Immuno & Safety Study of GSK Biologicals' Thio or Preservative Free Hepatitis B Vaccine in Subjects Aged 11-15 Yrs: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Immuno & Safety Study of GSK Biologicals' Thio or Preservative Free Hepatitis B Vaccine in Subjects Aged 11-15 Yrs (NCT00343915), a non-randomized phase 3 study. The primary outcome is Number of Subjects Seroprotected for Anti-hepatitis B Surface Antigen (Anti-HBs) Antibody..

2. Clinical Question

3. Trial Design

This was a Phase 3, non-randomized, unmasked, parallel study sponsored by GlaxoSmithKline. The registry lists enrollment of 267 and the condition as hepatitis B. The recorded study start was 2004-04-21, the primary completion date was 2008-01-10, and the status is completed. These design and status details describe the registry record; they do not establish which participants contributed to each reported outcome.

The comparison involved different vaccine schedules and formulations. In the 2-dose group, subjects received adult thiomersal-free HBV formulation at months 0 and 6 and physiological saline placebo at month 1. In the 3-dose group, subjects received paediatric preservative-free HBV formulation at months 0, 1 and 6. Because the allocation was non-randomized, the groups should not be treated as if assignment had created exchangeable groups. The record does not report a matching procedure or an adjusted comparative analysis.

The brief summary describes a long-term follow-up objective: assessing persistence of hepatitis B antibodies at 30, 42, 54 and 66 months after the first dose. It says subjects were aged 11 to 15 years at the primary vaccination course and 13 to 18 years at enrollment in the follow-up study. The listed primary endpoint, however, is seroprotection at Month 7. These are distinct time frames and should not be conflated: the Month 7 endpoint is the stated primary outcome, while the later timepoints belong to the follow-up objective.

The record lists secondary outcomes covering antibody titers, seroprotection, solicited symptoms, unsolicited adverse events and serious adverse events. It records 10 outcome measures and 0 analyses. The registry record does not report the statistical analysis plan, the reasons for the allocation approach, or how the target enrollment was selected.

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.

Primary Study

Reasons for not completing this period:

Month 30 Follow-up

Month 42 Follow-up

Month 54 Follow-up

Analysis population for the primary outcome, as defined in the registry: The ATP cohort for immunogenicity included all evaluable subjects who had post-vaccination immunogenicity results and who complied with the protocol, including the time schedule for vaccination and blood sample draw.

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 Seroprotected for Anti-hepatitis B Surface Antigen (Anti-HBs) Antibody., measured over At Month 7.

The primary endpoint is the number of subjects seroprotected for anti-hepatitis B surface antigen antibody at Month 7. A subject qualified as seroprotected if the anti-HBs concentration was at least 10 mIU/mL. This is a binary classification at the subject level: each evaluable subject either meets the stated threshold or does not. The endpoint is reported as a count, rather than as a continuous antibody concentration.

In general, a threshold-based seroprotection endpoint gives a clinically interpretable summary of whether measured antibody concentrations reach a prespecified level. It also loses information: two subjects on the same side of the threshold may have different concentrations, while subjects just on opposite sides are classified differently. Antibody titers are listed separately among the secondary outcomes, allowing concentration measurements to address a different question from the binary seroprotection count.

The primary endpoint's analysis population is described as the ATP cohort for immunogenicity. It includes evaluable subjects with post-vaccination immunogenicity results who complied with the protocol, including the vaccination and blood-sample timing schedule. The record does not provide the primary endpoint counts, group-specific proportions, uncertainty intervals, or a comparative test result. The Month 7 definition therefore describes what was to be measured, but does not by itself show the observed level of seroprotection in either group.

The follow-up objective concerns antibody persistence at 30, 42, 54 and 66 months after the first dose. Those later measurements are not the same as the stated Month 7 primary endpoint. The registry record does not report numerical persistence results at those later timepoints.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies a parallel, non-randomized study with no masking. A parallel design assigns participants to a group whose assigned schedule is followed over time, rather than switching each participant between schedules. In general, this structure makes between-group comparability important: the interpretation of a contrast depends on whether the groups differ in other ways as well as in the schedules being studied. Here, allocation was non-randomized, and the registry record does not report an adjustment method for group differences.

