Vitamin D/Calcium Polyp Prevention Study: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Vitamin D/Calcium Polyp Prevention Study (NCT00153816), a randomized phase 2, phase 3 study. The primary outcome is Colorectal Adenomas.

2. Clinical Question

3. Trial Design

This was the Vitamin D/Calcium Polyp Prevention Study, registered as NCT00153816. The registry lists the study as randomized, quadruple-masked, and factorial, with phases listed as PHASE2 and PHASE3. The recorded enrollment was 2813. Dartmouth-Hitchcock Medical Center is listed as sponsor. The study began in 2004-07, had primary completion in 2013-10, and is marked completed.

The full factorial design assigned participants to daily placebo, calcium carbonate, vitamin D3, or calcium carbonate plus vitamin D3. The listed daily amounts were 1200 mg calcium carbonate and 1000 IU vitamin D3. The four assignments allow the study to examine calcium and vitamin D within the same factorial structure, although the registry record does not report the specific statistical model used to estimate main effects or any interaction between the supplements.

The registry separately describes two arms for women choosing daily calcium: those women were randomized to daily placebo or daily vitamin D3 while taking 1200 mg calcium carbonate. These groups are not simply additional cells of the full factorial assignment; their description identifies a distinct choice and comparison. The record does not explain how the two-arm groups were combined with, or kept separate from, factorial analyses beyond the posted outcome groups and comparisons.

Masking is listed as quadruple, but the registry record does not identify which roles were masked. It also does not report the original sample-size justification, the planned allocation ratios, randomization procedures, interim analyses, or detailed analysis populations. These omissions limit reconstruction of the intended statistical plan from the registry entry alone.

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: 19,083 subjects were screened and 2,813 were enrolled into a run-in period of 56-84 days. 554 subjects were not randomized: 174 were ineligible due to out of range baseline lab values, 66 were ineligible for other reasons, 211 had \<80% tablet adherence, and 103 refused to participate.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Subjects are included if they had a follow-up exam at least one year after randomization and had sufficient histology available to ascertain the endpoint.

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.

Only the first 6 of 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Colorectal Adenomas, measured over 1 to 10 years.

The primary endpoint is colorectal adenomas, recorded as the percentage of subjects with the event. Its timeframe is 1 to 10 years. For this outcome, the registry says participants were included if they had a follow-up examination at least one year after randomization and sufficient histology available to ascertain the endpoint. The numbers contributing to the displayed arm results are smaller than the overall enrollment and differ by arm; the record does not give a complete account of why individual participants did not contribute.

An adenoma is a colorectal polyp category relevant to the study’s stated focus on colorectal neoplasia. A binary endpoint of this kind classifies each person according to whether the event was observed, then expresses the number with the event as a proportion of the outcome group. In general, this provides an interpretable summary of how often the endpoint occurred in each group over the specified observation period. It does not, by itself, describe when an adenoma was detected, how many adenomas a participant had, or the full course of follow-up.

The registry also posts advanced colorectal lesions as a secondary outcome, measured as a percentage of subjects over the same stated timeframe. The primary outcome description does not define adenoma subtypes or give additional endpoint-adjudication details. Those details are not reported in the registry record.

7. Results Posted in the Registry

• Full Factorial Calcium: 44.9% (171/381) versus 48.2% (183/380) on control; risk difference -3.3 percentage points (95% CI -10.4 to 3.8), risk ratio 0.93 (0.80–1.09); unadjusted two-proportion z-test p = 0.365.
• Full Factorial Vitamin D: 47.0% (179/381) versus 48.2% (183/380) on control; risk difference -1.2 percentage points (95% CI -8.3 to 5.9), risk ratio 0.98 (0.84–1.13); unadjusted two-proportion z-test p = 0.745.
• Full Factorial Calcium Plus Vitamin D: 45.7% (174/381) versus 48.2% (183/380) on control; risk difference -2.5 percentage points (95% CI -9.6 to 4.6), risk ratio 0.95 (0.82–1.10); unadjusted two-proportion z-test p = 0.492.
• Two Arm Placebo: 32.1% (88/274) versus 48.2% (183/380) on control; risk difference -16.0 percentage points (95% CI -23.5 to -8.6), risk ratio 0.67 (0.55–0.82); 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.

