Efficacy Study of COVID-19 mRNA Vaccine in Regions With SARS-CoV-2 Variants of Concern: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy Study of COVID-19 mRNA Vaccine in Regions With SARS-CoV-2 Variants of Concern (NCT05168813), a randomized phase 2, phase 3 study. The primary outcome is Part A: Number of Participants With a First Occurrence of CDC-defined COVID-19 Starting 1 Day After Month 0 Dose Until Month 6 Dose in FAS Cohort Between Analysis Group 1 (AG1) and Analysis Group 2-1 (AG2-1).

2. Clinical Question

3. Trial Design

The registry describes a randomized, parallel, multicenter study in Phase 2 and Phase 3. It lists quadruple masking for the study, but the Part A comparison used for the primary result is explicitly described as open-label. The record states an enrollment of 14237 and a completed status. The study evaluated different dosing regimens of Moderna COVID-19 mRNA vaccine in people living with HIV or with comorbidities associated with elevated risk of severe COVID-19. Regimen selection was determined in part by evidence of prior SARS-CoV-2 infection at enrollment.

The primary result is from Part A and compares Study Group 1, Analysis Group 1, with Study Group 2, Analysis Group 2-1. Both groups are described as adults living with HIV; their baseline SARS-CoV-2 status and dose schedules differ. Group 1 received doses on Months 0, 1, and 6, while Group 2 received doses on Months 0 and 6. The registry also describes groups of HIV-negative adults, but they are not part of this primary comparison.

Part B is described as blinded and includes a comparison of a Month 6 monovalent mRNA-1273 booster with a Month 6 bivalent mRNA-1273.222 booster. Those groups and outcomes are distinct from the primary Part A comparison. The record lists 25 outcome measures and 10 analyses, while only some secondary outcome results are represented in the posted results. The registry record does not report the original design assumptions for the primary comparison, the randomization procedure specific to this analysis, or the detailed analysis plan.

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: In Part A, participants were enrolled into 4 study groups defined by baseline HIV and SARS-CoV-2 anti-S status. All ppts received mRNA-1273 at Month 0. Study groups 1 and 3 also received mRNA-1273 at Month 1. 236 ppts were excluded due to major data issues. In Part B, ppts who received Month 6 vaccination under protocol v6+ were randomized to mRNA-1273 or mRNA-1273.222. Ppts who received M6 vaccination before protocol v6 were not randomized because mRNA-1273.222 was unavailable.

Part A (Open-label)

Reasons for not completing this period:

Part B (Blinded)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: FAS

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.

Population: Overall SARS-CoV-2 neg indicates all anti-Spike (anti-S), anti-Nucleocapsid (anti-N), and virus detection (NAAT) neg. Overall SARS-CoV-2 pos indicates one of anti-S, anti-N, or NAAT pos. Only anti-S is used to determine study groups (SG) in the Participant Flow table. Anti-S, anti-N, and NAAT are used to determine analysis groups (AG) for analyses. SG1 (anti-S neg) includes AG1 (overall neg) and AG2-2 (overall pos). SG3 (anti-S neg) includes AG3 (overall neg) and AG4-2 (overall pos).

Only the first 6 of 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Part A: Number of Participants With a First Occurrence of CDC-defined COVID-19 Starting 1 Day After Month 0 Dose Until Month 6 Dose in FAS Cohort Between Analysis Group 1 (AG1) and Analysis Group 2-1 (AG2-1), measured over 1 day after Month 1 dose until Month 6 dose.

The primary endpoint counts participants with a first occurrence of CDC-defined COVID-19 in the FAS cohort. The definition requires at least one specified systemic or respiratory symptom and a post-baseline nasal swab, or respiratory sample if hospitalized, that is positive for SARS-CoV-2 by NAAT. The symptom date and positive NAAT date must be within 14 days of one another. The endpoint date is the earlier of those dates that meets the stated criteria. The registry says that CDC criteria do not require adjudication.

The endpoint therefore captures symptomatic, laboratory-confirmed illness rather than infection detected only through testing without symptoms. Such a definition can make the clinical event more specific to symptomatic disease, while also making event ascertainment dependent on symptoms being recognized and testing being performed. The record notes that additional NAAT testing occurred at Month 1 for AG1 but not AG2-1, and that Month 1 NAAT results were not considered.

