Safety and Immunogenicity Study of BCG, H4:IC31, and H56:IC31 Revaccination in Healthy Adolescents: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Immunogenicity Study of BCG, H4:IC31, and H56:IC31 Revaccination in Healthy Adolescents (NCT02378207), a randomized phase 1 study. The primary outcome is Number of Participants With Adverse Events.

2. Clinical Question

3. Trial Design

The registry identifies NCT02378207 as a randomized, parallel-group phase 1 trial. Its official title describes it as a phase 1b, placebo-controlled, partially blinded study; the registry’s masking entry says quadruple. These are the descriptions recorded, and the registry record does not explain how the masking labels relate to one another. The trial is marked completed.

Enrollment was 84: 24 participants in each vaccine group and 12 in the sodium chloride control group. H4:IC31 and H56:IC31 were scheduled for intramuscular administration on Days 0 and 56, while BCG was scheduled for Day 0. The control regimen was scheduled on Days 0 and 56. The record identifies Aeras as sponsor.

The primary outcome counted participants with solicited and unsolicited adverse events, including serious adverse events, after vaccination, over up to 8 months. Other stated aims included assessing humoral responses, innate and adaptive immune responses, changes in innate cells, non-classical MHC-restricted T-cell responses, and QFT-GIT and ESAT-6 Free IGRA discordance and conversion or reversion. These are listed as secondary outcome areas; detailed results for those outcomes are not reported in the registry record.

The brief summary describes the study as intended to help identify assays and immune responses for later evaluation as correlates of risk and protection, and as complementary to a separate ongoing study. That context frames the trial as an early investigation of safety and immunogenicity, rather than a direct test of protection against tuberculosis.

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

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 Participants With Adverse Events, measured over Up to 8 months.

The primary outcome is the number of participants who experienced adverse events after vaccination, with solicited and unsolicited events and serious adverse events included. The timeframe is up to 8 months. The reported values are participant counts: 22 in the H4:IC31 group, 20 in the H56:IC31 group, 24 in the BCG group and 10 in the sodium chloride control group. The corresponding group sizes were 24, 24, 24 and 12.

Counting participants with an event is a binary outcome: each participant is counted according to whether at least one event meeting the outcome definition was recorded. It is not a count of the total number of individual events, so a participant with multiple events does not necessarily contribute multiple times to this endpoint. The registry record does not report further detail about how event episodes were grouped or how the solicited and unsolicited categories were operationalized.

Adverse-event monitoring is relevant to a safety-focused early-phase trial because it describes events observed after the regimens were administered. It does not, by itself, establish that a vaccine caused an event. A randomized control group can provide context for comparing event counts, but the small group sizes and the broad endpoint mean that such comparisons remain uncertain. The endpoint also does not measure vaccine-induced protection against tuberculosis.

7. Results Posted in the Registry

• Group 1 H4:IC31: 91.7% (22/24) versus 83.3% (10/12) on control; risk difference 8.3 percentage points (95% CI -15.5 to 32.1), risk ratio 1.10 (0.83–1.46); unadjusted two-proportion z-test p = 0.453.
• Group 2 H56:IC31: 83.3% (20/24) versus 83.3% (10/12) on control; risk difference 0.0 percentage points (95% CI -25.8 to 25.8), risk ratio 1.00 (0.73–1.36); unadjusted two-proportion z-test p = 1.000.
• Group 3 BCG (2-8 x 105 CFU): 100.0% (24/24) versus 83.3% (10/12) on control; risk difference 16.7 percentage points (95% CI -4.4 to 37.8), risk ratio 1.20 (0.93–1.55); unadjusted two-proportion z-test p = 0.040.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a binary participant-level measure: whether a participant was recorded as having an adverse event during the stated follow-up. For a binary outcome, a risk is the number of participants with the event divided by the number at risk. The posted group counts are the numerator and the group sizes are the denominator. The comparisons express group differences and ratios, but the registry record does not give a detailed analysis plan explaining the exact estimation or testing procedures.

What does the risk difference describe?

A risk difference describes the absolute separation between two event risks. It is expressed on the probability scale, rather than as a ratio. A value near zero indicates that the observed group risks are similar on this scale; a positive or negative value indicates the direction of the observed contrast, according to the comparison’s ordering. Because this is a participant-level event outcome, the risk difference concerns the proportion of participants with at least one qualifying event, not the total number of event episodes.

The posted risk differences are 0.083 for H4:IC31 compared with control, 0.0 for H56:IC31 compared with control, and 0.167 for BCG compared with control. These are estimates from groups with 24 participants in each vaccine arm and 12 in control. With group sizes of this scale, a small change in the number of participants counted can materially affect the estimated contrast. The confidence intervals communicate this uncertainty more directly than the point estimates alone.