Advanced Colorectal Lesions

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Full Factorial Calcium: 11.4% (44/386) versus 9.2% (35/380) on control; risk difference 2.2 percentage points (95% CI -2.1 to 6.5), risk ratio 1.24 (0.81–1.88); unadjusted two-proportion z-test p = 0.319.
  • Full Factorial Vitamin D: 10.9% (42/384) versus 9.2% (35/380) on control; risk difference 1.7 percentage points (95% CI -2.5 to 6.0), risk ratio 1.19 (0.78–1.82); unadjusted two-proportion z-test p = 0.428.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Full Factorial Calcium and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Full Factorial Vitamin D and the control. The interval gives the range of effects that remain compatible with the data.

Advanced colorectal lesions are the posted secondary endpoint. Across the full factorial groups, the observed percentages are relatively close to one another, while the two groups described for women choosing daily calcium have lower reported percentages than the full factorial placebo group. The direction and degree of separation are not uniform across every comparison. This broad pattern does not reproduce a clear separation among the four factorial groups on the primary adenoma outcome, and it should not be treated as a confirmation of the primary result.

For the factorial comparisons with posted estimates, the risk-difference confidence intervals include zero and the unadjusted p values are 0.319 and 0.428. The registry lists no reported analyses for this secondary outcome. As a result, the arm percentages and intervals support descriptive comparison, but do not supply a full set of trial-reported inferential tests for every secondary contrast. The two-arm groups also have the distinct enrollment description noted for the primary outcome, which remains relevant when considering the overall pattern.

In general, secondary outcomes can provide context about whether findings are consistent across related endpoints, but they often carry less evidential weight than a prespecified primary endpoint. When multiple outcomes or contrasts are examined, the chance of observing an apparently unusual result somewhere among them can increase. A multiplicity plan can address that issue, but no adjustment strategy is reported in the registry record. Without a reported plan, secondary findings should be treated as exploratory or descriptive rather than as independent confirmation.

The advanced-lesion results and the adenoma results also measure distinct endpoint definitions. Related endpoints may overlap conceptually, but one should not assume that a pattern in one is interchangeable with a pattern in the other. The registry does not report a formal joint analysis of these outcomes, a hierarchical testing strategy, or a prespecified rule for interpreting the secondary result alongside the primary result.

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

• Advanced Colorectal Lesions

9. Statistical Interpretation

10. Statistical Methods Explained

The registry reports a binary percentage outcome and unadjusted comparisons, but it does not describe a complete inferential analysis plan. The explanations below give general statistical context for reading the posted quantities; they should not be mistaken for methods confirmed as part of the trial’s original protocol.

What does a percentage outcome represent?

For a binary outcome, each participant is classified as having or not having the specified event. The percentage is the share of the outcome group classified as having it. For the primary outcome, the event is colorectal adenoma, and inclusion requires a qualifying follow-up examination and enough histology to ascertain the endpoint. Thus, the denominator is the group with evaluable outcome data, not necessarily all randomized participants.

In general, an arm percentage is descriptive: it reports the observed event frequency in that group. A comparison then expresses how the observed frequencies differ under a chosen measure. The registry lists risk differences and also relative measures, but the arm percentages and confidence intervals are the most direct way to understand the observed outcome frequencies. A descriptive difference does not by itself establish a causal effect; interpretation also depends on randomization, the analysis population, missing outcomes, and the prespecified analysis approach.

What does a risk difference say?

A risk difference is the event proportion in one group minus the event proportion in its comparator. A value below zero indicates a lower observed proportion in the named arm relative to the comparator; a value above zero indicates a higher proportion. In general, a confidence interval for the risk difference that includes zero is compatible with no difference at the interval’s confidence level. If it excludes zero, the interval is inconsistent with zero under that procedure, but the result still requires attention to multiplicity, study design, and the analysis plan.

The registry labels the reported comparison p values as unadjusted. A p value summarizes how incompatible the observed data are with a specified null hypothesis under the test assumptions; it is not the probability that the null hypothesis is true, nor a measure of clinical importance. The record does not identify the exact test or model behind each posted p value. Consequently, they are best described as the registry’s unadjusted results, rather than attributed to a particular regression or test procedure.

How does the factorial design matter?

In general, a two-factor factorial trial can evaluate assignments for two interventions within one design. It may estimate the average contribution of each intervention across levels of the other intervention and can examine whether the joint effect differs from what would be expected under a specified model. The registry describes four full factorial groups, but it does not report whether its primary analyses used main effects, individual arm contrasts, an interaction term, or a particular strategy for handling interaction. The posted pairwise comparisons should therefore not be assumed to be a complete factorial analysis.

