Study of Immune Cell Changes in the Genital Tract 2 Months After Initiation of an IUD for Contraception: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of Immune Cell Changes in the Genital Tract 2 Months After Initiation of an IUD for Contraception (NCT01240811), a randomized na study. The primary outcome is %CD4 Expressing CCR5 HIV Co-receptor in the Cervix and Endometrium.

2. Clinical Question

The biological motivation in the study description is a question about local immune-cell populations and features relevant to HIV transmission. In particular, the hypothesis concerns CD4 cells and CCR5 receptor density in the upper and lower genital tract after placement of a progestin-containing IUD. The primary outcome recorded, however, is the percentage of T-cells expressing CCR5, measured in the cervix and endometrium. It should not be read as a direct measure of HIV acquisition, transmission, or overall immune function.

The registry record describes a pilot study and lists change in vaginal flora and quantities of selected vaginal bacteria as secondary outcomes. The posted primary outcome results are continuous measurements summarized by group means and standard deviations. The question addressed by the displayed comparisons is narrower than the study's broad biological motivation: it concerns the recorded immune-cell measure at the stated follow-up, not clinical HIV outcomes or every aspect of genital-tract immunity.

3. Trial Design

The registry describes a randomized, parallel study with no masking. It lists an enrollment of 42 and a status of completed. The recorded start was 2010-11, and primary completion was 2011-11. The study is identified as a pilot; the registry record does not report a sample-size justification or explain how the enrollment target was selected.

Three groups are described: Control-No IUD, Levonorgestrel IUS, and Copper T380A IUD. The control volunteers were described as healthy, not at risk of pregnancy, and not using hormonal contraception. Volunteers seeking contraception were randomized to receive either the levonorgestrel IUS or Copper T380A IUD. The record does not explain whether or how the control group was randomized, nor does it describe allocation concealment, the random sequence method, or the timing of allocation.

The primary endpoint was assessed at two months and concerns the percentage of T-cells expressing CCR5 in the cervix and endometrium. The group summaries list 8 participants in the control group, 16 in the levonorgestrel group, and 16 in the Copper T380A group. The primary outcome description says that all 40 participants who completed the study were included in the analysis. It also says that one participant withdrew and one was excluded after enrollment, and that both were replaced to fill the study's goal enrollment of 40 participants. The record does not describe the reasons for withdrawal or exclusion.

These details matter when interpreting the groups. The control group is described differently from the IUD groups with respect to contraceptive intent and eligibility, while the two IUD groups comprise volunteers seeking contraception. Random allocation between IUD types does not by itself establish that the control comparison was randomized. The registry record does not provide enough detail to resolve that point or to assess how allocation was implemented.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All 40 participants who completed the study were included in the analysis. There were no lost-to-follow up participants. One participant withdrew and one participant was excluded post-enrollment. Both of these participants were replaced to fill our goal enrollment of 40 participants in the 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 %CD4 Expressing CCR5 HIV Co-receptor in the Cervix and Endometrium, measured over 2 months.

The primary endpoint is the percentage of T-cells expressing the CCR5 HIV co-receptor in the cervix and endometrium. It is a continuous measure, reported as a mean with a standard deviation for each group. The registry describes assessment at two months after IUD insertion and identifies flow cytometry as the measurement approach. The outcome is intended to characterize a feature of local genital-tract immune cells, rather than a clinical HIV outcome.

The study description gives a biological rationale for measuring CCR5: the receptor is relevant to HIV tropism, and the investigators were interested in immune-cell features that might relate to HIV transmission. A cell marker can help characterize a biological environment, but a difference in a marker is not itself evidence of a difference in HIV acquisition or transmission. Those outcomes are not the primary endpoint reported in this record.

The reported group means are 4.9 for Control-No IUD, 11.7 for Levonorgestrel IUS, and 7.2 for Copper T380A IUD. The corresponding standard deviations are 17.2, 29.8, and 32.8. These summaries describe the observed measurements within each group. The standard deviations indicate substantial variation in the measurements relative to the group means, especially in the IUD groups; they are not standard errors and do not directly give the precision of a group mean or a between-group contrast.

The outcome description refers to change after insertion, but the posted summaries are group-level means and standard deviations at the stated time point. The registry record does not provide baseline values or a numerical change estimate. Consequently, the posted summaries should not be interpreted as a quantified within-person change from baseline.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a group mean describe?

For a continuous endpoint, a group mean is the arithmetic average of the observations included in that group. Here, the endpoint is the percentage of T-cells expressing CCR5. The reported means describe the central level of the measurements in each analyzed group. A mean does not show the full distribution: different patterns of individual observations can have the same average. It also does not by itself quantify how much two groups differ, because a contrast requires comparing the group estimates directly.

