Post-authorisation Passive Enhanced Safety Surveillance of Seasonal Influenza Vaccines: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Post-authorisation Passive Enhanced Safety Surveillance of Seasonal Influenza Vaccines (NCT02893878), a registered phase not stated study. The primary outcome is Number of Subjects Reporting European Medical Agency (EMA) Defined Adverse Events of Interest (AEIs) (Any) & Onset Dates of AEIs Using a Card Based-adverse Drug Reaction (ADR) Reporting System in the Vaccinated_Fluarix Tetra Group, on a Weekly Basis..

2. Clinical Question

The study was intended to assess the combination of card-based reporting and routine data in a 2016/17 pilot, and to identify additional data that might be needed for pharmacovigilance requirements. The registry record does not report a causal comparison of vaccine products. Its primary endpoint is framed as a participant count in the Fluarix Tetra group, not as an adjusted relative effect between groups. Weekly and cumulative outcomes are also listed for any AEI and selected AEI categories. The registry record does not report the detailed weekly outcome values in the registry record for this analysis.

3. Trial Design

This was a completed post-authorisation passive enhanced safety surveillance pilot in England. The registry lists GlaxoSmithKline as sponsor, an enrollment of 19334, a start date of 2016-09-12, and primary completion on 2016-11-24. The study aimed to improve the combination of card-based adverse drug reaction reporting and routine data for collecting events after seasonal influenza vaccination. Its stated context was EMA guidance and PRAC requirements, including identifying additional data that may be needed.

The listed intervention is vaccine safety surveillance. The three described groups consist of volunteers vaccinated with Fluarix Tetra, volunteers vaccinated with non-GSK influenza vaccine, and volunteers whose vaccine manufacturer was unknown. Each group description refers to approximately 10 volunteer practices. Allocation is not stated. The registry record cautions that the arms are not a like-for-like control comparison, so their event counts should be read as descriptive rather than as evidence from randomized assignment.

The primary outcome was the number of participants in the vaccinated Fluarix Tetra group reporting any EMA-defined AEI and onset dates using a card-based ADR reporting system. The timeframe was within 7 days post vaccination, anticipated to be between 01 September 2016 and 30 November 2016. Weekly safety cohorts were defined for calendar Week 38 to 48 and included subjects vaccinated the week before the week of interest who were registered during follow-up from vaccination up to 7 days post vaccination. The primary analysis included all GSK vaccinees because electronic health records did not encode whether a subject had received an ADR card.

The registry also lists weekly and cumulative outcomes for any AEI, fever or pyrexia, local symptoms, and general non-specific symptoms. The record lists 120 outcome measures, but the registry entries available do not give the detailed result series for these outcomes. It reports no statistical analyses. Details such as a prespecified statistical analysis plan, an interim-analysis plan, or a sample-size justification are not reported in the registry record.

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: The study included vaccinated subjects only.

Overall Study

Analysis population for the primary outcome, as defined in the registry: The Weekly safety vaccinated cohorts (for each calendar Week 38 to 48) included subjects who were vaccinated the week before the week of interest and were registered during the safety follow-up period (from vaccination up to 7 days post vaccination).

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 Subjects Reporting European Medical Agency (EMA) Defined Adverse Events of Interest (AEIs) (Any) & Onset Dates of AEIs Using a Card Based-adverse Drug Reaction (ADR) Reporting System in the Vaccinated_Fluarix Tetra Group, on a Weekly Basis., measured over Within 7 days post vaccination (anticipated to be between 01 September 2016 and 30 November 2016).

The primary endpoint counts participants in the vaccinated Fluarix Tetra group who reported any European Medicines Agency-defined adverse event of interest through a card-based adverse drug reaction reporting system. It also concerns onset dates. The defined AEI categories included fever or other febrile illness, local reaction, general reaction such as fatigue or myalgia, and other presentations reported following vaccination. The specified observation window was within 7 days post vaccination.

A participant count is a descriptive way to summarize how many people reported an event, rather than how many separate event episodes occurred. It can help describe the volume and timing of reports captured by a surveillance approach. It does not, on its own, establish that vaccination caused an event, quantify an individual participant’s probability of experiencing one, or compare vaccine products fairly. Those interpretations require appropriate denominators, ascertainment, and a design and analysis suited to the question.