The trial also includes two groups described for women choosing daily calcium. Their assignment was placebo or vitamin D while calcium was taken daily. This creates a distinct comparison context from the four full factorial groups. The posted comparison reference is Full Factorial Placebo, including for the two-arm outcomes. The registry does not explain the rationale for that reference in the two-arm comparisons, so these results warrant careful reading and should not be silently re-labelled as a within-stratum comparison.

Why are confidence intervals useful?

In general, a confidence interval communicates uncertainty about an estimated quantity. A narrower interval indicates greater precision under the method used; a wider interval indicates that a broader range of values remains compatible with the data. The posted outcome intervals describe uncertainty around the arm percentages, while the comparison intervals describe uncertainty around risk differences. These are different quantities and should not be conflated.

Confidence intervals do not incorporate every possible source of uncertainty. They do not automatically account for selection caused by missing outcome assessments, errors in endpoint classification, deviations from the assigned intervention, or the fact that several outcomes and comparisons may be examined. The registry does not report enough analysis detail to determine how these issues were handled.

What is and is not reported about analysis?

The registry record does not report a detailed statistical analysis plan, the original sample-size justification, an interim-analysis plan, or a multiplicity adjustment. It does not report adjusted estimates or the analysis model used for the posted comparisons. The primary outcome inclusion description gives an evaluability criterion, but a complete account of all analysis populations and participant exclusions is not reported in the registry record.

For advanced colorectal lesions, the registry posts arm percentages and confidence intervals, but no reported analyses are listed for that outcome. The descriptive arm results can be compared at a broad level, but the absence of reported inferential analyses means no trial-specific significance conclusion for that secondary endpoint should be inferred from an independently chosen test. General statistical methods can explain how such analyses might be conducted, but the registry record does not establish that any particular method was used.

11. Confidence Intervals and Precision

The primary arm confidence intervals show appreciable uncertainty around the reported percentages. For the full factorial groups, the placebo interval is 43.0% to 53.3%, the calcium interval is 39.8% to 50.0%, the vitamin D interval is 41.9% to 52.1%, and the combined-assignment interval is 40.6% to 50.8%. The overlap indicates that the observed arm percentages are not sharply separated by these estimates. Overlap between separate arm intervals is not, by itself, a formal test of a difference; the registry’s comparison intervals and p values address the posted contrasts more directly.

For the full factorial contrasts, the comparison risk-difference intervals include zero. This leaves uncertainty about both the direction and size of any difference compatible with the data. The unadjusted p values range from 0.365 to 0.745 for those three comparisons. These results do not establish equivalence or prove that any underlying difference is absent. A claim of equivalence would require a specified margin and an analysis designed to evaluate it; such information is not reported in the registry record.

The two-arm outcome percentages have intervals of 26.6% to 38.0% and 26.8% to 38.5%. The posted risk-difference intervals for their contrasts against Full Factorial Placebo exclude zero, with p values below 0.001. That statistical separation is more distinct in the posted comparisons, but precision does not resolve the differences in the groups’ enrollment descriptions or the registry’s lack of explanation for how these comparisons were intended to be interpreted.

In general, confidence intervals describe sampling uncertainty under the analysis method and assumptions. They do not account for all uncertainty arising from endpoint ascertainment, missing outcome data, multiplicity, or unreported design choices. The registry does not report adjusted estimates or a detailed analysis plan, so the displayed precision should be read as precision for the posted summaries and comparisons, not as a complete accounting of every uncertainty in the study.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures describe hypothetical group sizes for future studies using the observed comparison effects as a starting point. They are not the enrollment targets reported for this trial. For these binary outcomes, the figures are participants per arm, not total participants and not total events. The registry’s planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each planning row uses the stated fraction of the observed effect as the effect to detect.

The 80% power column gives the participants per arm estimated to yield an 80% chance of detecting the specified effect under those assumptions if that effect is the true one. The 90% power column gives the corresponding per-arm planning figure for a 90% chance. Greater planned power generally requires more participants, all else held constant. The displayed rows consider planning for 100%, 75%, and 50% of the observed effect; the smaller the effect chosen for planning, the more participants are needed to detect it with the same power and assumptions.

For example, the calcium comparison lists 3,638 participants per arm for 80% power when planning for 100% of the observed effect, and 14,585 participants per arm when planning for 50% of that effect. The corresponding 90% figures are 4,870 and 19,524. These illustrations show how sensitive a sample-size calculation can be to the assumed effect size. They should not be read as evidence about how the original trial was designed: its own design assumptions and sample-size justification are not reported in the registry record.