There is a timing inconsistency within the record: the primary outcome title describes follow-up beginning one day after the Month 0 dose, whereas its timeframe field says one day after the Month 1 dose. The endpoint description also refers to the Month 1 testing difference. The registry record does not resolve the conflicting start-time wording. This matters when interpreting exactly which person-time and events are included.

7. Results Posted in the Registry

• Part A (Open-label): Study Group 2, Analysis Group 2-1 (AG2-1): 1.9% (151/7941) versus 3.1% (76/2476) on control; risk difference -1.2 percentage points (95% CI -1.9 to -0.4), risk ratio 0.62 (0.47–0.81); unadjusted two-proportion z-test p = <0.001.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Part A: Number of Participants With First CDC-based COVID-19 Occurrence From 1 Day After Month 0 Dose to Month 6 Dose in FAS Cohort, Comparing Baseline SARS-CoV-2 Negative vs Positive, Regardless of HIV Status

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Part A (Open-label): Analysis Group 2-1 (AG2-1) + Analysis Group 4-1 (AG4-1): 1.9% (179/9613) versus 3.0% (86/2867) on control; risk difference -1.1 percentage points (95% CI -1.8 to -0.5), risk ratio 0.62 (0.48–0.80); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Part A (Open-label): Analysis Group 2-1 (AG2-1) + Analysis Group 4-1 (AG4-1) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Part A: Number of Participants With First COVE-based COVID-19 Occurrence From 1 Day After Month 0 Dose to Month 6 Dose in FAS Cohort, Comparing Baseline SARS-CoV-2 Negative vs Positive, Regardless of HIV Statu

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Part A (Open-label): Analysis Group 2-1 (AG2-1) + Analysis Group 4-1 (AG4-1): 0.5% (45/9613) versus 1.2% (34/2867) on control; risk difference -0.7 percentage points (95% CI -1.1 to -0.3), risk ratio 0.39 (0.25–0.61); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Part A (Open-label): Analysis Group 2-1 (AG2-1) + Analysis Group 4-1 (AG4-1) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Part A: Number of Participants With First COVE-based Severe COVID-19 Occurrence From 1 Day After Month 0 Dose to Month 6 Dose in FAS Cohort, Comparing Baseline SARS-CoV-2 Negative vs Positive, Regardless of HI

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Part A (Open-label): Analysis Group 2-1 (AG2-1) + Analysis Group 4-1 (AG4-1): 0.1% (6/9613) versus 0.2% (7/2867) on control; risk difference -0.2 percentage points (95% CI -0.4 to 0.0), risk ratio 0.26 (0.09–0.76); unadjusted two-proportion z-test p = 0.008.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Part A (Open-label): Analysis Group 2-1 (AG2-1) + Analysis Group 4-1 (AG4-1) and the control. The interval gives the range of effects that remain compatible with the data.

The secondary clinical outcomes broaden the primary assessment by using a CDC-based endpoint across broader analysis groups, a COVE-based COVID-19 endpoint, and a COVE-based severe endpoint. The posted comparisons of baseline SARS-CoV-2-negative and positive groups generally show fewer event counts and estimates favoring the baseline-positive groups. The antibody and T-cell outcomes address immune measurements rather than clinical illness, so they represent a different type of evidence and should not be conflated with clinical endpoints.

The pattern is broadly compatible with the primary contrast, but it does not independently establish the same conclusion. Secondary outcomes may differ in case definition, analysis population, follow-up, and event frequency. In particular, the severe outcome has few observed events, limiting precision. A consistent direction across several outcomes can be informative, but the outcomes may be correlated and do not amount to independent replications.

Unadjusted secondary p-values carry less weight when many outcomes or comparisons are examined. With multiple opportunities to observe a small p-value, the chance of at least one apparently notable finding can increase. A multiplicity procedure can address this issue, but the registry record does not report one for these results. The record also does not establish which secondary outcomes were prespecified as confirmatory versus exploratory.

