Immunogenicity and Safety Study of Different Formulations of GlaxoSmithKline (GSK) Biologicals H7N1 Influenza Vaccine Administered to Adults 65 Years of Age and Older: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Immunogenicity and Safety Study of Different Formulations of GlaxoSmithKline (GSK) Biologicals H7N1 Influenza Vaccine Administered to Adults 65 Years of Age and Older (NCT01949090), a randomized phase 2 study. The primary outcome is Number of Seroconverted (SCR) Subjects for Hemagglutination Inhibition (HI) Antibodies Against the Flu A/Mallard/NL/12/2000 (H7N1) Virus Strain.

2. Clinical Question

This is an immunogenicity question, not a question about prevention of influenza illness. The primary endpoint records whether an antibody response met a specified threshold after vaccination. The registry also lists secondary immunogenicity outcomes, including antibody titres, seroprotection and geometric mean fold rise, as well as safety outcomes. Those additional measures provide different descriptions of antibody responses and safety; they do not change what the primary seroconversion endpoint measures.

The trial record describes a placebo-controlled study of different vaccine formulations in older adults. It does not report a primary comparison of one vaccine formulation against another. The reported risk differences therefore answer the vaccine-versus-placebo question for each formulation, rather than establishing which formulation has the strongest response relative to the others.

3. Trial Design

The registry identifies NCT01949090 as a completed Phase 2 randomized, parallel-group study with quadruple masking. Its official title describes the study as observer-blind. These are the registry’s design descriptions; the record does not explain how the masking labels relate to each other. The study concerned different formulations of GSK Biologicals’ H7N1 influenza vaccine in adults 65 years of age and older, with placebo as the control.

The arm descriptions identify four vaccine groups, F1 through F4, and a placebo group. Participants in each described group received two intramuscular administrations, at Day 0 and Day 21, in the deltoid region. The non-dominant arm was used at Day 0 and the dominant arm at Day 21. The registry reports total enrollment of 363. The analysis denominators for the primary outcome differ by group and are reported separately: 56 for F1, 56 for F2, 55 for F3, 53 for F4 and 112 for placebo. These are the reported denominators for the Day 42 primary outcome comparisons, not a substitute for total enrollment.

The registry lists the study start as 2013-09-25 and primary completion as 2013-12-04, and gives the overall status as completed. It describes the study purpose as evaluating safety-related events and antibody immune responses to the different formulations. The registry record does not report the allocation ratio, the sequence-generation or concealment procedures, or further operational details about masking. It also does not report a sample-size justification or a detailed statistical analysis plan. Those design details should not be inferred from the randomized and masked labels alone.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Out of 363 subjects enrolled in the study, 3 subjects did not receive any vaccination.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The analysis was performed on the According-to-Protocol (ATP) cohort for immunogenicity at Day 42, which included all evaluable subjects for whom data concerning immunogenicity outcome measure were available. This primary outcome measure was centered on the groups that received GSK2789869A vaccine.

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 Number of Seroconverted (SCR) Subjects for Hemagglutination Inhibition (HI) Antibodies Against the Flu A/Mallard/NL/12/2000 (H7N1) Virus Strain, measured over At Day 42.

The primary endpoint is the number of participants with seroconversion of HI antibodies against the vaccine-homologous Flu A/Mallard/NL/12/2000 H7N1 strain at Day 42. HI refers to hemagglutination inhibition, an antibody assay. The endpoint turns an antibody measurement into a participant-level classification: whether the post-vaccination response meets the registry’s seroconversion definition.

For participants who were initially seronegative, the definition requires an antibody titre below 10 before vaccination and a post-vaccination titre of at least 40. For participants who were initially seropositive, defined as having a pre-vaccination titre of at least 10, the post-vaccination titre must be at least 4-fold the pre-vaccination titre. The endpoint therefore applies a different threshold according to baseline serostatus.

Seroconversion is used as an immunogenicity measure because it captures a defined change in antibody response, rather than simply reporting a group average titre. It does not itself measure influenza illness, infection, hospitalization or transmission. Nor does meeting the endpoint establish clinical protection. The registry lists other immunogenicity outcomes, but the primary result is the count meeting this seroconversion definition in the Day 42 According-to-Protocol immunogenicity cohort.