The standard deviation describes the spread of individual observations around their group mean. It is not the uncertainty of the mean and is not a confidence interval for a treatment contrast. A standard error would describe sampling uncertainty in an estimate, while a confidence interval would express a range of values compatible with the data and analysis assumptions. The registry record reports standard deviations but does not report confidence intervals for the group means or comparisons.

What does ANOVA test?

Analysis of variance, or ANOVA, is a general method for assessing whether group means differ, under the assumptions of the model used. In a multi-group setting, an overall ANOVA can test whether the data are compatible with equal means across the groups. Specific contrasts can then address particular pairs of groups. The registry labels the reported analyses as ANOVA and gives p-values for the control group compared with each IUD group. It does not describe the precise ANOVA model, the contrast procedure, or whether the two comparisons came from one overall model.

General statistical interpretation: a p-value is calculated under a null hypothesis and a set of assumptions. It describes how unusual a test statistic at least as extreme as the observed one would be if that null model were true. It does not measure the probability that the null is true, the probability that results will replicate, or the size of a biological effect. The reported p-values, 0.80 and 0.97, should therefore be read as test results, not as estimates of the size of the group differences.

Why does uncertainty matter?

An estimated contrast and its confidence interval would indicate the direction and scale of a comparison and show its statistical uncertainty. A narrow interval suggests greater precision under the model; a wide interval indicates that a broader range of effects remains compatible with the observations. Confidence intervals are not guarantees about the true value. Their interpretation depends on the model, sampling process, and analysis assumptions.

No contrast estimates or confidence intervals are reported for the primary comparisons. As a result, the registry results do not show the scale of the observed group differences with an uncertainty range. The group means and standard deviations provide useful descriptive context, but they are not substitutes for contrast estimates and intervals. In particular, a large p-value cannot establish that any possible difference is small enough to be unimportant.

Why can variability affect the analysis?

When measurements vary widely within groups, it can be harder to estimate group means and distinguish a difference between means from background variation. The standard deviations reported for this endpoint are large compared with the reported group means. This is a descriptive observation, not a new calculation of an effect or a test. It makes the absence of uncertainty estimates especially relevant: the standard deviations alone do not indicate the precision of the comparison.

General statistical explanation: some continuous outcomes have variability that is not well described by a model's simplest assumptions. Over-dispersion refers to more variation than the assumed model allows. If present and unaddressed, it can make standard errors and p-values unreliable. The registry record does not report an assessment of dispersion, model diagnostics, or any alternative variance method, so it is not possible to determine whether over-dispersion was assessed or relevant to this analysis.

What do the reported p-values establish?

The two reported p-values correspond to comparisons of each IUD group with the control group. They do not provide a direct comparison between the levonorgestrel and Copper T380A groups. Nor do they quantify the difference in CCR5 expression. The record does not report a prespecified hypothesis-testing plan, a significance threshold for the primary analysis, or whether the comparisons were considered confirmatory or exploratory. The study is identified as a pilot, but its statistical role is not described in detail.

General statistical explanation: failure to reject a null hypothesis is not the same as demonstrating equivalence or non-inferiority. Those conclusions require an appropriate design and a prespecified margin defining what differences would count as sufficiently small. The registry record does not report such a margin or an equivalence analysis. The p-values should therefore not be interpreted as showing that the groups have the same outcome.

How should planning illustrations be read?

Planning calculations based on an observed effect are hypothetical illustrations for a future study, not outcomes of the completed study. Their interpretation depends on assumptions about the effect to detect, significance level, allocation, and statistical approximation. The registry record describes a planning approach using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the target effect expressed as a fraction of the observed effect. These assumptions are not evidence that the original study was designed using that approach.

Such calculations must not be called observed power. Observed power is not a substitute for reporting the observed effect and its confidence interval. The registry record does not report the original study's sample-size justification, so the planning assumptions should not be attributed to the trial's design. Any future-study illustration is conditional on its stated assumptions and does not change the interpretation of the results already observed.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Control-No IUD", "Levonorgestrel IUS", "Copper T380A IUD")
n   <- c(8, 16, 16)
m   <- c(4.9, 11.7, 7.2)
sp  <- c(17.2, 29.8, 32.8)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Control-No IUD  8  4.900  17.200  -9.480  19.280
CARM  Levonorgestrel IUS  16  11.700  29.800  -4.179  27.579
CARM  Copper T380A IUD  16  7.200  32.800  -10.278  24.678

11. Analyses Reported by the Sponsor

The registry reports two sponsor analyses, both labeled ANOVA. For Control-No IUD compared with Levonorgestrel IUS, the reported p-value is 0.80. For Control-No IUD compared with Copper T380A IUD, the reported p-value is 0.97. The record does not give the estimated contrasts or confidence intervals associated with these tests, so the size and precision of either comparison cannot be read from the p-values alone.

