H7N9 Vaccination With and Without AS03 and Unadjuvanted H3N2v Vaccination: Standard and Systems Biology Analyses: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for H7N9 Vaccination With and Without AS03 and Unadjuvanted H3N2v Vaccination: Standard and Systems Biology Analyses (NCT02921997), a randomized phase 2 study. The primary outcome is Percentage of Participants Achieving Seroconversion Based on HAI Titer (A/H7N9 Vaccine, With and Without Adjuvant).

2. Clinical Question

3. Trial Design

This was a Phase II, randomized, parallel-group study with enrollment listed as 30. The registry marks the design as double-masked, while the brief summary describes the study as partially blinded. The registry record does not explain the difference between those descriptions or specify which parts of the study were masked.

The three arm descriptions each list 10 subjects. One arm received a single dose of unadjuvanted A/H3N2v vaccine at Day 1. The other arms received two doses of A/H7N9 vaccine at Day 1 and Day 29: one with AS03 and one without adjuvant. The primary HAI comparison is between the two A/H7N9 arms, not between A/H7N9 and A/H3N2v.

The trial was conducted at a single center and is marked completed. Its recorded start date is 2016-11-07, and its primary completion date is 2017-03-28. These dates describe the registry record; they do not add information about recruitment duration or follow-up beyond the stated endpoint timeframes.

The objectives extended beyond the primary HAI response. The brief summary also describes gene-expression assessments after A/H7N9 vaccination and a set of immunological outcomes. The registry lists 15 outcome measures and 12 posted secondary results. A detailed statistical analysis plan, including the methods used for the secondary outcomes, is not reported in the registry record.

The randomized design supports a comparison of the assigned A/H7N9 regimens within this study, but the small arm sizes and the specialized healthy-adult population limit how broadly the posted results can be generalized. The record calls the study small and targeted; it does not report a sample-size justification or the intended precision for the primary comparison.

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

Analysis population for the primary outcome, as defined in the registry: The intent-to-treat (ITT) population, A/H7N9 + AS03 and A/H7N9 arms: Includes all participants in the A/H7N9 + AS03 and A/H7N9 arms who received at least one dose of study vaccine and contributed both pre- and at least one post-study vaccination blood samples for serological assessment (HAI or Neut antibody assays) for which valid results were reported.

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 Percentage of Participants Achieving Seroconversion Based on HAI Titer (A/H7N9 Vaccine, With and Without Adjuvant), measured over Day 29 Post-Vaccination 2 (Day 57).

The primary endpoint is the percentage of participants who achieved seroconversion based on HAI titer for the A/H7N9 vaccine, with and without AS03. The measurement was made at Day 29 after the second vaccination, identified as Day 57. HAI is a laboratory assay of serum antibody response. The registry specifies that serum came from venous blood samples and that the assay used the H7N9 strain A/Shanghai/02/2013xPR8.

Seroconversion was defined using either of two patterns. A participant could have a pre-vaccination HAI titer below 1:10 and a post-vaccination titer at least 1:40, or have a pre-vaccination titer at least 1:10 and show a minimum four-fold rise in post-vaccination antibody titer. Thus, the endpoint is not simply a post-vaccination titer threshold: its definition depends on the participant’s baseline titer as well as the later measurement.

As a percentage endpoint, it expresses how common the defined response was among the participants included in the analysis. It is used to summarize the proportion meeting a prespecified serological response definition, rather than the magnitude of each person’s antibody level. The record does not report a clinical outcome linked to this serological endpoint, so its interpretation should remain focused on the measured immune response.

The analysis population required at least one vaccine dose and both a pre-vaccination and at least one post-vaccination blood sample for serological assessment. The registry does not report further detail about how incomplete measurements were handled or whether the posted percentages use a different denominator from the listed arm sizes.

7. Results Posted in the Registry

• A/H7N9: 0.0% (0/10) versus 80.0% (8/10) on control; risk difference -80.0 percentage points (95% CI -104.8 to -55.2); unadjusted two-proportion z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

General: The primary outcome is binary at the participant level: each person either meets the registry’s seroconversion definition or does not. A percentage summarizes the share meeting that definition within an analysis group. For this trial, the posted values are 80.0% for A/H7N9 + AS03 and 0.0% for A/H7N9 without adjuvant.