What does non-randomized allocation imply?

In general, randomization uses chance to assign treatment and can reduce systematic differences between groups. Non-randomized allocation does not provide that protection. If factors related to health, prior care, or enrollment influence which schedule a subject receives, those factors may also relate to measured outcomes. This is the general concern known as confounding. The registry record does not describe why allocation was non-randomized, what factors influenced it, or whether baseline characteristics were adjusted in an analysis. The two arms should therefore not be read as a randomized control comparison.

The record lists no masking. In general, masking can reduce the influence of treatment knowledge on subjective reporting or assessment. Its relevance varies by outcome: antibody laboratory measurements may be less open to subjective interpretation than symptom reports, while reports of solicited symptoms can be influenced by awareness of treatment. The registry record does not describe outcome-assessor procedures, laboratory masking, or other steps intended to limit ascertainment differences.

How is the primary endpoint classified?

The endpoint counts subjects with anti-HBs concentrations at or above the stated seroprotection threshold of 10 mIU/mL at Month 7. This converts a measured antibody concentration into a binary outcome. In general, binary outcomes can be described with counts and proportions, and comparisons may be expressed through a difference or ratio of proportions. Those are possible statistical summaries, not reported results in this record. No primary endpoint counts, denominators, effect estimate, confidence interval, or test result are reported in the registry record.

The analysis population is the ATP cohort for immunogenicity: evaluable subjects with post-vaccination results who complied with the protocol, including the vaccination and blood-sample timing schedule. In general, restricting an analysis to participants meeting evaluability and compliance requirements can make the analysis depend on post-allocation events. If compliance or outcome availability differs between groups, the analyzed groups may not preserve the comparability that an analysis of all assigned subjects might aim to retain. The registry record does not report the group-specific size of this cohort or how excluded subjects were handled.

What do uncertainty and statistical significance mean?

In general, a confidence interval describes a range of values compatible with the observed data under a statistical model and its assumptions. It conveys precision as well as direction: a wide interval indicates less precision than a narrow interval. A p-value, when calculated under a stated test and null hypothesis, describes how unusual data at least as incompatible with that hypothesis would be if the model assumptions and null hypothesis held. Neither measure replaces assessment of study design, bias, or the practical meaning of an outcome.

The registry record does not report a confidence interval or p-value for the primary endpoint, and it records 0 analyses. There is therefore no posted inferential result from which to assess statistical precision or evidence against a null hypothesis. A numerical comparison should not be inferred from the study's enrollment or from adverse-event counts.

Why can multiplicity matter?

The record lists multiple secondary outcomes, including antibody titers, seroprotection, solicited local and general symptoms, unsolicited adverse events and serious adverse events. The follow-up objective also names assessments at 30, 42, 54 and 66 months after the first dose. In general, testing many outcomes or timepoints creates more opportunities for chance findings. Multiplicity procedures can control the chance of false-positive conclusions across a defined family of tests, but they need to be specified in relation to the study's analysis plan. The registry record does not report a multiplicity adjustment, a hierarchy among secondary outcomes, or the intended handling of repeated measurements.

What does the planning illustration represent?

The sample-size planning assumptions described for the posted material are an illustration based on an observed effect, not the trial's own design. They specify a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect defined as a stated fraction of the observed effect. Such calculations can show how a hypothetical future study's size would depend on assumptions about the target effect and statistical operating characteristics. They are not observed power for this completed study and should not be read as evidence that the trial was adequately or inadequately powered. The registry record does not report its own sample-size justification or design assumptions.

