Azithromycin Reduction to Reach Elimination of Trachoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Azithromycin Reduction to Reach Elimination of Trachoma (NCT04185402), a randomized phase 4 study. The primary outcome is Ocular Chlamydia Measured in a Population-based Sample of 0-9 Year-old Children.

2. Clinical Question

The clinical question is about a community-level distribution policy, not an individually assigned treatment comparison. Communities were randomized to continue or discontinue annual distribution, and the primary endpoint was assessed in a population-based sample of children. The registry describes cross-sectional data collection at baseline and endline, rather than follow-up of the same children as a cohort.

The registry record reports zero participants with the primary endpoint in both arms and a risk difference of 0. These posted values describe the recorded endpoint counts; they do not by themselves establish that the strategies are equivalent. Interpretation depends on how the outcome and analytic population were defined and on the statistical analysis used. Those details are not fully reported in the registry record.

3. Trial Design

This is a randomized, double-masked, parallel, Phase 4 trial. The trial is described as cluster-randomized: communities, rather than individual children, were assigned to continuation or discontinuation. It was conducted in hypoendemic communities of Maradi, Niger. The registry describes eligible communities as having up to 20% Trachomatous Inflammation - Follicular prevalence after at least 5 years of mass azithromycin distribution.

The planned comparison was discontinuation versus continuation of annual mass distribution. The continuation strategy was described as 3 additional years of annual distribution. Its arm description specifies distribution to individuals aged 1 month and older several weeks after baseline and 36-month monitoring visits. The discontinuation description says individuals receive no treatment. The registry lists azithromycin as the intervention. Further details of how the double masking was implemented are not reported in the registry record.

The record lists enrollment of 3931, a start date of 2021-05-29, primary completion on 2024-12-18, and status as completed. These are registry-level trial details; enrollment should not be assumed to equal the number analyzed for every endpoint. The primary endpoint description says participant flow reflects counts of clusters and enrolled children, whereas the number analyzed includes eligible individuals captured in cross-sectional villages and may exceed the participant-flow count. The rest of that explanation is not reported in the registry record.

For the primary outcome, the posted arm counts are 1993 for continuation and 1978 for discontinuation. The endpoint is assessed at 36 months. Since allocation occurred by community, an analysis would ordinarily need to account for clustering when estimating uncertainty. The specific analysis method, any cluster adjustment, and the masking procedures are not reported in the registry record, so the posted comparison should be read as the result recorded rather than as a fully described statistical analysis.

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: Communities (grappes) were randomized as the unit of allocation. Monitoring surveys for outcome assessment enrolled children residing in study communities; adults were not enrolled as study participants for outcome assessment.

Baseline Survey

36 Months Survey

Analysis population for the primary outcome, as defined in the registry: This is a cluster-randomized trial wit cross-sectional data collection at baseline and endline. Participant Flow reflects counts of clusters and enrolled children, not the analytic dataset. The number analyzed includes all eligible individuals captured in the cross-sectional villages, which may exceed the number 'Started' in Participant Flow.

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 Ocular Chlamydia Measured in a Population-based Sample of 0-9 Year-old Children, measured over 36 months.

The primary endpoint is titled “Ocular Chlamydia Measured in a Population-based Sample of 0-9 Year-old Children.” It is a count of participants, assessed by PCR at 36 months. In practical terms, it records how many eligible children in the sampled population had ocular Chlamydia detected by that method. It is a binary event summarized as a participant count, not a measure of organism load or a continuous score.

This endpoint is relevant to the stated trial question because it assesses ocular infection in children after communities followed different mass-distribution strategies. The registry also lists infectious load among infected children as a secondary outcome, distinguishing infection occurrence from the amount of infection among those found to be infected. Numerical results for that secondary outcome are not reported in the registry record.

The registry describes cross-sectional collection at baseline and endline. Accordingly, the endpoint is a population snapshot at the monitoring time, not a reported measure of each child's change over time. The description notes that the number analyzed includes eligible individuals captured in cross-sectional villages and may exceed the participant-flow count, but the explanation is truncated in the registry record. The precise eligibility rules, PCR procedures, and handling of missing or indeterminate tests are not reported in the registry record.

For the primary endpoint, the registry lists zero participants with the event in each arm. It also lists a risk difference of 0 with interval limits of 0 and 0. These are the recorded values; the record does not give enough methodological detail to determine how the interval was constructed or how the community-level allocation was incorporated.

7. Results Posted in the Registry

• Azithromycin Discontinuation: 0.0% (0/1978) versus 0.0% (0/1993) on control; risk difference 0.0 percentage points (95% CI 0.0 to 0.0); unadjusted two-proportion z-test p = n/a.

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.