How do risk ratios and odds ratios differ?

A risk ratio compares the event risk in one group with the event risk in another. A ratio of 1 indicates equal risks, while values on either side of 1 indicate a relative difference. An odds ratio compares event odds rather than event risks. When events are common, as they are in these posted group counts, odds ratios can be farther from 1 than risk ratios for the same underlying comparison. The two measures answer related but distinct questions and should not be treated as interchangeable.

The relative-effect intervals are wide. For H4:IC31, the posted risk ratio is 1.1 with an interval from 0.831 to 1.456; its odds ratio is 2.2 with an interval from 0.27 to 17.92. For H56:IC31, the risk ratio is 1.0 with an interval from 0.734 to 1.363 and the odds ratio is 1.0 with an interval from 0.156 to 6.42. The risk ratio for BCG is 1.2 with an interval from 0.932 to 1.546. Each of these intervals includes the value corresponding to no relative difference.

What does a confidence interval convey?

A confidence interval is a range of values compatible with the observed data under the method used to construct it. It shows how much uncertainty surrounds an estimate. A narrow interval indicates greater precision, while a wide interval indicates that the data are compatible with a broader range of effects. The interval is not a guarantee that the true effect lies within its endpoints, nor does it give the probability that a particular effect is true.

For risk differences, zero represents no group difference; for risk ratios and odds ratios, 1 represents no relative difference. The posted risk-difference intervals include zero for all three comparisons. This indicates that the estimates do not distinguish a nonzero difference from no difference at the interval’s stated level. The BCG unadjusted p-value is 0.04, while its risk-difference interval extends from -0.044 to 0.378. The registry record does not report enough method detail to explain the relationship between these summaries, so neither should be substituted for the other.

What does over-dispersion mean in general?

Over-dispersion is a general issue for count outcomes when observed variation exceeds what a statistical model assumes. It matters because a model that understates variation can produce overly narrow intervals and overly small p-values. The primary outcome in this trial is a count of participants with an event, not a recurrent-event count, and no over-dispersion parameter is reported in the registry record. More generally, if a study analyzes multiple event episodes per participant, the dependence among events from the same participant must be handled by the analysis method; treating them as unrelated observations can misstate uncertainty.

The registry record lists seven outcome measures and five posted secondary outcome areas, but no detailed analysis population definitions, missing-data approach, or adjustment procedures. It reports no formal analysis details for these comparisons. Interpretation should therefore stay close to the posted counts and estimates rather than assuming a particular model, covariate adjustment, or protocol-defined hierarchy.

10. Confidence Intervals and Precision

Precision is limited by the small group sizes and is visible in the width of the confidence intervals. The H4:IC31 risk-difference interval runs from -0.155 to 0.321 around an estimate of 0.083. For H56:IC31, the interval runs from -0.258 to 0.258 around an estimate of 0.0. For BCG, it runs from -0.044 to 0.378 around an estimate of 0.167. Each range includes zero, so the posted data are compatible with no difference as well as differences in either direction.

The relative-effect intervals also illustrate uncertainty. The H4:IC31 odds-ratio interval, from 0.27 to 17.92, covers a particularly broad range of relative effects. The H56:IC31 odds-ratio interval, from 0.156 to 6.42, is also broad. The risk-ratio intervals for H4:IC31, H56:IC31 and BCG include 1. Taken together, these ranges do not support fine distinctions among the groups based on the primary outcome.

A confidence interval gives a range of effects consistent with the data and the method used; it is not a range containing every plausible value with certainty. The endpoints are affected by sample size and by the observed event counts. With only 12 participants in control and 24 in each vaccine group, the estimates can be sensitive to a small number of participants. The registry record does not report the detailed interval-construction methods, so the posted endpoints should be interpreted as the reported uncertainty summaries rather than as a basis for reconstructing the analysis.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are an illustration based on the observed H4:IC31-versus-control effect, not the original trial design. They are participant counts per arm, not total events. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set at the stated fraction of the observed effect.

The 80% power column gives the number of participants per arm required under those assumptions for an 80% chance of detecting the specified effect if that effect were the true one. The 90% power column gives the corresponding per-arm count for a 90% chance. These are design-planning quantities, not predictions about what happened in the completed trial. They are not observed power.

