Treatment of Gardnerella Vaginalis Vaginal Colonization With Amoxicillin: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Treatment of Gardnerella Vaginalis Vaginal Colonization With Amoxicillin (NCT03211156), a randomized phase 2 study. The primary outcome is The Proportion of Participants With Eradication of Gardnerella Vaginalis (GV) in Each Study Arm at Visit 2 (Day 15-21) as Assessed by Nucleic Acid Amplification Test (NAAT)..

2. Clinical Question

3. Trial Design

This was a randomized, double-blind, parallel-group, placebo-controlled Phase 2 study of amoxicillin for vaginal colonization or infection with GV. The record lists enrollment as 97 and describes two study arms. The arm descriptions list n=49 for placebo and n=49 for amoxicillin; for the primary outcome, the posted denominators are 49 and 48, respectively. The record does not explain the difference between the arm-description count and the reported outcome denominator.

The planned regimen was 2 x 250 mg amoxicillin capsules by mouth twice daily for 7 days, compared with placebo capsules twice a day for 7 days. The primary assessment was at Visit 2, Day 15 to 21, when a test of cure was performed using NAAT. The primary endpoint counted a negative GV NAAT as eradication. Participants with a missing test result at the test-of-cure visit were classified as treatment failures.

The trial summary describes screening approximately 245 healthy adult females to enroll approximately 98 participants and achieve 82 evaluable participants at the test-of-cure visit. These are planning figures in the study description, not the observed enrollment or outcome denominators. The summary also describes study activity at two clinics in the United States: University of Alabama at Birmingham and Wake Forest University Health Sciences. The registry lists the study as completed, with start on 2017-09-11 and primary completion on 2019-08-22.

The record does not report the randomization sequence method, allocation concealment procedures, details of blinding implementation, or a statistical analysis plan. The specified design features establish the broad comparison, but they do not supply those operational or analytical details.

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: Participants included in the ITT population are those who randomized, regardless of whether they received study treatment or were compliant with the administration procedures or schedule.

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 randomized subjects, receiving or not receiving the study product, are included in the baseline analysis population.

6. Primary Endpoint

The registry lists the primary outcome as The Proportion of Participants With Eradication of Gardnerella Vaginalis (GV) in Each Study Arm at Visit 2 (Day 15-21) as Assessed by Nucleic Acid Amplification Test (NAAT)., measured over Day 15 to 21.

The primary endpoint was the proportion of participants in each arm with eradication of GV at Visit 2, scheduled for Day 15 to 21. Eradication was operationally defined as a negative GV NAAT result. The outcome is binary for each participant: the test either met the negative-result definition or it did not. A missing GV NAAT result at the test-of-cure visit was classified as treatment failure, and the proportion denominator was based on the number randomized to the specified group.

A laboratory test provides a prespecified way to classify the presence or absence of the organism at the assessment visit. Using the same test and visit definition in both randomized groups makes the endpoint suitable for comparing the observed proportions. The registry does not report additional rules for interpreting the NAAT, such as handling indeterminate results, beyond its rule for missing results.

This endpoint measures test-defined eradication at a particular follow-up visit. It does not directly measure symptoms, longer-term persistence, recurrence, or clinical BV. The record lists adverse events as a secondary outcome, but the primary outcome remains the NAAT-based proportion. The endpoint’s interpretation therefore depends on its specified test, timing, and missing-result rule; it should not be treated as a broader measure of clinical status than the registry defines.

7. Results Posted in the Registry

• Amoxicillin: 0.0% (0/48) versus 0.0% (0/49) on control; risk difference 0.0 percentage points (95% CI 0.0 to 0.0); unadjusted two-proportion z-test p = n/a.

8. Statistical Interpretation

9. Statistical Methods Explained

What does a proportion estimate?

The endpoint is binary: each participant is classified according to whether the Visit 2 GV NAAT was negative under the registry’s definition. A proportion expresses the share of the specified analysis group that met that definition. The posted arm estimates are 0.16 for placebo and 0.21 for amoxicillin. They summarize the observed results under the stated denominator rule; they are not guarantees about the result for a future participant or another study population.

Because the study uses randomized parallel groups, the arm proportions can be considered side by side. Randomization is intended to support a fair comparison, but a randomized design does not remove sampling uncertainty. The estimates can vary from one study to another even if the underlying probabilities are unchanged. The registry does not report further details about randomization implementation or whether any adjusted analysis was planned.