Conjunctival Inflammation

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Azithromycin Discontinuation: 0.2% (4/1981) versus 0.3% (6/1999) on control; risk difference -0.1 percentage points (95% CI -0.4 to 0.2), risk ratio 0.67 (0.19–2.38); unadjusted two-proportion z-test p = 0.536.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Azithromycin Discontinuation and the control. The interval gives the range of effects that remain compatible with the data.

The registry names infectious load among infected children and conjunctival inflammation as secondary outcomes. Numerical results are posted for conjunctival inflammation, but numerical results for infectious load are not reported in the registry record. The secondary evidence is therefore incomplete: it cannot provide a broad numerical picture across all the listed outcomes.

The conjunctival-inflammation comparison is centered close to zero on the risk-difference scale, and its reported interval includes zero. The ratio measures likewise have intervals that include 1. This pattern does not establish that the secondary endpoint is identical between strategies; rather, the estimates are uncertain and compatible with no difference as well as differences in either direction under the reported calculations. These observations should be considered alongside, not substituted for, the primary endpoint, which records zero events in both arms.

Secondary outcomes generally have less confirmatory weight than a designated primary outcome, especially when several outcomes may be examined. Each additional comparison creates more opportunity for an apparently notable result to arise by chance. Adjusting for multiplicity can help control the overall false-positive risk, but the registry record does not report whether any adjustment was planned or used. The conjunctival-inflammation p-value is explicitly unadjusted, and no adjusted result is listed.

The record does not report a hierarchy for secondary outcomes, a multiplicity strategy, or numerical findings for every listed secondary outcome. Consequently, the secondary findings are best treated as supportive or exploratory descriptions rather than a separate definitive test of the trial's main question. Their relationship to the primary result is contextual: they concern different manifestations or measures, and they cannot fill gaps in the primary analysis details.

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

• Infectious Load of Chlamydia Among 0-9 Year-old Children Infected With Ocular Chlamydia
• Conjunctival Inflammation

9. Statistical Interpretation

10. Statistical Methods Explained

For a binary endpoint, the event proportion is the number of participants with the event divided by the number assessed. A risk difference compares two such proportions. It is expressed on the proportion scale: zero means the recorded proportions are equal, a negative value means the first named arm has a lower recorded proportion than the reference arm, and a positive value means it has a higher recorded proportion. Those general interpretations do not establish causality or account automatically for the way participants were sampled.

The primary comparison is posted as a risk difference of 0, with both interval limits at 0. The event counts are zero in each arm. The registry does not report a primary p-value, an analysis model, or the method used to construct the interval. A confidence interval is generally a range of estimates compatible with the observed data under a specified statistical procedure. Its usefulness depends on the procedure's assumptions. When the posted limits are identical to the estimate, the interval as recorded displays no variation, but the record does not explain how this was obtained. It should not be read as proof that the true effect is known without uncertainty.

The trial randomized communities. In general, outcomes from people in the same cluster can be correlated because they share a setting. This correlation is called within-cluster dependence. If it is ignored, the effective information in a sample can be overstated and standard errors can be too small. Cluster-aware methods can account for this dependence, but the registry record does not identify the method used for the primary result, report the number of randomized communities, or describe any adjustment. The arm counts are participant counts and should not be mistaken for counts of randomized communities.

For conjunctival inflammation, the posted risk difference is -0.001, with limits -0.004 and 0.002. Since the interval extends across zero, the result does not clearly distinguish a difference in either direction from no difference under the reported calculation. The unadjusted p-value is 0.536. A p-value measures how unusual data at least as inconsistent with a specified null hypothesis would be if that hypothesis and the statistical model held. It is not the probability that the null hypothesis is true, nor a measure of clinical importance. The registry calls this value unadjusted; no adjusted p-value is reported.

The record also gives a risk ratio and an odds ratio for conjunctival inflammation. A risk ratio compares the event proportions as a ratio; a value of 1 corresponds to equal proportions. An odds ratio compares event odds rather than event proportions, and 1 similarly corresponds to equal odds. The posted risk ratio is 0.673 with limits 0.190 and 2.38. The posted odds ratio is 0.672 with limits 0.189 and 2.385. Both intervals include 1, indicating substantial uncertainty in the ratio estimates under the procedures used. The registry does not report the method or adjustments behind these estimates.

Over-dispersion is general statistical variation greater than a chosen model expects. It can arise when outcomes are more variable or more dependent than the model assumes. In a cluster trial, unaccounted similarity within communities can contribute to underestimating uncertainty. No over-dispersion parameter or related model details are reported in the registry record. It is therefore not possible to verify from the record whether the posted uncertainty reflects the trial's clustered structure.