For planning for 100% of the observed effect, the figures are 247 participants per arm for 80% power and 329 per arm for 90% power. For planning for 75% of the observed effect, they are 470 and 628 participants per arm. For planning for 50% of the observed effect, they are 1,126 and 1,506 participants per arm. The pattern illustrates a general principle: smaller effects are harder to distinguish from random variation, so detecting them with the same target power requires more participants. These numbers are conditional on the listed assumptions and on using the observed effect as the planning reference; they do not establish what the trial’s original sample-size justification was, which is not reported in the registry record.

The planning curve represents the same general relationship between effect size and required sample size: as the target effect becomes smaller, the required number of participants rises. Its values should be read as planning illustrations rather than as observed trial outcomes or evidence that the completed study had a particular power.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.833; treatment proportion p₁ = 0.917; average p̄ = 0.875.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.4677 = 0.9167.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.4640 = 0.3905.
4. n = (0.9167 + 0.3905)² / (0.0833)² = 246.1, rounded up to 247 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.833333333333333                          # control proportion
p1 <- 0.916666666666667                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(2378207)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 247 (80% power), 329 (90% power)
 75% of observed effect: n per arm = 470 (80% power), 628 (90% power)
 50% of observed effect: n per arm = 1,126 (80% power), 1,506 (90% power)
power.prop.test: 246.1 
Simulated power at n = 247 per arm: 0.807 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Multiplicity, Missing Data and Analysis Populations

The registry record does not report whether the trial specified a primary comparison hierarchy, whether adjustment for multiple comparisons was planned, or how the several vaccine-versus-control comparisons were to be interpreted together. The displayed p-values are unadjusted. When multiple comparisons are made, the chance of at least one apparently unusual result can increase; multiplicity procedures are used in some analyses to control that risk. Without a reported plan, the p-value of 0.04 for BCG should be treated as an unadjusted result rather than as a definitive finding.

The study also had several secondary aims, including humoral responses, innate and adaptive immune responses, changes in innate cells, non-classical MHC-restricted T-cell responses, and IGRA discordance and conversion or reversion. The record lists these aims but does not report detailed results for them in the information available. It also does not report whether secondary analyses were adjusted for multiplicity or how they were prioritized relative to one another.

Analysis populations are not defined in the registry record. It is not reported whether the primary analysis included all randomized participants, only participants who received an intervention, or another defined group. The record also does not describe how participants with incomplete follow-up, missing outcome information, or deviations from scheduled administration were handled. Missing-data methods can affect estimates when event status is not known for every participant, but no such method should be assumed.

Likewise, the record does not report whether the comparisons were adjusted for baseline characteristics or other covariates, or whether any interim analyses were planned or conducted. These omissions limit the ability to assess the prespecified statistical framework and to distinguish planned analyses from exploratory ones. They do not change the arithmetic meaning of the posted group counts, but they matter when interpreting p-values, uncertainty intervals, and the emphasis placed on individual comparisons.

14. Safety Results

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

Deaths

Other (non-serious) adverse events

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

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: impetigo (4.2%), road traffic accident (4.2%), cardiac murmur functional (4.2%).

Serious adverse events were reported for 0 of 24 participants in the H4:IC31 group, 1 of 24 in the H56:IC31 group, 1 of 24 in the BCG group and 0 of 12 in the sodium chloride control group. No deaths were reported in any group. The registry also lists Impetigo, Road traffic accident and Cardiac murmur functional among the top serious-event terms, each at 4.2%.

These figures count participants at risk and participants affected; they do not describe the total number of all adverse-event episodes. A serious adverse event is not necessarily caused by the intervention, and the registry record does not establish attribution for the listed events. The terms and counts should therefore be read as reported safety observations during the trial, rather than as proof of a causal relationship.

The numbers of participants in the groups are small, so zero reported serious events in a group does not establish that the risk is absent. Similarly, one participant with a serious event cannot by itself establish that a regimen has a higher serious-event risk. The data provide a descriptive account of serious events in this study, with limited precision for comparing groups.

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

The posted results describe adverse events reported during follow-up in a randomized phase 1b study of healthy adolescents. Most participants in each group were counted as having at least one adverse event under the primary outcome definition. The counts were 22 of 24 for H4:IC31, 20 of 24 for H56:IC31, 24 of 24 for BCG and 10 of 12 for sodium chloride control. Serious adverse events affected 0, 1, 1 and 0 participants in those groups, respectively, and no deaths were reported.

The comparisons do not give a precise estimate of differences between regimens. All three risk-difference confidence intervals include zero, and the relative-risk intervals include 1. The intervals are broad enough to accommodate materially different effects in either direction. The BCG comparison’s unadjusted p-value of 0.04 is one part of the statistical record, but it does not override the uncertainty in the interval or the absence of reported multiplicity methods.