General cautions also apply to calculations based on observed effects: an estimate from a study can be imprecise, and planning around it does not guarantee a future study will reproduce that effect. No numerical planning results are stated in the registry record. The illustrations cannot replace the missing primary endpoint results, and they do not resolve concerns about non-randomized allocation or missing analysis details.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("2-dose Engerix", "3-dose Engerix")
n      <- c(241, 113)
events <- c(233, 111)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  2-dose Engerix  241  233  96.7  93.6  98.6  93.6  98.3  93.6  98.3
ARM  3-dose Engerix  113  111  98.2  93.8  99.8  93.8  99.5  93.8  99.5

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 10 outcome measures, including the Month 7 seroprotection endpoint and secondary measures covering antibody titers, solicited symptoms, unsolicited adverse events and serious adverse events. Its brief summary also describes antibody persistence assessments at 30, 42, 54 and 66 months after the first dose. Considering multiple outcomes and timepoints can raise multiplicity concerns. In general, when many hypotheses are tested, some may appear notable by chance. Prespecified outcome hierarchies or statistical adjustments can help define how such findings should be interpreted. The registry record does not report a multiplicity strategy or testing hierarchy.

The primary endpoint population is the ATP cohort for immunogenicity: evaluable subjects with post-vaccination immunogenicity results who complied with the protocol, including the vaccination and blood-sample schedule. This is a defined analysis population, but the record does not give its size by group or explain how missing measurements and protocol deviations were handled. The number of participants contributing to the primary endpoint is not reported in the registry record.

Missing outcome data can affect both precision and interpretation. In general, missingness may be related to a participant's characteristics, treatment group, or outcome, and different assumptions about missing values can lead to different conclusions. The registry record does not report the amount or reasons for missing primary outcome data, an imputation method, or a sensitivity analysis. These details cannot be inferred from the enrollment figure.

The registry records 0 analyses. It does not report a statistical analysis plan, a method for comparing groups, or an adjustment for the non-randomized allocation. As a result, neither multiplicity handling nor the robustness of any primary estimate can be assessed from the registry record.

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: injury (0.8%), arthritis (0.4%), ileitis (0.4%), infection bacterial (0.4%).

The record reports serious adverse events in 4 of 257 subjects in the 2-dose Engerix group and 1 of 125 subjects in the 3-dose Engerix group. It reports other adverse events in 155 of 257 and 74 of 125 respectively, and 0 deaths in each group. These are counts with their listed at-risk denominators; they are not adjusted comparisons. The registry record does not provide a statistical test or uncertainty interval for these figures.

The listed serious-event terms are Injury at 0.8%, Arthritis at 0.4%, Ileitis at 0.4% and Infection bacterial at 0.4%. Those terms and percentages describe the entries in the record; they do not establish that an event was caused by vaccination. The record also lists secondary outcomes concerning solicited local symptoms, solicited general symptoms, unsolicited adverse events and serious adverse events, but the listed event counts alone do not describe severity, timing, duration or relatedness for every event.

In general, safety counts need to be interpreted with attention to definitions, observation period, completeness of follow-up and the population at risk. The registry record does not report enough detail to reconstruct those elements for every safety outcome. Because allocation was non-randomized, differences in counts should not be treated as a causal effect of one schedule versus the other.

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

The registry record documents a study intended to evaluate hepatitis B antibody persistence after different vaccine schedules in subjects aged 11 to 15 years at primary vaccination. It defines the primary outcome as the number of subjects reaching the anti-HBs seroprotection threshold at Month 7, and it describes follow-up assessments at 30, 42, 54 and 66 months after the first dose. This information clarifies the question the study set out to address and the population and schedules involved.

However, the primary endpoint results are not reported in the registry record. The record does not give group-specific seroprotection counts or denominators, an estimated comparison, a confidence interval, or a p-value. It therefore cannot establish from the posted results how many subjects met the Month 7 threshold or how precisely the two schedules differed on that outcome. Numerical results for the later persistence timepoints are also not reported in the registry record.

The selected safety entries report serious adverse-event counts of 4 of 257 and 1 of 125 in the 2-dose and 3-dose groups, other adverse-event counts of 155 of 257 and 74 of 125, and 0 deaths in each group. These data describe what the record lists; they do not, by themselves, identify a causal relationship or establish comparative safety. A zero count in the reported death entries does not describe outcomes beyond the recorded population and observation represented in the registry.