What does a confidence interval describe?

General: A confidence interval expresses uncertainty around an estimate. It is not a range containing a known fixed set of participant outcomes, nor does it guarantee that the true value lies within the particular interval. Repeated use of a method under its assumptions would produce intervals with the stated coverage property over the long run. In an individual trial, a wide interval signals that the estimate is not very precise.

The registry reports 95% confidence intervals for both arm percentages: 44.4% to 97.5% for A/H7N9 + AS03 and 0.0% to 30.8% for A/H7N9. Their breadth is consistent with the limited information available from small groups. The record does not state the exact interval method used, so it is not appropriate to attribute these bounds to a particular formula.

What is the risk difference?

General: A risk difference compares the probability of an event between groups by subtracting one group’s event probability from the other’s. Its interpretation depends on which group is placed first. The registry reports the A/H7N9 arm relative to the A/H7N9 + AS03 arm, with an estimate of -0.8. The negative sign is consistent with the lower observed seroconversion percentage in the first-named comparison arm. The posted comparison also gives p < 0.001, identified as unadjusted.

A p-value measures how incompatible the observed data, or data more extreme, would be with a specified null model under the assumptions of the test. It is not the probability that the null hypothesis is true, and it does not quantify the probability that the result occurred by chance. The registry record does not name the test or fully state the comparison’s modeling assumptions. The unadjusted p-value should therefore be reported as posted without inferring an unreported procedure.

Why does small-group estimation need care?

General: With a binary endpoint, each participant contributes a response or nonresponse. When groups are small, a change in the outcome of a single participant can make a noticeable difference to the estimated percentage. A result at a boundary, such as 0.0%, can coexist with a confidence interval whose upper limit is above zero. This is why the interval is important alongside the point estimate.

The trial’s arm descriptions list 10 subjects in each group. The endpoint population requires at least one vaccination and qualifying blood samples. The record does not report the exact analyzed denominator separately, so the listed arm sizes should not automatically be treated as confirmed endpoint denominators.

What does the planning illustration mean?

General: A sample-size planning calculation asks how many participants might be needed in a future study to detect an assumed effect under specified statistical assumptions. Such a calculation is conditional on the selected effect size, significance level, allocation, and approximation. It is not evidence that a completed study had a particular probability of detecting its result, and it is not a retrospective measure of power.

The planning illustration accompanying the posted material is described as using the observed effect, a two-sided significance level of 0.05, equal allocation, and a normal approximation. It considers effects expressed as fractions of the observed effect. The registry does not report that these were the original trial-design assumptions or give a sample-size justification for this study. Any participant counts in such an illustration are planning quantities, not observed trial results.

General: Over-dispersion matters when variability is greater than a statistical model assumes. If ignored, standard errors may be too small and confidence intervals or tests may be too optimistic. The primary endpoint is a binary participant-level response, but the registry record does not report an over-dispersion assessment or model-based adjustment. This general consideration does not imply that over-dispersion was present or that a particular correction was used.

The outcome list also includes gene-expression, cytokine, chemokine, and other serological measures. Those outcomes can involve repeated measurements and many comparisons, which raises distinct questions about correlation, modeling, and multiplicity. The registry record does not report the detailed methods for those analyses, so no specific model or correction should be inferred from the outcome titles alone.

10. Confidence Intervals and Precision

The precision of a percentage estimate is conveyed in part by its confidence interval. For A/H7N9 + AS03, the reported seroconversion percentage is 80.0%, with a 95% confidence interval from 44.4% to 97.5%. For A/H7N9 without adjuvant, the reported percentage is 0.0%, with a 95% confidence interval from 0.0% to 30.8%. Both intervals cover a broad range of plausible underlying percentages, so the point estimates should not be treated as exact population values.

The intervals are especially important because the arm descriptions list 10 subjects per group. With limited observations, a percentage can change substantially when the outcome status of one participant changes. A zero observed percentage therefore does not establish that the underlying response probability is zero; the reported interval itself extends above zero. Likewise, the interval around 80.0% indicates appreciable uncertainty despite the high point estimate.