What does the corrected chi-squared test address?

The sponsor reported a corrected chi-squared test for the comparison between the two groups, with p=0.757. In general, a chi-squared test assesses whether observed categorical outcome patterns are compatible with a null hypothesis of no association between group and outcome. A correction may be used to modify the test statistic or its approximation, but the registry does not describe the specific correction or provide the test statistic, degrees of freedom, or other calculation details.

A p-value is not the probability that the treatment has no effect. It is calculated under a null hypothesis and the assumptions of the test, and it describes the compatibility of the observed data with that hypothesis. A large p-value does not establish equivalence, while a small p-value would not by itself describe the size or importance of a difference. The record does not report a prespecified equivalence margin or a clinically important difference.

Why are confidence intervals useful?

In general, a confidence interval expresses the uncertainty associated with an estimate under a specified statistical procedure. The registry reports 95% confidence intervals for each arm’s proportion: 0.09 to 0.29 for placebo and 0.12 to 0.34 for amoxicillin. These intervals show a range of values compatible with each arm’s data under the interval method. The record does not identify the method used to calculate these intervals, so that detail is not reported in the registry record.

Arm-specific intervals do not substitute for an interval around the treatment contrast. The record does not provide an interpretable sponsor-estimated risk difference with a confidence interval. The posted recalculated risk difference and its limits are all 0.0, which does not agree with the posted arm-level proportions. That recalculated contrast cannot serve as a coherent estimate of the difference between arms.

How should missing outcomes be understood?

The registry states that participants with a missing GV NAAT at the test-of-cure visit were classified as treatment failures. This rule assigns missing results to the outcome category that does not meet eradication, rather than omitting them from the denominator. It therefore makes the endpoint definition depend on both the test result and the stated handling of missing results. The record does not report how many test results were missing, why any were missing, or whether sensitivity analyses using alternative assumptions were conducted.

The primary population is described as ITT: participants randomized, regardless of treatment receipt or adherence to administration procedures or schedule. This defines the intended analysis population, but the registry does not report enough detail to verify how every randomized participant was handled in the reported outcome calculation. The reported denominator is 49 for placebo and 48 for amoxicillin.

What do planning calculations mean?

In general, a sample-size illustration asks how many participants per arm might be needed to detect an assumed effect under stated design conditions. The registry’s planning assumptions describe a two-sided significance level of 0.05, equal allocation, and a normal approximation, with the effect set as a stated fraction of the observed effect. Such illustrations are planning calculations, not observed power and not evidence that the trial was designed around those assumptions. The registry does not report the trial’s own sample-size justification, the effect size originally targeted, or whether those planning assumptions were used in the protocol.

10. Confidence Intervals and Precision

The placebo eradication estimate was 0.16, with a 95% confidence interval from 0.09 to 0.29. The amoxicillin estimate was 0.21, with a 95% confidence interval from 0.12 to 0.34. Each interval spans a broad range of possible underlying proportions relative to its point estimate. This reflects uncertainty in the arm-specific estimates and cautions against treating either point estimate as exact.

In general, a confidence interval conveys the uncertainty of an estimate under the procedure used to construct it. A 95% confidence interval does not mean there is a 95% probability that the fixed underlying value lies within this particular interval. The registry does not identify the interval calculation method, so the precise assumptions behind these bounds are not reported in the registry record.

The intervals overlap, but their overlap is not a formal test of the difference between groups. More importantly for interpreting a treatment contrast, the registry does not report an interpretable sponsor-estimated risk difference with a confidence interval. The posted recalculated contrast and interval are 0.0, despite the distinct arm-level estimates; those values should not be taken as a precise estimate of no difference.

Thus, the record supports reading the arm-specific proportions as uncertain estimates and the sponsor’s p=0.757 as its reported comparison test. It does not support a precise numerical estimate of the between-arm difference or a claim that the groups are equivalent. The absence of a contrast interval limits direct assessment of how large a difference remains compatible with the study results.

11. Analyses Reported by the Sponsor

The sponsor-reported analysis compares the arms using a corrected chi-squared test and gives p=0.757. The registry does not provide a sponsor-estimated risk difference, a confidence interval for that contrast, or the detailed calculation inputs for the test. The sponsor result is therefore best described as the reported test result, alongside the arm-level outcome estimates.