For these reasons, secondary results are useful for context and for identifying patterns that may warrant further study, but they should not override the primary endpoint or be interpreted as a collection of definitive tests. The breadth of listed outcomes includes clinical events and immune assays, and each requires its own attention to endpoint definition, analysis population, uncertainty, and missingness.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Part A: Number of Participants With First CDC-based COVID-19 Occurrence From 1 Day After Month 0 Dose to Month 6 Dose in FAS Cohort, Comparing Baseline SARS-CoV-2 Negative vs Positive, Regardless of HIV Status
• Part A: Number of Participants With First COVE-based COVID-19 Occurrence From 1 Day After Month 0 Dose to Month 6 Dose in FAS Cohort, Comparing Baseline SARS-CoV-2 Negative vs Positive, Regardless of HIV Statu
• Part A: Number of Participants With First COVE-based Severe COVID-19 Occurrence From 1 Day After Month 0 Dose to Month 6 Dose in FAS Cohort, Comparing Baseline SARS-CoV-2 Negative vs Positive, Regardless of HI
• Part B: Number of Participants With a First Occurrence of CDC-based COVID-19 Starting 14 Days After Month 6 Dose Until End of Follow up in RM6 Cohort Who Have Month 6 Hybrid Immunity Between Month 6 Monovalent and Bivalent Boost Groups
• Part B: Number of Participants With a First Occurrence of CDC-based COVID-19 Starting 14 Days After Month 6 Dose Until End of Follow up in RM6 Cohort Who Have Month 6 Vaccine Immunity Between Month 6 Monovalent and Bivalent Boost Group
• Part A: Geometric Mean Titers (GMTs) of SARS-CoV-2 Antibodies as Measured by Neutralizing Antibody (NAb) Assay
• Part A: Geometric Mean Titers (GMTs) of SARS-CoV-2 Antibodies as Measured by Meso Scale Discovery-electrochemiluminescence Assay (MSD-ECL)
• Part A: Geometric Mean Titers of T-Cell Responses as Measured by Intracellular Cytokine Staining (ICS) Assay at Month 0

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is binary at the participant level: each participant is counted as having or not having a first qualifying event during the defined follow-up. The posted arm summaries give the number of participants with an event and the analysis-group size. A simple unadjusted comparison uses those summaries to estimate the difference in event risk between groups. It does not account for time to event or adjust for other characteristics.

What does the risk difference describe?

A risk difference is the event risk in one group minus the event risk in the reference group, using the comparison's stated group order. Its sign indicates the direction of the contrast; a negative value in this comparison corresponds to lower event risk in the baseline-positive group. The confidence interval describes a range of values compatible with the data under the calculation's assumptions. An interval that excludes zero is evidence against equal risks under that calculation. This is a general explanation of the statistic, not a claim that every source of bias or uncertainty has been addressed.

The displayed recalculation is unadjusted and based on posted summary counts. It is not necessarily the same estimand or method as the sponsor's own time-to-event analysis. Small differences between measures can arise because risk over a specified interval and the instantaneous event-rate contrast modeled by a hazard ratio are not interchangeable. The recalculation also does not recover participant-level timing, censoring, covariates, or missing outcome information.

How does a Cox regression differ?

The registry reports a Cox regression analysis with a hazard ratio. In general, Cox regression models the time until an event and can incorporate the timing of events and censoring, subject to its assumptions. A hazard ratio below one indicates a lower modeled event hazard in the comparison group relative to the reference group, given the model and group ordering. It is not a risk ratio, and it should not be read as a direct percentage change in the cumulative probability of an event. The registry record does not report whether the Cox model was adjusted, which covariates were included, how proportional hazards were assessed, or how censoring was handled.

The primary outcome is labeled as a count of participants, but the sponsor analysis also uses time-to-event information. The risk-difference calculation and Cox analysis therefore offer complementary summaries rather than duplicate versions of the same calculation. The posted results give a Cox hazard ratio of 0.58 and a 95% confidence interval from 0.44 to 0.77. That interval summarizes uncertainty for the sponsor's model-based estimate, not for the unadjusted risk difference.

Why do the endpoint definition and timing matter?

Outcome definitions determine which events enter an analysis. Here, a qualifying event requires symptoms, a positive NAAT, and dates within the stated window. The registry notes that Month 1 NAAT results were not considered because additional testing was performed at Month 1 for AG1 but not AG2-1. This restriction is relevant to the consistency of ascertainment between the groups. The primary outcome title and timeframe also give different follow-up start points, so the exact interval is not fully clear from the record.

General statistical practice treats endpoint definition, observation period, analysis population, and handling of incomplete follow-up as integral parts of an estimate. If participants have different observation time or incomplete outcome ascertainment, the interpretation of a simple risk comparison can be affected. The registry record does not report detailed missing-data handling, follow-up completeness, or sensitivity analyses for this primary result.