The posted risk-difference estimate is -0.8, and the registry reports a comparison p-value of p < 0.001. A small p-value and broad interval estimates address different aspects of the evidence. The p-value concerns compatibility with a null comparison under the analysis used, while intervals communicate uncertainty about the estimated response rates. Neither provides a guarantee about replication or generalizability.

The exact method used to calculate the intervals is not reported in the registry record. Nor does the record describe sensitivity analyses or adjustments for missing endpoint measurements. Precision should therefore be read from the posted intervals without assuming unreported methodological details.

11. Multiplicity, Missing Data and Analysis Populations

The trial assessed several types of outcomes, including HAI and neutralization seroconversion, gene expression, and cytokine and chemokine concentrations. The registry lists 15 outcome measures and 12 posted secondary results. When many outcomes, time points, or comparisons are examined, the chance of observing apparently notable findings can increase. Statistical methods may address this multiplicity, but an adjustment strategy or hierarchy for interpreting outcomes is not reported in the registry record.

The primary HAI result is reported with an unadjusted comparison p-value of p < 0.001. The record does not state whether this comparison was prespecified in a full statistical analysis plan, whether any multiplicity adjustment was planned, or how the primary and secondary findings were to be interpreted together. The presence of a primary endpoint label does not by itself establish those details.

The primary analysis population included participants in the two A/H7N9 arms who received at least one dose and contributed pre-vaccination and at least one post-vaccination blood sample for serological assessment. This definition makes sample availability part of inclusion in the endpoint analysis. The registry record does not report how many randomized participants lacked required samples, why any measurements were missing, or whether missing values were imputed.

The available description does not establish whether all randomized participants were included in the primary analysis or whether the posted analysis followed any additional exclusions. The exact endpoint denominators are not separately stated in the results included in the registry record. These gaps limit assessment of potential bias from missing measurements and the relationship between the analysis population and the original randomization.

12. 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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: intussusception (10.0%).

The registry lists serious adverse event figures for the three study arms. For A/H3N2v, the record reports 0 serious adverse events among 10 at risk; for A/H7N9 + AS03, it reports 0 among 10 at risk; and for A/H7N9 + PBS, it reports 1 among 10 at risk. The registry’s top serious-event term is Intussusception, listed as 10.0%. The record does not provide enough detail in these figures to establish the event’s timing, clinical course, or relationship to vaccination.

Other adverse events are listed separately: 8 among 10 at risk for A/H3N2v, 10 among 10 at risk for A/H7N9 + AS03, and 7 among 10 at risk for A/H7N9 + PBS. Deaths are recorded as 0 in each group. These counts describe the registry’s posted safety summaries and should not be confused with the primary HAI endpoint.

Because the groups are small, these counts provide limited information about the frequency of uncommon events. The registry record does not report further detail about event severity beyond the serious-event classification, event definitions, or comparative statistical analyses. The figures are descriptive and do not establish whether an event was caused by a vaccine.

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

The primary result shows different observed HAI seroconversion percentages in the two A/H7N9 arms at the stated assessment time: 80.0% with AS03 and 0.0% without adjuvant. The reported risk difference is -0.8 for the A/H7N9 arm relative to the A/H7N9 + AS03 arm, and the unadjusted comparison is reported as p < 0.001. These findings describe the trial’s measured serological response in its defined analysis population.

The result does not establish that every participant receiving A/H7N9 with AS03 responded, or that no participant in the unadjuvanted arm could respond in another setting. The confidence intervals are wide, and the reported 0.0% estimate has an upper confidence limit of 30.8%. The estimates should therefore be understood as uncertain, particularly in light of the small groups.

The outcome is an HAI seroconversion definition based on pre- and post-vaccination titers. It is a laboratory measure, not a direct measure of illness, hospitalization, or any other clinical outcome. The registry record does not report a clinical efficacy endpoint for this primary comparison. The result therefore does not by itself quantify protection against influenza or establish a clinical benefit.