7. Results Posted in the Registry

• GSK2789869A F1 Group: 69.6% (39/56) versus 0.0% (0/112) on control; risk difference 69.6 percentage points (95% CI 57.6 to 81.7); unadjusted two-proportion z-test p = <0.001.
• GSK2789869A F2 Group: 80.4% (45/56) versus 0.0% (0/112) on control; risk difference 80.4 percentage points (95% CI 70.0 to 90.8); unadjusted two-proportion z-test p = <0.001.
• GSK2789869A F3 Group: 76.4% (42/55) versus 0.0% (0/112) on control; risk difference 76.4 percentage points (95% CI 65.1 to 87.6); unadjusted two-proportion z-test p = <0.001.
• GSK2789869A F4 Group: 86.8% (46/53) versus 0.0% (0/112) on control; risk difference 86.8 percentage points (95% CI 77.7 to 95.9); unadjusted two-proportion z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary at the participant level: each evaluable participant either met the seroconversion definition or did not. The registry reports the number meeting the definition and the analysis denominator for each group. The associated risk difference expresses the difference between the vaccine-group proportion and the placebo-group proportion. It is an absolute contrast on the proportion scale, rather than a relative measure such as a risk ratio.

For the placebo comparison, the registry reports a zero count among 112 participants. The vaccine-group counts and denominators are 39 of 56 for F1, 45 of 56 for F2, 42 of 55 for F3, and 46 of 53 for F4. The risk differences and confidence limits are reported in the same proportion scale. They should not be confused with antibody titres: the endpoint is a count of participants crossing the seroconversion threshold, while titres are continuous measurements and appear among the registry’s secondary outcomes.

A confidence interval describes a range of values compatible with the observed data under the statistical procedure used. It reflects sampling uncertainty, not uncertainty about whether the recorded counts were entered correctly, and it does not account automatically for every design or analysis limitation. The confidence limits reported for the risk differences give readers information about precision as well as the point estimates. The registry record does not identify the specific method used to calculate these intervals, so a particular interval procedure should not be attributed to the sponsor.

The posted p-values are labelled unadjusted and are <0.001 for each vaccine-versus-placebo contrast. A p-value is a measure of how unusual data at least as incompatible with a null hypothesis would be if that null hypothesis and the test assumptions held. It is not a measure of effect magnitude, and it cannot establish the probability that a treatment effect is real. The exact test used for these comparisons is not reported in the registry record. The reported p-values therefore should be quoted as posted rather than assigned to a specific test or recalculated using an assumed method.

General statistical context: confidence intervals and p-values answer related but different questions. An interval emphasizes a range of plausible effect estimates under a procedure, while a p-value assesses compatibility with a specified null value under a test. Neither measure resolves whether an endpoint is clinically important, whether missing observations could bias the result, or whether multiple comparisons were adequately addressed. Those questions require information about the clinical context and analysis plan.

Why does the analysis population matter?

The primary endpoint population is identified as the According-to-Protocol cohort for immunogenicity at Day 42. The description says it included evaluable subjects for whom immunogenicity outcome data were available. This is the population underlying the reported primary outcome denominators. It is not described as all randomized participants, and the registry does not provide the number excluded from this cohort, reasons for exclusion, or a comparison of included and excluded participants. The result should therefore be understood as an analysis of the stated evaluable cohort rather than as an automatically complete account of everyone enrolled.

General statistical context: restricting analysis to participants with evaluable data can affect the interpretation if availability is related to response, safety, adherence or other participant characteristics. An analysis population defined by protocol adherence may answer a different question from an analysis based on all randomized participants. The registry record does not report sensitivity analyses or methods for handling missing endpoint data, so the effect of those choices cannot be assessed from the posted primary results.

What can planning calculations tell readers?

The registry record includes assumptions for planning illustrations based on the observed effect: a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is described as a stated fraction of the observed effect. For time-to-event outcomes, the planning convention is total events; for other outcomes, it is participants per arm. These are planning assumptions, not evidence that the original trial was designed using those assumptions.

General statistical context: a sample-size illustration describes the amount of information a future study might plan for under specified assumptions. If based on an effect observed in a completed study, it depends on that estimate and on the chosen fraction of that effect. It is not observed power, does not establish that the completed study was adequately sized, and should not be read as a retrospective probability that the trial detected an effect. The original trial’s sample-size justification and planning assumptions are not reported in the registry record. The record also does not report interim analyses or a detailed statistical analysis plan.

These distinctions matter because the numerical results are interpretable only alongside the population, endpoint definition and comparison being reported. The record gives the endpoint counts, risk differences, intervals and unadjusted p-values, but it does not give enough information to reconstruct every analytical decision. General explanations of the statistical measures clarify their meaning; they do not fill in methods that the registry does not document.

10. Confidence Intervals and Precision

The point estimates for the vaccine-versus-placebo risk differences are 0.696 for F1, 0.804 for F2, 0.764 for F3 and 0.868 for F4. Their reported confidence intervals are 0.576 to 0.817, 0.700 to 0.908, 0.651 to 0.876, and 0.777 to 0.959, respectively. These intervals describe uncertainty around each vaccine-versus-placebo contrast, not around a comparison between vaccine formulations.

General statistical context: a confidence interval conveys precision by showing how much uncertainty remains around an estimate under the method used. A narrower interval indicates greater precision than a wider interval under comparable conditions, but interval width alone does not determine whether an effect is important. The interval is not a guarantee that the true value lies inside it, nor does it include uncertainty from unreported design or analysis choices.