No recalculated comparisons are listed in the record, and the discrepancy indicator is false. There is therefore no recalculated estimate to set against the sponsor's reported tests. In general, a simple unadjusted recalculation from summary statistics can differ from an analysis that uses participant-level data or model adjustments. When a discrepancy is flagged, the sponsor's own analysis takes precedence; that condition is not indicated for this study.

The sponsor's reported p-values remain limited to the comparisons and method described in the registry. They do not demonstrate equivalence, provide an effect size, or establish that any observed difference is clinically important. The registry record does not give further detail about the ANOVA specification, adjustments, or analysis plan, so the sponsor results should be interpreted within those reporting limits.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists three outcome measures and identifies two secondary outcomes: change in vaginal flora, and change in quantities of vaginal H2O2-producing Lactobacillus species and Gardnerella vaginalis. It also indicates five analyses, but the analysis details reported for the primary endpoint specify two comparisons: each IUD group against the no-IUD control group. The record does not describe the other analyses in enough detail to determine their hypotheses, outcomes, or results.

When several outcomes or comparisons are examined, the chance of obtaining at least one apparently unusual result can increase. Statistical methods for addressing multiplicity may include adjusting the testing procedure or identifying a primary comparison in advance. The registry record does not report whether any multiplicity adjustment was used, whether the reported contrasts were prespecified, or how the five analyses relate to the listed outcomes. The two p-values should therefore not be treated as a complete multiplicity-aware account of all analyses.

The primary outcome population description states that all 40 participants who completed the study were included in the analysis. It also records that one participant withdrew and one was excluded post-enrollment, and that both were replaced to fill the goal enrollment of 40. The record does not report an intention-to-treat analysis, explain the post-enrollment exclusion, or give outcome values for the withdrawn or excluded participants. It does not say whether missing values were imputed or whether sensitivity analyses were performed. The statement that there were no lost-to-follow-up participants does not resolve the separate withdrawal and exclusion details.

The arm summaries list 8 participants in the control group and 16 in each IUD group. The adverse-event section uses different at-risk counts for the IUD groups. The registry record does not explain this difference in denominators. Counts from separate outcome sections should not be assumed to represent identical analysis populations unless the record says so.

These reporting gaps do not establish that a particular analysis method was or was not used. They mean that the registry record does not provide enough information to assess the handling of multiplicity and missing data fully or to reproduce the analysis plan.

13. 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 1% reporting threshold.

The adverse-event record lists zero serious adverse events in the control group, with 8 at risk; zero in the levonorgestrel group, with 17 at risk; and zero in the Copper T380A group, with 17 at risk. The record also lists zero other adverse events in each group, using the same at-risk counts. These are the posted event counts and denominators for the adverse-event section.

The deaths field is not populated for these groups, so death counts are not reported in the registry record. No serious adverse-event terms are listed. The at-risk counts in this discussion differ from the primary endpoint group summaries for the IUD groups, and the record does not explain why. Safety denominators should therefore be read as reported rather than substituted for the primary outcome group sizes.

Zero recorded events in a study does not prove that an event cannot occur or that risk is absent. Safety conclusions depend on the number of people observed, the duration and completeness of follow-up, and how events were defined and collected. The registry record does not provide detailed safety definitions, follow-up duration for adverse events, or a fuller account of collection procedures. These counts describe what was recorded, not a guarantee of safety.

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

The posted results describe the distribution of the primary immune-cell measure in the three study groups and report ANOVA p-values for comparisons of each IUD group with the no-IUD control group. The means were 4.9 for the control group, 11.7 for the levonorgestrel group, and 7.2 for the Copper T380A group. The reported p-values were 0.80 and 0.97 for the respective comparisons. These are descriptive and inferential summaries of CCR5 expression as recorded at the stated time point.

The results do not establish that the groups are equivalent. There are no reported contrast estimates or confidence intervals with which to assess the magnitude and precision of the comparisons, and no equivalence margin is reported. The p-values alone cannot establish that the groups have the same outcome or that any possible difference is negligible.

The endpoint is a biological marker in genital-tract T-cells. It is not an HIV infection outcome, a direct measure of transmission, or a comprehensive measure of immune status. The study's rationale refers to HIV-related biology, but the posted primary results do not show whether the marker predicts HIV acquisition or transmission. They also do not establish what changes may occur in vaginal flora; those outcomes are listed separately.