These figures are planning illustrations, not observed power. Observed power is not a useful substitute for interpreting the estimated effects and their confidence intervals; the uncertainty intervals describe the precision of the reported results more directly.

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 = 0.482; treatment proportion p₁ = 0.449; average p̄ = 0.465.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7054 = 1.3825.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.7050 = 0.5934.
4. n = (1.3825 + 0.5934)² / (-0.0328)² = 3,637.8, rounded up to 3,638 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 <- 0.481578947368421                          # control proportion
p1 <- 0.448818897637795                          # 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(153816)
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 = 3,638 (80% power), 4,870 (90% power)
 75% of observed effect: n per arm = 6,476 (80% power), 8,668 (90% power)
 50% of observed effect: n per arm = 14,585 (80% power), 19,524 (90% power)
power.prop.test: 3637.8 
Simulated power at n = 3,638 per arm: 0.801 (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. Multiplicity, Missing Data and Analysis Populations

The registry reports a primary outcome, a secondary outcome, and several arm comparisons, but does not report a multiplicity strategy. Multiplicity arises when more than one hypothesis is tested: even if each test uses a conventional threshold, the chance of at least one false-positive finding across a set of tests can be greater than for a single test. In general, prespecifying a primary comparison, using an adjustment method, or defining a hierarchy can help clarify how multiple results should be interpreted. The registry record does not state whether any such approach was planned or used.

The posted p values are described as unadjusted. They should therefore be read as individual comparison results rather than as values adjusted for the number of outcomes, arms, or contrasts. This is particularly relevant when interpreting the secondary endpoint and the two-arm comparisons alongside the factorial results. The record does not report whether the listed comparisons were the complete set of planned tests or a subset selected for posting.

Missing outcome information also matters because the primary endpoint population includes participants with a follow-up examination at least one year after randomization and sufficient histology for endpoint ascertainment. The registry lists different outcome denominators by arm, but does not provide a full participant flow, the number without evaluable histology, reasons for missing assessments, or a comparison of included and excluded participants. It does not report imputation or sensitivity analyses for missing outcomes.

The record does not identify whether the primary result follows an intention-to-treat approach, a per-protocol approach, or another analysis population beyond the stated evaluability criterion. Nor does it describe how adherence, withdrawals, or deviations from assigned intervention were handled. These omissions do not invalidate the posted summaries, but they limit assessment of potential bias and reproducibility. The original analysis plan, interim analyses, and sample-size justification are not reported in the registry record.

15. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

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

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: joint replacement (5.0%), breast cancer/ductal carcinoma in situ (4.8%), prostate cancer (3.6%), endometrial cancer/hyperplasia (3.3%), fibroids (3.3%).

The registry reports serious adverse events and other adverse events by group as counts among participants at risk. For example, in Full Factorial Placebo, 84 participants were affected by serious adverse events among 415 at risk, and 46 were affected by other adverse events among 415 at risk. The corresponding counts for the other groups are also posted. These figures describe affected participants, not the number of separate events; a participant may have experienced more than one event, but the record does not clarify event-counting rules.

Terms appearing among the listed serious adverse events include joint replacement, breast cancer/ductal carcinoma in situ, prostate cancer, endometrial cancer/hyperplasia, fibroids, and back pain/degenerative spinal disorders. The registry gives these terms with percentages, but does not provide a causal assessment for each event. The death counts are not reported in the registry record.

In general, adverse-event counts should be interpreted with attention to group size, follow-up duration, event definitions, and whether events were judged related to an intervention. The registry summary does not report enough detail to compare safety formally or to attribute events to a study assignment. These counts are descriptive and should not be treated as a complete assessment of harms.

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

The posted results show that the observed primary adenoma percentages were similar across the four full factorial groups, with confidence intervals that overlap and comparison intervals that include zero. The unadjusted p values for those factorial contrasts do not indicate a clear separation in the posted data. This pattern is evidence about the estimates and their uncertainty; it is not proof that the assignments have identical effects or that no difference exists.

The two groups described for women choosing daily calcium have lower observed adenoma percentages than the Full Factorial Placebo group, and the registry’s unadjusted comparisons report p values below 0.001. However, the two-arm groups have a different enrollment description from the full factorial groups. Because the registry does not explain the analysis rationale for using the full factorial placebo group as the comparison reference for those groups, the contrast should be interpreted cautiously. It cannot be assumed to answer the same randomized question as comparisons within the factorial design.