What does a p-value tell the reader?

A p-value measures how inconsistent the observed data are with a specified null model, assuming that model and the analysis conditions hold. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical value of an effect. The primary recalculation reports p <0.001, while the sponsor's Cox analysis reports p < .001. Those formats are retained as recorded. The record does not state whether the primary test was adjusted for other analyses or whether it was selected before examining the data.

Because the primary contrast is between groups defined by baseline infection status, statistical adjustment alone would not necessarily remove confounding. The record's reported analysis details are not sufficient to establish which potential confounders were measured or controlled. A precise estimate can still be biased if group differences, endpoint ascertainment, or model assumptions are not adequately addressed.

11. Confidence Intervals and Precision

Precision is reflected in the width of a confidence interval: narrower intervals indicate less sampling uncertainty under the analysis assumptions, while wider intervals indicate greater uncertainty. The sponsor's primary Cox estimate is 0.58, with a 95% confidence interval from 0.44 to 0.77. The interval lies below 1, the general null value for a hazard ratio, and the reported p-value is < .001. This supports a lower modeled hazard for the comparison group under the reported analysis, while still allowing a range of effect sizes.

The recalculated risk-difference interval also excludes no difference, and its point estimate favors the baseline-positive group. The registry lists the unadjusted p-value as <0.001. The risk-difference interval and hazard-ratio interval refer to different measures, so their widths and limits should not be compared as though they were on a shared scale.

Precision is more limited for the severe COVE-based secondary outcome. Only 7.0 and 6.0 participants were counted with events in the two groups. Its recalculated risk-difference interval includes no difference, while the sponsor's Cox analysis reports a hazard ratio of 0.27 with a 95% confidence interval from 0.09 to 0.82 and p = 0.02. Sparse events can make estimates sensitive to a small number of outcomes; statistical significance in one method does not erase that limitation.

Confidence intervals quantify uncertainty from sampling within the specified model and data. They do not automatically capture uncertainty from endpoint timing ambiguity, unmeasured differences between groups, missing information, or model misspecification. The registry does not report enough methodological detail to assess those sources fully. Thus, the intervals describe statistical precision, not certainty that the estimated contrast is causal or will be reproduced in another population.

12. 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 illustrations based on the observed effect, not the trial's original design calculation. For the binary outcomes, figures are participants per arm. The planning assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is set to a stated fraction of the observed effect. The record does not report the trial's own original sample-size assumptions or justification.

The 80% power column gives the planned participants per arm for a hypothetical future study to have an 80% chance of detecting the specified effect under those assumptions, if that effect and the assumptions are correct. The 90% power column applies the same interpretation with a higher target chance of detection. These are design calculations, not statements about the probability that the observed trial result is true. This is not observed power.

Rows for planning at 100% of the observed effect use the full observed contrast as the target. Rows for planning at 75% or 50% use smaller target effects. Smaller effects generally require more participants to distinguish them from random variation at the same significance and power targets. These figures should be treated as planning illustrations rather than recommendations: the observed estimate may be uncertain, and the record does not report whether future studies would share the same population, event rate, endpoint definition, or follow-up.

The sample-size assumptions state that time-to-event calculations give total events using the Schoenfeld approach, while the other calculations give participants per arm. The primary comparison is binary and its planning figures therefore use participants per arm. Any curve should be read in the context of these assumptions and the corresponding endpoint and target effect; it does not show the number of participants actually enrolled in this trial.

13. 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 = 0.031; treatment proportion p₁ = 0.019; average p̄ = 0.025.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.2202 = 0.4315.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.2200 = 0.1852.
4. n = (0.4315 + 0.1852)² / (-0.0117)² = 2,788.0, rounded up to 2,789 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 <- 0.0306946688206785                          # control proportion
p1 <- 0.0190152373756454                          # 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(5168813)
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 = 2,789 (80% power), 3,732 (90% power)
 75% of observed effect: n per arm = 5,241 (80% power), 7,016 (90% power)
 50% of observed effect: n per arm = 12,429 (80% power), 16,639 (90% power)
power.prop.test: 2788 
Simulated power at n = 2,789 per arm: 0.806 (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.