The endpoint is relevant to the pilot’s stated purpose because it focuses on whether the reporting and routine-data approach can collect AEI information and onset dates. However, the primary analysis included all subjects vaccinated with GSK vaccine rather than only subjects known to have received the ADR card: card receipt was not encoded in the electronic health record. The registry record does not report detailed weekly primary endpoint results in the registry record for this analysis. It also does not report an estimate of sensitivity, completeness, or agreement between the card system and routine data.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a participant count tell the reader?

A count reports the number of participants meeting an outcome definition. In this study, the primary outcome is a count of people reporting any EMA-defined AEI in the Fluarix Tetra group, along with onset dates. A participant count is not automatically a rate or a risk. To interpret a rate or risk, the outcome definition, observation period, eligible population, and denominator must align. Counts can also depend on reporting behavior and data capture, which are central concerns in a surveillance pilot.

The registry separately describes people at risk and people affected by serious and other adverse events for each vaccine group. Those summary counts should be distinguished from the primary endpoint wording. The detailed weekly primary endpoint series are not reported in the registry record details. In particular, the summary adverse-event counts should not be assumed to provide every weekly count, onset date, or AEI category required by the primary outcome description.

Why does the group structure matter?

Randomization aims to make comparison groups comparable, on average, with respect to factors other than the assigned intervention. This study’s allocation is not stated, and the registry cautions that the arms are not a like-for-like control comparison. Therefore, differences in the observed counts could reflect group composition, exposure or follow-up, reporting, or other factors, rather than a vaccine effect. The groups are useful descriptive categories for the surveillance report, but they do not by themselves create a controlled comparison.

The unknown-vaccine group has an explicitly uncertain product classification. Keeping it as a distinct descriptive category preserves that uncertainty; assigning its members to either known-product group would require information not reported in the registry record. Likewise, the primary endpoint is focused on the Fluarix Tetra group, while the adverse-event summary also lists non-GSK and unknown-vaccine groups. These different roles should not be conflated.

What is the role of an observation window?

The primary outcome specifies events within 7 days post vaccination. The weekly safety cohorts cover calendar Week 38 to 48, selecting subjects vaccinated the week before the week of interest and registered during follow-up from vaccination through 7 days post vaccination. A defined time window helps establish which events fall within the stated surveillance period. It does not establish that all events were captured, that participants had identical observation, or that events were caused by vaccination. Details of completeness of follow-up and event ascertainment are not reported in the registry record.

How should uncertainty be described?

A confidence interval is a statistical range that expresses uncertainty around an estimated quantity under a specified method and its assumptions. It is not a guarantee that the true value lies within that interval. The registry entries available for this analysis do not report confidence intervals or other uncertainty estimates. Without an interval or an inferential model, the reported counts alone do not show how precisely a rate, risk, or group contrast has been estimated.

The registry lists 0 statistical analyses. This means no inferential analysis is reported in the record details; it does not mean that an event count is itself a comparative test. No p-values, adjusted estimates, or model-based comparisons are reported. There is also no reported specification for how potential confounding, practice-level clustering, or reporting differences were handled. These are important considerations in general when data arise from groups recruited through practices and outcomes depend on reporting.

Why might over-dispersion matter in general?

General explanation: when counts vary more widely than a statistical model assumes, standard errors based on that model may be too small and confidence intervals too narrow. This is called over-dispersion. It can arise, for example, when observations are clustered or when participants have different underlying event rates. This general issue is relevant to count data, but the registry record does not report a count model, a dispersion assessment, or any adjustment for clustering. No conclusion about over-dispersion in these data can therefore be drawn.

What do planning illustrations mean?

General explanation: a sample-size calculation estimates how many participants might be needed under stated assumptions to detect a specified effect with a chosen power and significance level. It is a planning exercise, not a measure of how much power the completed study actually had. The stated illustration assumptions use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the target effect expressed as a fraction of the observed effect. Any sample sizes are planning illustrations, never observed power. The registry record does not report a trial-specific sample-size justification, and no inference about the adequacy of the completed enrollment follows from an illustration.