The secondary advanced-lesion outcome provides additional descriptive context but has no reported analyses listed in the registry. Its pattern should not be treated as a separate confirmation of the primary endpoint, particularly given the absence of a reported multiplicity strategy and the distinct outcome definition.

These results do not establish that a treatment works or does not work, and they do not by themselves determine clinical relevance. They report event percentages and statistical uncertainty in the registry’s defined outcome groups. The registry record does not report detailed adjusted analyses, a complete missing-data assessment, or a protocol-level account of planned comparisons. Those limitations constrain how broadly the posted estimates can be interpreted.

In general, a statistically imprecise estimate can remain compatible with a range of effects, including no difference. Conversely, a small p value does not by itself establish that a difference is important or that the comparison is free from design and analysis concerns. Interpretation should combine the endpoint definition, group structure, magnitude and precision of estimates, and the limitations of the registry record.

17. Why This Trial Matters Statistically

This trial is statistically interesting because it combines a four-group factorial design with an additional two-arm comparison described for women choosing daily calcium. The factorial structure can support questions about calcium and vitamin D assignments, while the additional groups represent a distinct comparison context. Keeping those structures conceptually separate is important when reading the posted contrasts, especially because the registry lists Full Factorial Placebo as the control for all comparisons.

The results also illustrate why an estimate and its uncertainty should be considered together. The full factorial adenoma percentages are fairly close, and their confidence intervals and unadjusted comparisons do not provide a sharply resolved difference. The two-arm comparisons have a different posted pattern, but their distinct group description limits direct comparison with the factorial results. A single headline based only on the smallest p value would miss these design and interpretation details.

The endpoint definition highlights another important issue: outcome denominators depend on follow-up examination and sufficient histology. The total enrollment and the group sizes contributing to the endpoint are not the same. Without a complete account of missing assessments and analysis populations, readers cannot fully assess how evaluability may affect generalizability or introduce selection concerns.

Finally, the registry reports both primary and secondary outcomes but does not report a multiplicity plan or detailed analysis strategy. This makes the distinction between confirmatory evidence and descriptive context especially important. The trial record demonstrates how design structure, endpoint ascertainment, uncertainty, and reporting completeness all shape the statistical meaning of a result.

18. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

Several trial-specific limitations follow from what is not reported in the registry. The primary endpoint is restricted to participants with a follow-up examination at least one year after randomization and sufficient histology, but the record does not give a complete flow of participants through follow-up or explain the differing outcome denominators. It is therefore not possible from this record alone to assess whether participants with evaluable outcomes differ from those without them.

The study includes both full factorial groups and two arms described for women choosing calcium. The registry does not explain how these groups were intended to contribute to the same or separate estimands, nor why the Full Factorial Placebo group is the stated control for the two-arm comparisons. This complicates direct interpretation of those comparisons relative to the factorial contrasts.

The registry does not report the original sample-size justification, the detailed statistical analysis plan, interim analyses, an interaction analysis, multiplicity adjustment, or adjusted estimates. It also does not specify the exact tests or models behind the unadjusted p values. These omissions make it difficult to assess whether the posted analyses match the planned analyses or whether all intended contrasts are represented.

Endpoint ascertainment depends on follow-up and histology, while additional details about adjudication and missing-data handling are not reported. The advanced colorectal lesion endpoint has posted arm summaries but no reported analyses. Safety figures are summarized as affected participants, without enough information about event attribution or counting rules to support comparative causal conclusions. These limitations are specific to interpreting the registry record and do not establish that the underlying study lacked such procedures.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Vitamin D/Calcium Polyp Prevention Study — NCT00153816.
• Registry record last updated 2017-03-15; status last verified 2017-02.
• Results first posted 2015-10-20.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

22. Record Summary

NCT00153816 is the completed Vitamin D/Calcium Polyp Prevention Study, sponsored by Dartmouth-Hitchcock Medical Center. The registry lists randomized allocation, quadruple masking, a factorial model, PHASE2 and PHASE3, and enrollment of 2813. The recorded study start was 2004-07 and primary completion was 2013-10.

The primary outcome is colorectal adenomas, reported as a percentage of subjects over 1 to 10 years among those with a qualifying follow-up examination and sufficient histology. Advanced colorectal lesions are listed as a secondary outcome. The registry posts outcome percentages, confidence intervals, unadjusted comparisons, and adverse-event counts. It does not report a detailed statistical analysis plan, original sample-size justification, multiplicity approach, or complete missing-outcome account.