The intervals are separated from zero as posted, and each corresponding unadjusted p-value is <0.001. This is consistent with the reported comparisons indicating differences from the placebo result under the statistical procedures used. However, the interval method and test method are not reported in the registry record, and no adjustment for the four comparisons is reported. The confidence intervals should not be treated as simultaneous, multiplicity-adjusted intervals.

Precision also depends on the analysis denominators and endpoint availability. The primary results use the stated According-to-Protocol cohort for immunogenicity, with denominators that vary by group. The registry does not report the uncertainty associated with any departures from the planned or randomized population, nor does it provide sensitivity analyses for missing data. Thus, the intervals quantify only the sampling uncertainty represented by the posted calculations, not every possible source of uncertainty in the study.

11. Multiplicity, Missing Data and Analysis Populations

The primary outcome is one seroconversion measure assessed against placebo for each of four vaccine formulations. The record reports four unadjusted p-values, each <0.001. It does not report a multiplicity adjustment, a prespecified testing hierarchy, or another method for controlling the chance of false-positive findings across the set of comparisons. The p-values should therefore remain identified as unadjusted; they should not be presented as if family-wise error or another multiplicity target had been controlled.

The registry lists 34 outcome measures and reports 20 posted secondary outcomes. The secondary outcomes include immunogenicity measures such as antibody titres, seroprotection, seroconversion and geometric mean fold rise, for vaccine-homologous H7N1 and vaccine-heterologous H7N9 strains. The registry record does not describe how these outcomes were prioritized for statistical testing, whether confirmatory testing was planned, or whether multiplicity was addressed across primary and secondary outcomes. The number of outcomes alone does not determine the appropriate adjustment; interpretation depends on the prespecified questions and analysis plan, which are not reported.

The primary analysis population is the According-to-Protocol cohort for immunogenicity at Day 42, described as evaluable subjects with available outcome data. The record does not report how many participants were not evaluable, why data were unavailable, or how missing values were handled. It also does not report an all-randomized analysis of the primary endpoint or sensitivity analyses for missing data. These omissions limit assessment of how robust the estimate is to the choice of analysis population and to incomplete outcome availability.

The registry record does not report interim analyses, detailed stopping rules, or a full statistical analysis plan. It also does not report a sample-size justification. Those items cannot be inferred from the reported enrollment or from planning illustrations based on observed effects. In general, prespecification helps distinguish confirmatory comparisons from exploratory analyses; without the relevant details in the record, the results are best interpreted as the posted endpoint estimates and unadjusted comparisons, with the analysis-plan uncertainties kept explicit.

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: angina pectoris (3.3%), coronary artery disease (3.3%), lung neoplasm malignant (3.3%), angina unstable (1.7%), arrhythmia (1.7%).

The registry reports serious adverse event counts and the number at risk by group. The counts are 3 among 60 for F1, 10 among 60 for F2, 7 among 60 for F3, 7 among 60 for F4, and 13 among 120 for placebo. The registry also reports deaths as 0 for F1, 1 for F2, 2 for F3, 0 for F4 and 0 for placebo. These are descriptive counts; they do not by themselves establish that an event was caused by the assigned intervention.

Serious adverse events are not the same endpoint as the primary antibody response. The registry lists serious-event terms including angina pectoris, coronary artery disease and lung neoplasm malignant, each reported as 3.3%, as well as angina unstable, arrhythmia and atrial fibrillation, each reported as 1.7%. The record excerpt does not provide a denominator or a group assignment for these term-level percentages, so they should not be used to infer which group had each term or to make between-group comparisons.

General statistical context: safety counts describe events observed during the reported follow-up and depend on how events were defined, collected and attributed. A count is not a causal assessment. The registry record does not provide further detail about event timing, duration, seriousness criteria, relatedness assessments or exposure-adjusted analyses. The counts should therefore be read as the posted safety information, separate from the immunogenicity findings and without a causal conclusion.

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

The primary results show that, in the reported Day 42 According-to-Protocol immunogenicity cohort, participants in each vaccine formulation group met the registry’s seroconversion definition more often than participants in the placebo group. The reported risk differences are positive for all four vaccine-versus-placebo comparisons, and each posted unadjusted p-value is <0.001. The confidence intervals show the uncertainty attached to those estimates under the calculation used for the posted results.

That finding is specific to a laboratory-defined antibody response against the vaccine-homologous H7N1 strain. It does not establish that participants were protected from influenza, that influenza cases were prevented, or that the vaccine formulations have a particular clinical effect. The primary endpoint is not a clinical illness endpoint. The registry record does not report illness outcomes as part of the primary result.