The record contains a planning illustration based on the observed conjunctival-inflammation effect. It assumes a two-sided significance level of 0.05, equal allocation, and a normal approximation, and considers planning for 100%, 75%, and 50% of the observed effect. Such calculations describe hypothetical participant requirements under their assumptions; they are not observed power, do not show the trial's original sample-size justification, and do not establish what effect the trial was designed to detect. The original sample-size assumptions and justification are not reported in the registry record.

The registry lists no reported analyses. It does not describe interim analyses, missing-data procedures, covariate adjustment, prespecified analysis populations, or a statistical analysis plan. These omissions matter because analysis choices can affect estimates and uncertainty. Without those details, interpretation should remain tied to the posted counts and comparisons rather than assuming a particular model or analysis strategy.

11. Confidence Intervals and Precision

Precision concerns how narrowly an estimate identifies the range of effects compatible with the data under an analysis method. A confidence interval expresses this uncertainty; a narrow interval generally indicates more precision than a wide one, provided the method and assumptions are appropriate. Precision is not the same as clinical importance, and an estimate near zero is not by itself evidence of equivalence.

For the primary endpoint, the registry reports a risk difference of 0 and interval limits of 0 and 0. This interval has no width as posted. However, the record does not report how the interval was calculated, a primary p-value, or whether the calculation accounts for community randomization. Given those omissions and the zero event counts in both arms, the interval should be described as the registry's reported interval, not treated as a complete measure of uncertainty about the underlying event risks.

For conjunctival inflammation, the risk-difference interval runs from -0.004 to 0.002 around the estimate of -0.001. It includes zero and allows values on both sides of no difference under the reported calculation. The risk-ratio interval, from 0.190 to 2.38, and odds-ratio interval, from 0.189 to 2.385, each include 1, the value corresponding to equal groups on those scales. These intervals are broad relative to their point estimates, so the secondary comparison is imprecise in the sense that a range of effects remains compatible with the calculation.

These intervals do not answer whether any remaining range would be important in practice. The registry does not report a threshold for a clinically meaningful difference, an equivalence margin, or the uncertainty method. Cluster-aware uncertainty is especially relevant because allocation was by community. Until the analysis approach is clear, precision should be judged cautiously and separately from the point estimates.

12. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several outcomes, comparisons, or analyses are considered. If each is evaluated separately, the chance of at least one apparently positive result can increase. Statistical plans sometimes address this through prespecified outcome hierarchies or adjustment procedures. The registry lists secondary outcomes and one unadjusted secondary p-value, but does not report a multiplicity plan, adjusted results, or a hierarchy governing interpretation. The extent to which multiplicity was addressed is therefore not reported in the registry record.

The registry lists no reported analyses. It does not state which analysis population was prespecified, whether all randomized communities were included in each analysis, or how eligibility for the cross-sectional endpoint was applied. The primary description distinguishes participant flow from the analytic dataset and says that the number analyzed includes eligible individuals captured in cross-sectional villages and may exceed the participant-flow count. The explanation ends without clarifying the circumstances or exact analysis population. That missing detail limits assessment of how the displayed arm counts relate to enrollment and participant flow.

Missing-data handling is also not reported in the registry record. There is no description of incomplete PCR assessments, missing endpoint status, exclusions, or sensitivity analyses. Zero recorded events are not the same as a statement that there were no missing measurements. Without a denominator definition and missingness account for each endpoint, the completeness of the outcome data cannot be assessed from the registry values alone.

Other analysis details are not reported in the registry record: the statistical model, cluster adjustment, covariate adjustment, interim analyses, and the statistical analysis plan. These are not safe to infer from the randomized, double-masked, parallel design. The record identifies community-level allocation but does not say how that design was reflected in inference. These omissions do not negate the posted results, but they constrain how much can be concluded from them and make independent evaluation of the analysis difficult.

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 reports zero serious adverse events in the continuation group among 1999 at risk and zero in the discontinuation group among 1981 at risk. It also reports zero other adverse events and zero deaths in both groups. These are the safety figures recorded for the trial; the registry does not list serious adverse-event terms.

A count of zero recorded events means no events are listed in the relevant registry fields. It does not establish that an event is impossible or that risk is absent. The number at risk is the registry's stated denominator for these safety summaries and should not be assumed to match the primary endpoint population exactly. The safety fields do not provide event timing, severity criteria, assessment procedures, or follow-up completeness. Those details are not reported in the registry record.

Because the record lists no serious events in either group, it does not support a comparative estimate of relative serious-event risk from observed events. The absence of listed events should be interpreted narrowly, as a report of zero counts, rather than as proof of equal safety or a guarantee about individuals.

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

The posted primary result is a record of zero ocular Chlamydia events in each arm and a risk difference of 0 at 36 months. It describes the observed registry counts in the cross-sectional child sample. It does not, on its own, show that the underlying risks are exactly equal, that discontinuation and continuation are equivalent, or that either strategy causes a particular outcome.