14. Analyses Reported by the Sponsor

The sponsor-reported primary analysis is a Cox regression with a hazard ratio of 0.58, a 95% confidence interval from 0.44 to 0.77, and p < .001. The recalculated result uses the posted event counts to make an unadjusted risk comparison, reporting a negative risk difference and p <0.001. Both summaries indicate fewer observed events, or a lower modeled event hazard, in the baseline SARS-CoV-2-positive group than in the baseline-negative group.

The results are directionally consistent but should not be treated as numerically interchangeable. A risk difference compares event proportions over the stated interval; a hazard ratio is a time-to-event measure from the Cox model. The latter may use event timing and censoring, while the recalculation is based on aggregate counts. The registry does not report enough detail to determine whether the Cox analysis adjusted for covariates or to reproduce its model from participant-level data.

The record marks no discrepancy between the recalculation and the reported analysis. Accordingly, the recalculation is best understood as a transparent secondary summary of the posted counts, and the sponsor's Cox estimate as the registry-reported model-based analysis. Neither result by itself resolves the limits of a comparison between groups defined by baseline infection status or the ambiguity in the follow-up start wording.

15. Multiplicity, Missing Data and Analysis Populations

The registry record lists 25 outcome measures and 10 analyses, with several clinical and immune endpoints. Testing multiple outcomes raises a multiplicity concern: even when each individual test uses a conventional threshold, repeated testing can increase the chance of at least one result appearing statistically notable by chance. The record does not report a multiplicity adjustment, a testing hierarchy, or which outcomes were designated confirmatory. The unadjusted secondary p-values should therefore be interpreted cautiously rather than as if each were the sole planned test.

The primary analysis is described as using the FAS cohort, but the registry record does not provide the detailed FAS definition in the posted endpoint description. It also does not report how participants with missing outcome assessments, incomplete follow-up, or deviations from the planned schedule were handled. These omissions limit assessment of the analysis population and the extent to which the posted counts represent all participants intended for analysis.

The event definition requires symptoms and a positive NAAT within a specified window. The record notes a difference in additional Month 1 testing between AG1 and AG2-1 and says Month 1 NAAT results were not considered. That procedural detail is relevant to ascertainment, but it does not fully explain whether testing access, follow-up, or symptom reporting differed in other ways.

Information about interim analyses, a prespecified statistical analysis plan, and sensitivity analyses is not reported in the registry record. The record also does not state how competing events or censoring were treated in the Cox analysis. These are not reasons to assume an analysis was inappropriate; they are limits on what can be evaluated from the posted statistical details. Clear reporting of populations, missing-data handling, multiplicity, and model specifications is important for judging how robust an estimate is.

16. 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: pneumonia (0.2%), death (0.2%), diabetic ketoacidosis (0.2%), necrotising fasciitis (0.2%), ankle fracture (0.2%).

The registry reports serious adverse event counts by study group, with an at-risk count and an affected count for each group. For example, the Part A Study Group 1 entry reports 84.0 serious adverse events among 3740 at risk, while the Part A Study Group 2 entry reports 126.0 among 7941 at risk. These are group-specific registry counts; the groups have different sizes and are not a randomized head-to-head safety contrast in this presentation. They should not be compared as though the raw counts alone represented comparable risks.

The record also provides other adverse-event counts and deaths by group, and lists serious terms including pneumonia, death, diabetic ketoacidosis, necrotising fasciitis, ankle fracture, and wrist fracture, each shown as 0.2%. The posted serious-event figures describe what was recorded, but do not establish that an event was caused by vaccination. The registry record does not provide a causal attribution analysis or a detailed account of event timing relative to doses in these summary fields.

Safety counts should be interpreted alongside group composition, observation time, event definitions, and ascertainment. The record does not report all of those details in the adverse-event summary. These figures are descriptive and are not a substitute for a comparative safety analysis.

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

The posted primary results show a lower observed frequency of CDC-defined COVID-19 in the baseline SARS-CoV-2-positive group than in the baseline-negative group among adults living with HIV in the stated Part A comparison. The recalculated risk comparison and sponsor-reported Cox analysis both favor the baseline-positive group, and each reports strong statistical evidence against its respective null comparison. The sponsor's hazard ratio provides a time-to-event summary, while the recalculation summarizes event risk from the posted counts.

That pattern does not establish that prior SARS-CoV-2 infection caused the difference. Participants were grouped by baseline infection status; the record does not describe randomized assignment to that status. The groups also differ in the scheduled Month 1 dose. Differences in participant characteristics, exposure history, testing, or follow-up could contribute to the contrast, and the registry record does not give enough information to assess all such possibilities.