The comparison also does not answer every question about the vaccine regimens. It concerns the two-dose A/H7N9 arms and the specified Day 57 assessment. The separate A/H3N2v arm received a different vaccine on a different schedule and is not part of this primary contrast. Findings from these arms should not be conflated.

Finally, the result is from a Phase II study of healthy adults at a single center. The registry record does not establish that the same response percentages would apply to other populations, settings, or vaccination schedules. The result is a description of posted trial evidence, not medical advice or a basis for an individual treatment decision.

14. Why This Trial Matters Statistically

This trial illustrates why an immunogenicity result should be read together with its definition, analysis population, and uncertainty. Seroconversion is not simply a positive-or-negative laboratory result independent of baseline: the registry defines it using baseline and post-vaccination HAI titers. That definition determines which participants count as responders and gives the percentage its specific meaning.

The two reported arm percentages are markedly different, while their confidence intervals remain broad. This contrast makes the trial useful for discussing how a strong separation in observed results can coexist with limited precision about the underlying rates. A p-value alone does not communicate that uncertainty, and the point estimates alone do not show the range of values compatible with the interval procedure.

The design also highlights the importance of keeping comparisons aligned with the endpoint. The primary HAI result compares the A/H7N9 regimens with and without AS03. The A/H3N2v group had a different vaccine and schedule, so it is not interchangeable with the unadjuvanted A/H7N9 comparator. Correctly identifying the comparison prevents a misleading interpretation of the posted results.

Finally, the registry record describes multiple immune outcomes but does not report a detailed multiplicity strategy or full analysis plan. This makes it important to distinguish the primary result from secondary findings and to avoid treating every posted outcome as if it had the same inferential status. The study is a useful example of how endpoint definitions, small-group precision, analysis populations, and reporting detail shape statistical interpretation.

15. 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 primary comparison involves small groups: each A/H7N9 arm description lists 10 subjects. This limits precision for percentage outcomes and makes the estimates sensitive to individual response classifications. The confidence intervals reflect substantial uncertainty around both arm percentages.

The stated analysis population required at least one vaccine dose and qualifying blood samples. The registry record does not report separate endpoint denominators, the number of participants excluded from analysis, or the reasons for missing samples. This prevents a full assessment of how the analyzed group relates to all randomized participants and whether missing measurements could affect the result.

The registry identifies the design as double-masked, while its brief summary calls the study partially blinded. The record does not explain this discrepancy. Details of allocation concealment, masking procedures, and who remained masked are not reported in the registry record.

The primary endpoint is a serological response measured at the stated time point. The registry record does not report a clinical efficacy endpoint for this comparison, so the HAI finding cannot be interpreted as a direct estimate of clinical protection. The single-center, healthy-adult population also limits the scope of direct generalization.

Finally, the registry does not report a full statistical analysis plan, the interval calculation method, a multiplicity adjustment strategy, or details of missing-data methods. Although the record lists many secondary outcomes, those omissions limit assessment of how the wider set of analyses was planned and interpreted.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: H7N9 Vaccination With and Without AS03 and Unadjuvanted H3N2v Vaccination: Standard and Systems Biology Analyses — NCT02921997.
• Registry record last updated 2024-10-09; status last verified 2019-08-08.
• Results first posted 2022-01-26.
• 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.

19. Record Summary

The registry record describes a completed, randomized, parallel-group Phase II study of influenza vaccines in healthy adults. Enrollment is listed as 30, with three arms: one A/H3N2v dose, two A/H7N9 doses with AS03, or two A/H7N9 doses without adjuvant. The primary result compares HAI seroconversion in the two A/H7N9 arms at Day 57.

Seroconversion was reported in 80.0% of the A/H7N9 + AS03 arm and 0.0% of the A/H7N9 arm. The posted unadjusted comparison reports a risk difference of -0.8 and p < 0.001. Both arm confidence intervals are broad. The endpoint population required vaccination and qualifying blood samples, while the record does not report separate analyzed denominators or detailed missing-data methods.

The record also lists secondary immune outcomes and safety summaries. It does not report a complete statistical analysis plan, a multiplicity strategy, or a clinical efficacy result for the primary comparison.