The trial's community-level randomization is relevant to causal interpretation, but a credible comparison also depends on how the analysis handled community clustering, the endpoint population, and missing data. Those methods are not reported in the registry record. The primary interval is listed as 0 to 0, but the record does not explain its calculation. That interval cannot replace the absent methodological detail or establish certainty about the true effect.

The conjunctival-inflammation result is a secondary comparison with an unadjusted p-value and intervals that include the null values on the reported scales. It does not provide confirmatory evidence that the groups differ or are the same. The other named secondary outcome, infectious load, has no numerical result reported in the registry record. Safety fields list zero serious adverse events, other adverse events, and deaths in both groups; these counts do not establish that risks are absent.

The record supports a careful description of what was entered: a completed, cluster-randomized comparison with recorded zero primary events in each arm and limited secondary numerical reporting. It does not support a treatment recommendation, a claim of benefit or lack of benefit, a claim of equivalence, or a complete assessment of safety. Those broader conclusions would require sufficiently detailed outcome, analysis, and uncertainty reporting.

15. Why This Trial Matters Statistically

This trial raises a statistical question that is important in community-level research: how should a comparison be interpreted when assignment occurs by community but outcomes are counted among individual participants? General statistical reasoning recognizes that people within the same community may have correlated outcomes. The number of individuals alone does not describe all the information available for estimating an effect; the number and structure of clusters also matter. The registry identifies cluster randomization but does not report the cluster-level analysis details needed to evaluate this aspect of precision.

The record also illustrates why a point estimate and an uncertainty interval should be read together with their methods. The primary estimate is zero, and the posted interval has identical limits, yet the record does not explain how that interval was obtained. Without that explanation, the apparent precision cannot be evaluated. A zero estimate is a description of the recorded comparison, not proof of no underlying difference.

Another statistical issue is the distinction between the primary question and secondary outcomes. The primary endpoint concerns ocular Chlamydia detected by PCR in children, while the listed secondary outcomes address infectious load and conjunctival inflammation. These outcomes are related to the trial's subject but measure different things. Their evidential weight also depends on prespecification and multiplicity handling, neither of which is described in detail in the registry record.

Finally, the record differentiates participant flow from the cross-sectional analytic population, but does not fully explain that distinction. Clear definitions of denominators, missingness, cluster inclusion, and analysis methods are central to reproducible interpretation. The trial is therefore statistically informative as a case in which recorded estimates need to be separated from the strength of the evidence that can be assessed from registry reporting.

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 limitations specific to this record constrain interpretation. The primary endpoint counts are zero in both arms, and the posted risk-difference interval is also zero to zero. The registry does not explain the interval calculation or report a primary p-value. This leaves uncertainty about how the displayed result was derived and how much uncertainty should be attributed to it.

The trial is cluster-randomized, but the registry does not report the number of randomized communities, the distribution of communities by arm, or the statistical method used to account for clustering. Participant counts cannot substitute for those details. The record also does not report whether cluster-level imbalance, covariates, or other design features were incorporated in the analysis.

The primary population description says the number analyzed includes eligible people captured in cross-sectional villages and may exceed the participant-flow count, but the explanation is incomplete in the registry record. The exact denominator rules, eligibility criteria, and relationship between enrollment, participant flow, and endpoint analysis are not fully described. Missing-data handling and sensitivity analyses are not reported.

Numerical results are available for only one of the named secondary outcomes. The record does not report multiplicity control, a secondary-outcome hierarchy, or a statistical analysis plan. The planning illustration is based on the observed secondary effect and is not the trial's design justification. These reporting gaps mean the posted results can be described, but a comprehensive evaluation of the analysis and its robustness is not possible from the registry record alone.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Azithromycin Reduction to Reach Elimination of Trachoma — NCT04185402.
• Registry record last updated 2026-03-25; status last verified 2026-03.
• Results first posted 2026-03-25.
• 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, double-masked, parallel Phase 4 cluster trial in hypoendemic communities of Maradi, Niger. It compares continuation with discontinuation of annual mass azithromycin distribution and lists enrollment of 3931. The primary endpoint is ocular Chlamydia detected by PCR among children aged 0-9 years at 36 months.

The primary results list zero participants with the endpoint in each arm and a risk difference of 0, with interval limits of 0 and 0. Conjunctival inflammation is the secondary outcome with numerical results; infectious-load results are not reported in the registry record. The registry also lists zero serious adverse events, other adverse events, and deaths in both groups.

Important analysis details—including the primary statistical method, cluster adjustment, missing-data handling, and multiplicity strategy—are not reported in the registry record. The estimates should therefore be read as posted counts and comparisons, not as proof of equivalence or a complete account of uncertainty.