The result also does not directly compare the vaccine with no vaccine, nor does it isolate the effect of a specific vaccine product against another. Both primary groups are described as receiving the COVID-19 mRNA vaccine, with schedules linked to their baseline status. Part B's monovalent and bivalent booster comparison addresses a separate question.

Statistical significance does not establish clinical importance, and confidence intervals do not account for every source of uncertainty. Endpoint timing is described inconsistently in the record, and details about missing data, covariate adjustment, and multiplicity are not reported. The secondary results offer context but do not eliminate these limitations. Overall, the registry supports a statistical association in the analyzed groups, while leaving causal interpretation and broader applicability uncertain.

18. Why This Trial Matters Statistically

This trial illustrates why the question behind a comparison matters as much as the numerical result. The primary analysis compares groups defined by baseline SARS-CoV-2 status, not simply two interventions assigned at random. Reading the estimate as a causal effect of prior infection would therefore go beyond what the registry record establishes. Clear descriptions of who was compared, how groups were formed, and what each estimate measures are central to biostatistical interpretation.

It also demonstrates the value of distinguishing absolute-risk summaries from time-to-event models. A risk difference can be easier to relate to event proportions over a defined interval, whereas a Cox hazard ratio uses event timing and model assumptions. Their agreement in direction is informative, but they are not interchangeable and may answer different questions.

The endpoint itself shows how operational definitions shape evidence. Symptom criteria, NAAT confirmation, timing windows, and the treatment of Month 1 tests determine which events count. The registry's conflicting start-time descriptions make that detail particularly important. Transparent outcome definitions and consistent reporting allow readers to judge whether the estimand is clear and whether ascertainment could differ between groups.

Finally, the secondary endpoints and safety summaries emphasize the need to consider precision, multiple testing, analysis populations, and reporting completeness together. A result can have a small p-value while still leaving questions about confounding, missingness, or clinical relevance. Statistical literacy means reading the estimate, its uncertainty, the design, and the limits of the record as one connected account.

19. 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 groups are defined by baseline SARS-CoV-2 status, and the registry does not report that this status was randomized. The groups also have different dose schedules: Study Group 1 is described as receiving doses on Months 0, 1, and 6, while Study Group 2 is described as receiving doses on Months 0 and 6. These features limit a causal interpretation of the group contrast as the effect of prior infection alone.

The primary outcome title and timeframe field disagree about whether follow-up begins one day after the Month 0 or Month 1 dose. The endpoint description notes different additional NAAT testing at Month 1 and says those results were not considered. The registry record does not explain the timing discrepancy or provide enough detail to assess all potential differences in outcome ascertainment.

The result record does not report the full FAS definition, missing-data handling, follow-up completeness, the covariates in the Cox model, model diagnostics, censoring details, or sensitivity analyses. It also does not report a multiplicity strategy for the many listed outcomes. The absence of these details in the registry record does not show that the underlying study lacked such procedures, but it limits independent assessment of the posted results.

Some secondary outcomes have few events, especially the severe endpoint, and so estimates can be imprecise. The posted immune outcomes are not substitutes for clinical outcomes. Safety figures are descriptive, with different at-risk counts across groups, and the summary does not establish event causality. Finally, the record's broad enrollment and completed status do not by themselves establish how well the estimates generalize to populations or circumstances not described in the primary comparison.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Efficacy Study of COVID-19 mRNA Vaccine in Regions With SARS-CoV-2 Variants of Concern — NCT05168813.
• Registry record last updated 2026-04-21; status last verified 2026-04.
• Results first posted 2025-11-26.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The registry record describes a completed, randomized, parallel, multicenter study in Phase 2 and Phase 3, with quadruple masking listed at the study level. The primary result is an open-label Part A comparison among adults living with HIV, grouped by baseline SARS-CoV-2 status and receiving different dose schedules. The endpoint counts first CDC-defined symptomatic, NAAT-confirmed COVID-19 events.

The posted counts and unadjusted recalculation favor the baseline-positive group, as does the sponsor-reported Cox analysis. The record also reports secondary clinical and immune outcomes and descriptive adverse-event counts. Interpretation is limited by the nonrandomized basis of the primary group distinction, differing schedules, an unresolved follow-up timing inconsistency, and statistical details not reported in the registry record.