In general, a completed study record is not equivalent to a complete statistical report. Study design, endpoint definitions and counts can be informative, but conclusions about effect size and precision require the outcome results and analysis details. Here, the non-randomized allocation, absence of reported primary results, and lack of documented analysis methods restrict what can be concluded. The record supports a description of the intended comparison and selected safety counts, not a claim that one schedule is superior, equivalent, or causally safer than the other.

14. Why This Trial Matters Statistically

This study illustrates why an endpoint definition and a result need to be distinguished. The registry defines seroprotection by a clear antibody threshold and states when the primary outcome was assessed. Those details make the intended measurement understandable. But without the observed counts and denominators, the endpoint definition cannot tell readers how often the threshold was reached or how uncertain a comparison would be.

It also highlights the interpretive limits of non-randomized parallel groups. In general, a difference between groups may reflect treatment schedules, differences in the participants assigned to those schedules, or both. Without a reported adjustment strategy and relevant group characteristics, a numerical contrast would not automatically represent a causal vaccine effect. This is especially important when a comparison might otherwise be read as if the groups had been formed by random assignment.

The record's long-term objective and Month 7 primary endpoint further demonstrate the need to distinguish outcome timepoints. A study can describe a later follow-up aim without reporting results at those timepoints. Readers need the actual measurements, denominators and analysis to evaluate persistence and precision. The registry record does not report those numerical outcomes.

Finally, the safety counts show why denominators, event definitions and design context matter. Counts can describe events recorded in groups, but they do not alone establish causality or a comparative risk. Statistical interpretation depends not only on a calculated estimate, but also on how participants entered groups, which outcomes were available, and how analyses were planned and conducted. The record makes some of those design features clear while leaving key analysis and result details unreported.

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 most immediate limitation for interpretation is that the registry record does not report the primary Month 7 seroprotection counts or a comparative estimate. It also does not report confidence intervals, p-values or group-specific denominators for that endpoint. As a result, the magnitude and precision of the primary outcome cannot be assessed from the record.

Allocation was non-randomized, and the registry record does not describe how subjects entered the two groups or whether the groups were comparable at baseline. It does not report an adjustment strategy for confounding. Any between-group contrast would therefore require caution, even if the outcome estimate were available. The study was also unmasked according to the record, while details of outcome assessment and laboratory procedures are not reported.

The analysis population is restricted to evaluable subjects with post-vaccination immunogenicity results who complied with the protocol and timing requirements. The registry record does not report the number of subjects meeting these criteria by group, the amount or reasons for missing outcome data, or how missing observations were handled. The extent to which the analyzed population differs from the enrolled population cannot be determined.

The record lists multiple outcomes and later antibody persistence timepoints but does not report a multiplicity strategy or a hierarchy of analyses. It also records 0 analyses and does not provide a statistical analysis plan or sample-size justification. Although selected safety counts are reported, detailed event definitions, follow-up information and adjusted comparisons are not reported in the registry record. These limitations constrain interpretation of both immunogenicity and safety findings.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Immuno & Safety Study of GSK Biologicals' Thio or Preservative Free Hepatitis B Vaccine in Subjects Aged 11-15 Yrs — NCT00343915.
• Registry record last updated 2019-06-10; status last verified 2019-02.
• Results first posted 2009-03-31.
• 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

NCT00343915 is a completed Phase 3, non-randomized, unmasked, parallel study sponsored by GlaxoSmithKline. The record lists enrollment of 267, a start date of 2004-04-21 and a primary completion date of 2008-01-10. Subjects were aged 11 to 15 years at primary vaccination and 13 to 18 years at enrollment in the long-term follow-up study.

The groups received either 2 doses of adult thiomersal-free Engerix-B with placebo at month 1, or 3 doses of paediatric preservative-free Engerix-B. The primary outcome was the number of subjects seroprotected for anti-HBs antibody at Month 7, defined by a concentration of at least 10 mIU/mL. The follow-up objective named assessments at 30, 42, 54 and 66 months after the first dose.

Primary endpoint results and numerical antibody persistence results are not reported in the registry record. Selected safety counts are reported, but allocation was non-randomized and the record does not report a comparative analysis or statistical analysis plan.