More generally, a sample-size result depends on the outcome scale, expected event frequency or variability, target effect, allocation, and analysis method. Changing those assumptions changes the planning estimate. Since the registry entries do not provide a protocol rationale or a complete statistical plan, the illustrations should not be treated as proof that the study was designed to test a particular comparison. They are separate from the descriptive counts and do not supply missing inferential results.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 120 outcome measures. The secondary outcomes include weekly and cumulative counts of participants reporting any EMA-defined AEI, fever or pyrexia, local symptoms, and general non-specific symptoms, with onset dates by vaccine group. Repeated outcomes across time and symptom categories can create multiple opportunities to observe patterns. In general, if many hypotheses are tested, the chance of at least one apparently notable result can increase unless analyses account for multiplicity or are interpreted as exploratory.

The registry entries do not report a multiplicity strategy, a hierarchy among secondary outcomes, or a prespecified approach for selecting which comparisons to emphasize. They also report 0 statistical analyses, so no adjusted or unadjusted inferential results are available in the registry record. This does not establish whether analyses were conducted outside the registry; those details are not reported in the registry record.

Missing-data handling is also not reported in the registry record. The registry entries do not state how incomplete follow-up, missing onset dates, absent event reports, or unknown vaccine manufacturer were handled analytically. The unknown-manufacturer group is described as its own vaccinated group, but no additional rules for missing classification are reported. The registry entries do not say whether any participant-level data were excluded or how the analysis treated records lacking required outcome information.

The primary analysis included all subjects vaccinated with GSK vaccine, because receipt of an ADR card was not encoded in the electronic health record. Thus, the endpoint’s card-based reporting system and the defined analysis population do not map to a known set of card recipients in the record. The registry does not report a separate analysis restricted to confirmed ADR-card recipients or describe how data sources were linked. Details of analysis populations beyond the stated primary group and weekly cohort definition are not reported in the registry record.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent non-serious adverse event terms

The registry’s adverse-event summary lists serious events as a count of participants affected and gives 0 for the Fluarix Tetra group, 0 for the non-GSK group, and 0 for the unknown-vaccine group. The corresponding counts at risk are 13861, 2295, and 3178. The summary reports no serious-event terms. These entries describe what is recorded in the registry summary; they do not show that the underlying risk is zero or that no serious event could occur.

The same summary lists other-adverse-event affected counts of 1049 for the Fluarix Tetra group, 62 for the non-GSK group, and 54 for the unknown-vaccine group. They should be read as reported counts, not as a comparison of product safety. The record does not provide confidence intervals, adjusted estimates, or a like-for-like design, and the groups differ in size. The registry’s explicit warning that these are not like-for-like control arms is especially important for interpreting the counts.

The primary endpoint describes any AEI reports, whereas the adverse-event summary uses serious and other adverse-event categories. The registry entries do not establish that the other-adverse-event affected counts are identical to the primary endpoint counts or give weekly timing for those summary figures. No death counts are reported in the adverse-event summary. Absence of reported serious events in this summary should therefore be interpreted as a registry finding, not as proof of absence of serious risk or evidence comparing vaccines.

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

The posted information shows that the registry describes a completed pilot focused on post-vaccination safety surveillance and reports adverse-event counts in three vaccinated groups. It documents a primary endpoint based on participants reporting any EMA-defined AEI and onset dates within 7 days post vaccination. It also describes weekly and cumulative outcomes for several AEI categories. The purpose was to improve the combined use of a card-based ADR system and routine data and to identify additional data that may be needed.

The record does not establish that a particular vaccine caused or prevented an event. Participants were not described as randomized, allocation is not stated, and the registry cautions that groups are not like-for-like controls. The results available for this analysis contain event counts but no inferential analyses, confidence intervals, p-values, or adjusted comparisons. The counts do not quantify uncertainty around a comparative effect.

The registry’s adverse-event summary reports 0 serious-event affected participants in each of the three groups, alongside the stated at-risk counts. This is a description of the registry entry and should not be interpreted as proof that serious events cannot occur. The other-adverse-event counts are also descriptive and do not demonstrate a product difference. Their relationship to the primary endpoint’s specific AEI definition is not stated in the registry record.

The pilot’s value is therefore principally in its surveillance question: whether card-based reporting and routine data can be combined to collect information relevant to post-vaccination safety monitoring. The registry record does not report completeness or accuracy measures for that data collection approach, the detailed weekly endpoint results, or a validated comparison between the sources. Further claims about measurement performance or comparative safety are not supported by the information reported in the registry record.