The study description concerns change after IUD insertion, but the posted primary summaries do not provide baseline values or a numerical change from baseline. Therefore, the group means should not be interpreted as measured within-person changes. Nor do the registry results establish that any group difference was caused by the IUD, particularly given the distinct description of the control volunteers and the limited detail about allocation and analysis.

The findings are best read as pilot-study observations with substantial within-group variability and limited reporting of inferential uncertainty. They can inform questions for further research, but the registry record does not establish a clinical effect on HIV risk, a causal mechanism, equivalence between groups, or the absence of safety concerns.

15. Why This Trial Matters Statistically

This study is statistically informative because it illustrates the gap between a biological hypothesis and the evidence available from a small pilot study. The investigators were interested in immune-cell features relevant to HIV transmission, while the primary endpoint is a cellular marker measured in cervical and endometrial samples. Careful interpretation keeps those levels of evidence separate: a marker can motivate research without establishing a clinical outcome.

The group summaries make variability visible. The reported standard deviations are large relative to the group means, yet the registry gives no contrast estimates or confidence intervals. This highlights why a p-value alone is an incomplete result. A reader needs the scale of the estimated difference and its uncertainty to judge what effects remain compatible with the data. A large p-value cannot provide that information or demonstrate equivalence.

The record also shows why trial design details matter. Although the study is described as randomized and parallel, the control volunteers differ in their stated contraceptive context from volunteers entering the IUD groups. The record describes randomization of volunteers seeking contraception to the two device groups but does not explain the control-group allocation. These distinctions shape which comparisons can be interpreted as randomized and what conclusions are warranted.

Finally, the registry separates the primary immune endpoint from secondary vaginal-flora outcomes and adverse-event reporting. Differences in outcome populations and incomplete details about multiplicity, missing data, and analysis methods constrain the conclusions. A transparent account of estimands, contrasts, confidence intervals, and analysis decisions would make the statistical evidence easier to assess. In this record, the main lesson is to distinguish descriptive findings, hypothesis tests, and broader biological implications rather than treating them as interchangeable.

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.

The registry record does not report a sample-size justification, so the reason for the enrollment target and the precision anticipated for the primary endpoint cannot be assessed. The study is identified as a pilot, but the record does not explain whether its aim was feasibility, estimation, hypothesis testing, or a combination. Planning illustrations based on observed effects are not the study's original design rationale.

Reporting of the primary comparison is incomplete for evaluating effect magnitude. Group means and standard deviations are available, along with two p-values, but contrast estimates and confidence intervals are not reported. The specific ANOVA model, contrast procedure, model assumptions, and diagnostic checks are also not described. The record does not say whether a direct comparison between the two IUD groups was planned or performed.

The record describes volunteers in the control group differently from volunteers seeking contraception in the IUD groups. It does not report how the control group was allocated, whether baseline characteristics were balanced, or how any baseline differences were handled. It also does not give baseline endpoint values, despite describing the outcome as change after insertion. These omissions limit interpretation of the summaries as changes attributable to IUD placement.

The population description notes a withdrawal and a post-enrollment exclusion, with both participants replaced. Reasons, outcome availability, analysis handling, and sensitivity analyses are not reported in the registry record. The relationship between the primary endpoint group counts and the adverse-event denominators is also unexplained. Multiplicity handling for the listed outcomes and analyses is not reported.

These limitations concern what can be evaluated from the registry record; they do not show that a particular design or analysis choice was necessarily inappropriate. They do, however, restrict conclusions about precision, causation, generalizability, and the robustness of the reported comparisons.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Study of Immune Cell Changes in the Genital Tract 2 Months After Initiation of an IUD for Contraception — NCT01240811.
• Registry record last updated 2017-12-26; status last verified 2017-11.
• Results first posted 2013-07-22.
• 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

NCT01240811 is a completed randomized, parallel study with no masking, described as a pilot of immune-cell changes after IUD initiation. The primary endpoint is the percentage of T-cells expressing CCR5 in the cervix and endometrium at two months. Group means and standard deviations are reported for Control-No IUD, Levonorgestrel IUS, and Copper T380A IUD.

The registry reports ANOVA p-values of 0.80 for the control versus levonorgestrel comparison and 0.97 for the control versus Copper T380A comparison. It does not report contrast estimates or confidence intervals. The study description also lists vaginal-flora outcomes, while the adverse-event record reports zero serious and zero other adverse events in each group; deaths are not reported.

The results describe a biological marker, not HIV acquisition or transmission. Incomplete reporting of effect estimates, analysis details, allocation, multiplicity, and missing-data handling limits how strongly the results can be interpreted.