These findings do not establish that any regimen prevents tuberculosis, nor do they show that the regimens cause the adverse events recorded. The primary outcome is an adverse-event count at the participant level, not an efficacy endpoint. The trial’s stated interest in immune responses and potential future correlates of risk or protection is not itself evidence that those immune responses predict protection. Detailed results for the stated secondary outcome areas are not reported in the registry record.

The study can be read as an early safety and immunogenicity investigation with a control group and several vaccine regimens. Its small group sizes make the estimates uncertain, and the record does not contain enough information about the analysis plan, missing data or analysis populations to reconstruct all intended inferential decisions. The appropriate statistical conclusion is descriptive: the posted counts and estimates summarize the observed data, while their uncertainty limits strong claims about comparative differences.

16. Why This Trial Matters Statistically

This trial is statistically interesting because it places several vaccine regimens alongside a sodium chloride control while measuring a broad participant-level safety outcome. The primary endpoint is easy to state, but it compresses different kinds of adverse events into a single indicator of whether a participant had an event. That simplicity aids a group-level comparison while concealing variation in event type, timing, severity and number of episodes unless those details are examined separately.

The allocation was unequal: each vaccine group included 24 participants and the control group included 12. The resulting group counts can describe what occurred, but the small numbers limit precision. The confidence intervals reflect that limitation: each risk-difference interval includes no difference and allows for effects on either side. This is a useful reminder that a point estimate is not a precise answer merely because it is expressed with several decimal places.

The BCG comparison also illustrates why p-values should be read alongside effect estimates, intervals and the analysis context. Its unadjusted p-value is 0.04, while its interval includes zero. The registry record does not report multiplicity procedures or enough detail about the inferential methods to resolve the interpretation further. A single threshold-based reading would miss the broader uncertainty and the multiple comparisons.

Finally, the stated aims extend beyond the primary safety count to immune-response measurement and exploration of potential future correlates. Keeping those aims distinct matters: an immune response measured in an early-phase study is not automatically a validated marker of protection. The registry record gives the study a role in generating and evaluating candidate responses, while the posted primary outcome supports only a limited description of adverse events.

17. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several features of the record constrain interpretation of this trial’s results. The groups were small, with 24 participants in each vaccine group and 12 in control. The confidence intervals are broad, so the estimates do not sharply distinguish among the regimens. The primary endpoint also combines solicited and unsolicited adverse events, including serious adverse events, into a participant count. The registry record does not provide a fuller event-level breakdown for the primary outcome.

The record does not report detailed eligibility criteria beyond the description of healthy, HIV-1-uninfected adolescents in the official title, nor does it give demographic characteristics or baseline group summaries. It does not define the analysis population, explain missing-data handling, report protocol deviations, or describe whether the comparisons were adjusted for baseline factors. These details can affect how representative and comparable the analyzed groups are, but should not be inferred.

The registry identifies randomized allocation and quadruple masking, while the official title describes the trial as partially blinded. The registry record does not clarify the difference between those descriptions. It also does not report detailed statistical methods for confidence intervals and tests, an adjustment for multiple comparisons, interim analyses, or the trial’s original sample-size justification. The planning illustration is separate from the original design rationale.

Finally, the stated secondary aims are broader than the posted primary result. Detailed outcomes for humoral, innate, adaptive and non-classical MHC-restricted responses, as well as IGRA discordance and conversion or reversion, are not reported in the registry record. Conclusions about those immune outcomes, or about protection against tuberculosis, cannot be drawn from the adverse-event counts.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Safety and Immunogenicity Study of BCG, H4:IC31, and H56:IC31 Revaccination in Healthy Adolescents — NCT02378207.
• Registry record last updated 2019-11-18; status last verified 2018-01.
• Results first posted 2019-11-18.
• 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.

21. Record Summary

NCT02378207 is a completed randomized parallel-group phase 1 study sponsored by Aeras, with 84 participants. Its official title describes a phase 1b, placebo-controlled, partially blinded trial in healthy, HIV-1-uninfected adolescents; the registry lists quadruple masking. Participants received H4:IC31, H56:IC31, BCG or sodium chloride 0.9% control.

The primary outcome counted participants with solicited or unsolicited adverse events, including serious adverse events, over up to 8 months. The counts were 22 of 24, 20 of 24, 24 of 24 and 10 of 12, respectively. Serious adverse events affected 0, 1, 1 and 0 participants, with no deaths reported. Posted group comparisons have broad confidence intervals, and detailed analysis methods, multiplicity procedures and analysis populations are not reported in the registry record.