The results also do not establish a preferred formulation. Each reported contrast is a vaccine formulation versus placebo. The F1, F2, F3 and F4 estimates should not be treated as a direct ranking or as evidence that one formulation is superior to another. Such a conclusion would require a direct comparison, with an appropriate analysis and uncertainty, not just comparison of the separate point estimates.

The findings do not establish a complete safety profile. Serious adverse event counts and deaths are reported separately, but the counts alone do not determine causation or comparative safety. The registry’s record does not include all the detail needed to assess event attribution, missing safety information or the full context of follow-up.

In general, an observed difference can be statistically distinct from zero while its clinical meaning remains uncertain. Interpretation depends on the endpoint’s relevance, the prespecified analysis, precision, multiplicity and the completeness of the data. The registry supports a description of the observed immunogenicity contrast at Day 42; it does not resolve those broader clinical and methodological questions.

14. Why This Trial Matters Statistically

This trial illustrates how the choice of endpoint determines what a clinical study can answer. The primary measure is a participant-level seroconversion classification based on HI antibody titres, not influenza illness. Reporting both endpoint counts and vaccine-versus-placebo risk differences makes the contrast interpretable on the scale of proportions, while the confidence intervals show that the point estimates have uncertainty.

The record also demonstrates why comparison structure matters. There are four separate vaccine-versus-placebo contrasts, but those contrasts do not answer whether one formulation differs from another. A reader who ranks formulations by their point estimates would be making a different comparison from the one reported. The unadjusted p-values also require attention to multiplicity because several related contrasts are present and no adjustment is reported.

Population definition matters as well. The primary endpoint analysis used an According-to-Protocol immunogenicity cohort of evaluable participants with available data. That is important context for interpreting the reported denominators and for judging how far the estimates apply beyond the analyzed cohort. The registry does not report the missing-data details or sensitivity analyses needed to assess the effect of exclusions or incomplete outcome data.

Finally, the safety counts and immunogenicity results address distinct questions. Antibody response estimates cannot replace safety assessment, and adverse-event counts alone cannot establish causation. Read together, the design labels, outcome definitions, analysis population, uncertainty intervals and omissions in the registry record provide a useful example of why statistical interpretation should stay close to the exact question each reported measure answers.

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.

The primary endpoint results are reported for an According-to-Protocol immunogenicity cohort at Day 42. The registry describes this as including evaluable subjects with available immunogenicity outcome data, but does not report the number or characteristics of participants outside that cohort, reasons for missing or excluded data, or sensitivity analyses. The possible influence of those exclusions cannot be assessed from the record.

The primary result is an antibody response against the vaccine-homologous H7N1 strain. It is not a measure of influenza illness or clinical protection. The registry lists other immunogenicity outcomes, including a vaccine-heterologous H7N9 strain, but those outcomes do not broaden the meaning of the primary H7N1 endpoint. The posted primary comparison is formulation versus placebo; direct statistical comparisons among the four vaccine formulations are not reported.

The record presents unadjusted p-values for four vaccine-versus-placebo comparisons. It does not report a multiplicity procedure, detailed statistical analysis plan, or interim analysis information. The method used for the risk-difference confidence intervals and the specific test behind the p-values are not reported in the registry record. Consequently, the posted estimates can be described, but all analytic choices cannot be independently evaluated from the record.

The safety information is also limited in detail. Serious adverse event counts, deaths and selected serious-event terms are listed, but further event-level context and causal assessments are not reported in the information summarized. The record does not provide a sample-size justification, so enrollment alone cannot establish the study’s intended precision or ability to address every outcome. These limitations concern the information available in the registry record and should not be resolved by assumptions about the protocol or analyses.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Immunogenicity and Safety Study of Different Formulations of GlaxoSmithKline (GSK) Biologicals H7N1 Influenza Vaccine Administered to Adults 65 Years of Age and Older — NCT01949090.
• Registry record last updated 2017-11-17; status last verified 2017-11.
• Results first posted 2017-11-17.
• 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

NCT01949090 is a completed randomized, parallel-group Phase 2 study of four GSK2789869A H7N1 vaccine formulations and placebo in adults aged 65 years and older. The registry describes quadruple masking, while the official title calls the study observer-blind. Total enrollment was 363; the listed study start was 2013-09-25 and primary completion was 2013-12-04.

The primary endpoint was Day 42 seroconversion of HI antibodies against vaccine-homologous H7N1 in the According-to-Protocol immunogenicity cohort. The reported counts were 39 of 56 for F1, 45 of 56 for F2, 42 of 55 for F3, 46 of 53 for F4, and 0 of 112 for placebo. Risk differences versus placebo ranged from 0.696 to 0.868, with each unadjusted p-value <0.001. The record also reports serious adverse event counts and deaths, but does not report a multiplicity adjustment, detailed missing-data handling or a full statistical analysis plan.