. Thus the degree to which missing data could affect the results cannot be assessed from the posted record.

The study is identified as cluster-randomized, but the registry does not report the cluster definition, cluster numbers, or how clusters and participants were handled in the analysis. The analysis notes caution against an individual-level re-analysis. This is important because the analysis unit and dependence among participants affect both uncertainty and interpretation. The posted data do not support a reconstruction of those design details or an independent adjustment. A detailed multiplicity plan, missing-data strategy, and analysis-population definition are not reported in the registry record.

13. 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 lists serious adverse events as 0 for the high-intensity group, 0 for the medium-intensity group, and 0 for usual care. Deaths are also listed as 0 in each group. The numbers at risk for serious adverse events are recorded as 746, 696, and 587, respectively; the numbers at risk for other adverse events are the same for the corresponding groups. Other adverse events are listed as 0 in each group, and no serious adverse event terms are listed.

These entries describe the adverse-event figures recorded in the registry. They should not be read as proof that no adverse events occurred outside the registry’s reporting definitions or observation procedures. The record does not explain the collection period, event definitions, ascertainment process, or whether the figures include events regardless of relationship to study procedures. It also does not provide detail on how adverse events were solicited or verified.

Zero recorded events in the listed categories do not establish that the study interventions or screening options are free of risk. They indicate only that the registry reports zero events in these categories for the groups and denominators shown. The available figures do not permit a more detailed comparison of safety between groups or an assessment of rare events.

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

The posted results concern whether participants completed a qualifying colorectal cancer screening episode within 1 year after randomization and intervention. The registry reports a statistically significant p-value for a two-group comparison and an overall three-group p-value below 0.01. These results indicate evidence of group differences under the reported analyses. The registry does not provide enough information to identify the direction and size of each arm-specific difference or to determine which pairs account for the overall result.

The primary endpoint is a screening-process measure. It combines colonoscopy with specific MT-sDNA pathways and requires follow-up colonoscopy after a positive MT-sDNA result. It is not an outcome measuring colorectal cancer incidence, mortality, diagnostic accuracy, or long-term health. The registry summary describes a long-term study goal of improving screening and reducing colorectal cancer-attributable mortality, but the reported primary endpoint does not establish an effect on mortality.

The result does not show from the posted fields alone whether high-intensity outreach, medium-intensity outreach, or usual care had the highest screening rate. The two-group odds ratio lacks named group mapping and event coding in the reported analysis, while the three-group result is an overall test without reported group-specific comparisons. Nor does the p-value provide an estimate of the number of people whose screening behavior changed.

Interpretation is also limited by missing details about clustering, outcome counts, missing data, the analysis population, and multiplicity. The registry explicitly warns against simple individual-level recalculation for this cluster-randomized trial. Accordingly, the reported tests can be described, but they should not be expanded into an unsupported ranking of the strategies or a claim about effects beyond the defined screening endpoint.

General statistical principle: evidence for a difference and the size, precision, and practical importance of that difference are separate questions. The reported confidence interval supplies some information about uncertainty for the two-group odds ratio, but absent arm-specific rates and group mapping, the result cannot be translated into an absolute screening difference. The findings are specific to the study setting and the strategies as described in the registry; broader applicability is not established by the posted results alone.

15. Why This Trial Matters Statistically

This trial is statistically notable because it studies screening uptake in a care setting described by the registry as not previously evaluated for the new at-home multi-target stool DNA approach. The intervention is not simply a test comparison: it combines outreach intensity, culturally appropriate educational material, telephone contact, and access to MT-sDNA when selected. The primary endpoint captures a completed screening episode through more than one route, matching the implementation question of whether outreach is associated with screening completion rather than only test preference.

The three-arm structure allows comparison of two outreach intensities and usual care in the registry’s overall analysis. However, the posted results do not include arm-specific screening rates or identify the named groups in the two-group odds-ratio analysis. The overall p-value indicates evidence of differences among groups under the reported test, but does not resolve which strategy differs or by how much. This distinction between an overall test and interpretable pairwise effects is important when reading multi-arm trials.

The cluster-randomized design is another key statistical feature. Outcomes from participants sharing a cluster may be related, so analysis needs to reflect the assignment structure. The registry explicitly warns that simple individual-level recalculation would be misleading and reports only the sponsor’s analysis. This makes clear why an apparently straightforward outcome count cannot be independently reanalyzed without the design and analysis details.

Finally, the registry’s endpoint definition emphasizes completion and follow-up. A request for a test is a secondary outcome, while the primary measure requires a qualifying screening episode under the stated rules. That distinction matters when considering implementation: interest, test request, completion, and follow-up are separate stages. The posted primary results address the defined screening episode, not every step in the pathway or longer-term clinical outcomes.

16. 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 trial-specific details needed for a fuller interpretation are not reported in the registry record. Although the analysis notes identify cluster randomization, the record does not define the clusters, report their number, or explain how the analysis accounted for clustering. It also does not provide the randomization procedure or allocation concealment details. The masking designation is single, but who was masked is not stated.

The primary endpoint’s definition is specific, but the posted record does not give the number of participants meeting it in each arm, the arm-specific screening proportions, or the handling of missing screening information. The analysis population and participant flow are not reported. As a result, it is not possible to assess how incomplete follow-up was treated or to reconstruct the primary outcome analysis from the available summary.

The two-group odds ratio is not linked to named arms in the analysis information, and event coding is not specified. The reported 95% confidence interval has a lower endpoint of 1.0 alongside p = 0.04; the record does not explain whether rounding or differing calculation details account for this presentation. The three-group p-value is not accompanied by an effect estimate or pairwise follow-up tests. Details of multiplicity handling are not reported.

The record also does not report the sample-size justification, interim analyses, or a full statistical analysis plan. Intervention fidelity, participant adherence, and variation across participating Tribal health organizations are not detailed. The secondary outcomes are named, but their results are not included in the registry record recorded for this analysis. These omissions do not negate the posted findings, but they constrain reproducibility, assessment of precision by arm, and interpretation of how the interventions were implemented.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Stool DNA to Improve Colorectal Cancer Screening Among Alaska Native People — NCT04336397.
• Registry record last updated 2026-09-10; status last verified 2025-11.
• Results first posted 2025-11-19.
• 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.

20. Record Summary

The registry record describes a completed, randomized, parallel-group, single-masked trial with enrollment of 2029. It compares high-intensity and medium-intensity tribal health worker outreach with usual care for colorectal cancer screening among Alaska Native people. The sponsor is the Alaska Native Tribal Health Consortium.

The primary endpoint is a qualifying screening episode within 1 year after randomization and intervention, defined through colonoscopy or specified MT-sDNA pathways. The posted analyses report an odds ratio of 1.2 with a 95% confidence interval of 1.0 to 1.5 and p = 0.04 for a two-group comparison, plus p < 0.01 for an overall three-group comparison. Arm-specific outcome counts and the named groups in the two-group result are not reported in the registry record.

The analysis notes identify a cluster-randomized design and state that a simple individual-level recalculation would be misleading. The record does not report the cluster details, analysis population, missing-data approach, or multiplicity plan. Serious adverse events, other adverse events, and deaths are listed as 0 in each group.