13. Why This Trial Matters Statistically

Safety surveillance depends not only on which events occur, but also on whether the data systems can capture them with usable timing and classification. This pilot directly concerned the combination of a card-based ADR reporting system and routine data. Its stated aim was to improve that combination and identify additional data that might be needed for pharmacovigilance requirements. That makes the data-collection process itself a central statistical issue, rather than treating the reported event counts as a complete answer about vaccine effects.

The endpoint includes onset dates and a defined post-vaccination window, while the secondary outcomes describe weekly and cumulative counts across AEI categories. In principle, such information can help characterize when events are reported and how reporting varies by category. However, interpretation relies on consistent definitions, complete observation, and clarity about which participants had access to or received the reporting card. The registry specifically notes that receipt of the ADR card was not encoded in the electronic health record, and the primary analysis included all GSK vaccinees.

The design also illustrates a general distinction between descriptive surveillance and causal comparison. Categorizing participants by vaccine received can organize event counts, but it does not create randomized groups or remove differences between groups. The record cautions against treating these arms as like-for-like controls. Statistical readers should therefore separate evidence about the reporting process from evidence about comparative product safety.

Finally, the record reports 120 outcome measures but no statistical analyses. That combination makes transparent outcome definitions and multiplicity handling particularly relevant to interpretation. The registry entries leave these analytic details unreported. The study matters statistically because it highlights how surveillance design, data linkage, participant classification, follow-up, and reporting completeness shape what event counts can support.

14. 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 limitation for interpretation is that the study is not described as randomized. Allocation is not stated, and the registry warns that the vaccinated groups are not like-for-like control arms. Differences in observed counts therefore cannot be attributed to vaccine product on the basis of these group labels. The unknown-vaccine group adds a further classification limitation: its members’ vaccine manufacturer is not known in the registry description.

The primary endpoint refers to card-based reports, but the primary analysis included all GSK-vaccinated subjects because receipt of the ADR card was not encoded in the electronic health record. The registry entries do not report whether card receipt was available through another source, how the card and routine data were linked, or whether a card-recipient-only analysis was performed. This makes it difficult to assess how closely the endpoint reflects reporting among known card recipients.

The available record details do not include detailed weekly results for the primary outcome, onset-date distributions, confidence intervals, p-values, adjusted comparisons, or other inferential estimates. Although the registry lists 120 outcome measures, it reports 0 analyses in the registry record. Multiplicity handling, missing-data rules, follow-up completeness, and the treatment of duplicate reports are not reported in the registry record.

The adverse-event summary provides affected counts and counts at risk, but it does not report a comparative model or explain how the summary categories map onto the primary AEI definition. It also does not provide details of event adjudication or data-source validation. As a pilot about surveillance methods, the study’s stated purpose is distinct from establishing vaccine safety differences; the registry record does not provide enough analytical detail to assess the completeness or comparative validity of event ascertainment.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Post-authorisation Passive Enhanced Safety Surveillance of Seasonal Influenza Vaccines — NCT02893878.
• Registry record last updated 2019-02-11; status last verified 2018-09.
• Results first posted 2019-02-11.
• 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.

18. Record Summary

NCT02893878 is a completed post-authorisation passive enhanced safety surveillance pilot of seasonal influenza vaccines in England, sponsored by GlaxoSmithKline. The registry reports enrollment of 19334, a start date of 2016-09-12, and primary completion on 2016-11-24. Its stated aim was to combine card-based ADR reporting with routine data for collecting post-vaccination events and identify additional data needs.

The primary endpoint counts participants in the vaccinated Fluarix Tetra group reporting any EMA-defined AEI and onset dates within 7 days post vaccination. The described vaccine groups are Fluarix Tetra, non-GSK, and unknown manufacturer; allocation is not stated, and the registry says the groups are not like-for-like controls. The adverse-event summary reports 0 serious-event affected participants in each group and other-adverse-event counts of 1049, 62, and 54, respectively.

The registry lists 120 outcome measures and 0 statistical analyses. Detailed weekly results, inferential estimates, multiplicity and missing-data methods, and a trial-specific sample-size justification are not reported in the registry record.