The posted arm proportions are 0.16 for placebo and 0.21 for amoxicillin. By contrast, the posted recalculated risk difference is 0.0, and both interval limits are also 0.0. Those recalculated fields do not correspond to the arm-level proportions and should not be presented as a meaningful numerical contrast. The registry marks its discrepancy indicator as false, but that flag does not resolve the inconsistency visible between the posted contrast fields and arm estimates.

Since the record does not give enough information to reconstruct the sponsor’s comparison or reconcile it with the recalculated fields, the two should not be blended into a single estimate. The sponsor’s p-value can be reported as the trial’s own test result; the recalculated contrast should be identified as internally inconsistent with the arm proportions rather than interpreted as evidence of zero effect. The registry does not report why the contrast fields take those values.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists two outcome measures and two analyses. It identifies the proportion with GV eradication at Visit 2 as the primary outcome and lists related adverse events as a secondary outcome. However, the statistical plan for handling multiple outcomes or analyses is not reported in the registry record. It is not reported whether any multiplicity adjustment was planned, whether one outcome was prioritized for formal inference beyond its primary designation, or whether additional analyses were prespecified.

Multiplicity matters because conducting several statistical tests can increase the chance of at least one apparently notable result when all null hypotheses are true. In general, an analysis plan can specify how such testing is handled, but the registry does not give those details. The reported p=0.757 is the sponsor’s corrected chi-squared test for the primary comparison; the record does not explain the scope or interpretation of the correction beyond naming the method.

The primary endpoint description states that missing GV NAAT results at the test-of-cure visit were classified as treatment failures. This is a specific missing-data rule and should be included when interpreting the proportions. The record does not report the number of missing results, their distribution by arm, reasons for missingness, or sensitivity analyses under alternative missing-data assumptions. Therefore, the effect of missing outcome data on the reported estimates cannot be assessed from the registry record.

The primary population is defined as ITT and includes randomized participants regardless of whether they received treatment or complied with administration procedures or schedule. The outcome denominators reported are 49 for placebo and 48 for amoxicillin. The record does not explain the difference between the amoxicillin arm-description count of n=49 and the outcome denominator of 48, nor does it provide participant-level disposition details that would clarify that difference. The registry also does not report whether analyses outside the stated ITT definition were conducted.

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.

Most frequent non-serious adverse event terms

The registry lists serious adverse events affecting 0 participants at risk in the placebo group and 0 participants at risk in the amoxicillin group. It also reports 0 deaths in each group. These figures describe the serious-event and death counts recorded in the registry; they do not establish that no such event could occur in a larger or different population.

For other adverse events, the record reports 3 affected participants among 47 at risk in the placebo group and 10 among 45 at risk in the amoxicillin group. These are participant counts, not percentages. The record’s secondary outcome is the proportion reporting related adverse events, but the posted group-level figures do not state whether the affected participants’ events were judged related to study product. No rate should be inferred from these counts without a clearly specified denominator and event definition for that calculation.

In general, safety counts require attention to the population at risk, event definition, duration of observation, and attribution method. The registry states that the secondary outcome concerns related AEs and SAEs following the first dose through Visit 2, but it does not provide details of adjudication or a list of serious event terms. The figures are descriptive and should not be treated as a complete assessment of safety beyond the reported observation window.

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

The trial record reports that 0.16 of the placebo group and 0.21 of the amoxicillin group met the test-defined eradication endpoint at Visit 2. The sponsor-reported corrected chi-squared test gave p=0.757. The proportions are accompanied by broad 95% confidence intervals, and the registry does not report an interpretable sponsor-estimated contrast with a confidence interval. These registry entries describe the observed endpoint results and their limits; they do not establish that the treatments are equivalent or that there is no difference.

The endpoint is narrowly defined: a negative GV NAAT at Day 15 to 21. It does not establish longer-term eradication, address recurrence after that visit, or directly measure clinical BV. The study population was women colonized or infected with GV but without clinical evidence of BV, so the results should not automatically be generalized to people with clinical BV or to other populations not described in the record.

The trial’s missing-result rule classified missing NAAT results as failures. The registry does not report how many results were missing or the reasons for missingness. The reported ITT definition includes all randomized participants regardless of treatment receipt or adherence, but the record does not provide enough participant-level information to independently assess implementation of the analysis population.

The serious adverse event figures are zero in both groups, while other adverse events were reported in 3 participants at risk in the placebo group and 10 participants at risk in the amoxicillin group. These are descriptive registry counts and do not alone determine the safety profile. The registry record does not provide enough detail about event attribution, event terms, or follow-up beyond Visit 2 to support broader conclusions.

Overall, the posted results are one comparison from a completed Phase 2 study with a short, test-defined primary assessment. They provide estimates for the recorded groups and test, while leaving uncertainty about the contrast, missing outcomes, and the scope of conclusions beyond the studied population and visit.

15. Why This Trial Matters Statistically

This study illustrates why a trial result should be read at more than one statistical level. The arm proportions summarize the outcome in each group, confidence intervals convey uncertainty around those estimates, and the sponsor-reported p-value addresses a statistical comparison. None of these quantities alone answers every question about effect size, precision, or clinical relevance.

It also shows why outcome definitions and missing-data rules matter. Eradication was defined through a negative NAAT at a specified follow-up visit, while missing results were classified as failures. Those choices shape what the proportion represents. The endpoint is not interchangeable with longer-term eradication or clinical improvement, and the registry does not report those as the primary measure.

A further statistical issue is consistency between reported quantities. The arm proportions are 0.16 and 0.21, but the posted recalculated risk difference and its limits are 0.0. Because those values do not align, the recalculated contrast cannot be treated as a coherent estimate. Reporting the sponsor’s test separately from that inconsistent contrast avoids giving false precision to a result the record does not reconcile.

Finally, the record distinguishes trial planning from observed findings. Its study summary gives approximate screening and enrollment targets, while the registry reports enrollment as 97 and outcome denominators of 49 and 48. Planning illustrations based on an observed effect are not observed power and do not explain the original design unless the registry reports that they were used. The trial is therefore useful for understanding how design, endpoint definition, reporting consistency, and uncertainty jointly shape interpretation.

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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record leaves several trial-specific details unreported. It does not describe the randomization sequence, allocation concealment, the practical implementation of blinding, or the statistical analysis plan. These details would help readers assess how the study was conducted and how its inferential analyses were selected. The registry also does not report the rationale for its target sample size or the effect size used in the original design.

The reported primary outcome denominator for amoxicillin is 48, while the arm description lists n=49. The placebo arm description and outcome denominator both list 49. The record does not explain the difference for amoxicillin, give participant disposition by reason, or report the number of missing NAAT results. Although the endpoint definition classifies missing test results as failures, the impact of that rule cannot be assessed without those details.

The posted contrast fields are also limited: the recalculated risk difference and both interval limits are 0.0, which does not align with the reported arm proportions of 0.16 and 0.21. The registry does not provide a sponsor-estimated contrast and confidence interval that would clarify the size and precision of the between-group difference. The sponsor’s corrected chi-squared result is reported as p=0.757, but the record does not describe the correction in detail or provide the test statistic.

The primary outcome is measured at Visit 2, Day 15 to 21. The record does not report a longer-term outcome for eradication or recurrence. The safety information is also limited: the registry lists participant counts but does not provide serious adverse event terms, detailed attribution procedures, or information beyond the stated follow-up. These limitations constrain interpretation to the recorded endpoint, assessment time, and outcomes.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Treatment of Gardnerella Vaginalis Vaginal Colonization With Amoxicillin — NCT03211156.
• Registry record last updated 2020-08-28; status last verified 2017-06-28.
• Results first posted 2020-08-28.
• 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

NCT03211156 is a completed randomized, double-blind, parallel-group Phase 2 study comparing amoxicillin with placebo for GV colonization or infection in women without clinical evidence of BV. The registry reports enrollment of 97. The amoxicillin regimen was 2 x 250 mg capsules by mouth twice daily for 7 days; placebo was given as 2 capsules twice a day for 7 days.

The primary outcome was a negative GV NAAT at Visit 2, Day 15 to 21. The posted proportions were 0.16 for placebo and 0.21 for amoxicillin, with 95% confidence intervals of 0.09 to 0.29 and 0.12 to 0.34, respectively. The sponsor-reported corrected chi-squared test gave p=0.757. The posted recalculated contrast fields do not align with the arm estimates, and the registry does not provide an interpretable sponsor contrast with a confidence interval.

Missing test results were classified as treatment failures. Serious adverse event and death counts were 0 in each group. The registry lists other adverse events in 3 participants at risk in the placebo group and 10 in the amoxicillin group.