Trial of Two Versus Three Doses of Human Papillomavirus (HPV) Vaccine in India: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Trial of Two Versus Three Doses of Human Papillomavirus (HPV) Vaccine in India (NCT00923702), a non-randomized phase 4 study. The primary outcome is Median Florescent Intensities (MFI) of the Total Antibodies to Vaccine-included HPV Types (16/18/6/11) at Different Time Points.

2. Clinical Question

3. Trial Design

This is a phase 4, parallel study with non-randomized allocation and no masking. The registry lists enrollment of 22729 and identifies cervical cancer and cervical precancerous lesions as conditions of interest. The study began in 2009-09, and primary completion is listed as 2017-01. Its status is recorded as unknown.

The arms represent distinct dose histories rather than randomly assigned regimens. The three-dose group received vaccine at days 1, 60 and 180+. The two-dose group received doses at days 1 and 180+. The group called two doses by default received doses at days 1 and 60 by default, while the single-dose group received one dose by default; both are described as incomplete-dose groups. A separate cohort was unvaccinated. These differing histories and non-randomized allocation limit direct causal interpretation of differences between groups.

The registry summary states that the Government of India stopped vaccination in all HPV vaccine trials in the country in April 2010 for reasons unrelated to this study. The record does not describe how this interruption affected follow-up, dose completion, outcome measurement, or analysis. Those consequences are therefore not reported in the registry record.

The primary outcome is antibody measurement at specified months, while the listed secondary outcomes concern infection by other non-targeted high-risk HPV types and cervical neoplasia associated with HPV types not included in the vaccine. The record lists five outcome measures and two analyses, but does not describe the analysis methods or clarify which outcomes were included in each analysis.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Data used for this outcome are from a subset of participants that had their immunogeneicity samples tested in the laboratory. We are indicating the total of HPV 16 L1 antibodies of the HPV 16 L1 antiboties at month 7 for the 3-dose and 2-dose groups and at month 12 for the 2 doses by default and single-dose groups. No immunogeneity data was collected from the unvaccinated cohort as it was recruited adhoc to assess the HPV infections and CIN endpoints.

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 Median Florescent Intensities (MFI) of the Total Antibodies to Vaccine-included HPV Types (16/18/6/11) at Different Time Points, measured over Month 7 (for 3-dose and 2-dose groups), 12 (for 2 doses by default and single-dose groups), 18, 36, 48.

The primary endpoint is the geometric mean of median fluorescence intensity (MFI) measurements for antibodies to vaccine-included HPV types 16, 18, 6 and 11. Samples were treated with EDTA and analysed using Luminex-based multiplex serology, which measures binding antibodies to the L1 major capsid protein. The record describes MFI as a measure of antibody concentration and says these measurements are directly comparable with optical densities measured by ELISA.

Antibody measurements are used to assess immunogenicity: the immune response indicated by measured binding antibodies after vaccination. This is relevant to the study’s comparison of dose patterns, but it is not itself a direct measurement of persistent infection or cervical neoplasia. The primary outcome description lists measurements at month 7 for the three-dose and two-dose groups; month 12 for the two doses by default and single-dose groups; and months 18, 36 and 48. The population description says that the outcome uses a subset whose immunogenicity samples were tested in the laboratory, but the record’s description is truncated and does not fully specify that subset.

The endpoint specifies a 95% confidence interval around the geometric mean. Numerical MFI estimates and confidence intervals are not reported in the registry record, so the magnitude and precision of the measured antibody responses cannot be assessed from the posted results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Interpreting this study requires separating the endpoint definition, the study design and the statistical evidence actually posted. A well-defined measurement can explain what investigators intended to assess, but it cannot substitute for an estimate, an uncertainty interval or a clearly specified comparison. For this study, the registry describes the antibody assay and time points but does not report numerical primary outcome results or a detailed statistical analysis plan.

What does a geometric mean represent?

A geometric mean summarizes measurements on a multiplicative scale. In general, it is useful when values are positive and their variation is better described in relative terms than as a fixed amount above or below an arithmetic average. It can be interpreted as a central value after accounting for the multiplicative structure of the measurements. That general interpretation does not determine which statistical model was used in this study; the model is not reported in the registry record.

The endpoint description calls the primary measure median fluorescence intensity and identifies the reported parameter as a geometric mean. These terms refer to distinct aspects of a measurement and summary: fluorescence intensity is the assay readout, while the geometric mean is a way to summarize values across observations. The record states that MFI values quantify binding antibody concentration and are directly comparable with ELISA optical densities. It does not provide assay thresholds, rules for values below a detection limit, or details on transformations before analysis.

How should a confidence interval be read?

In general, a confidence interval expresses statistical uncertainty around an estimate under the assumptions of the method used to calculate it. A narrower interval indicates greater precision under those assumptions; a wider interval indicates greater uncertainty. A confidence interval is not a guarantee that the true value falls within the specific interval, and it does not address systematic bias from study design, missing data or measurement procedures.

The primary endpoint specifies 95% confidence intervals, but their numerical limits are not reported in the registry record. As a result, neither the estimates’ precision nor the extent of overlap or separation between groups can be assessed from the posted record. An interval would also need to be interpreted alongside a defined contrast. Separate group intervals are not, by themselves, a formal test of the difference between groups.

Why does allocation matter?

Random allocation, in general, is intended to make treatment groups comparable on average with respect to both measured and unmeasured baseline characteristics. This study is explicitly recorded as non-randomized. Dose groups may therefore reflect differences in participant characteristics, timing, or completion of vaccination. Statistical adjustment can sometimes address measured differences, but it cannot guarantee removal of confounding, and the record does not report an adjustment strategy. The stated design therefore limits causal interpretation of group contrasts.

The parallel model indicates that groups are followed in parallel rather than describing a crossover. However, that design label does not establish that the groups had the same follow-up, were assessed at identical times, or were analyzed using a particular model. The record lists multiple assessment months and distinct initial measurement months for several groups. Methods for handling repeated observations or differing timing are not reported.

What does the planning illustration mean?

General sample-size planning asks how many participants or events would be needed under specified assumptions to detect a chosen effect with a chosen statistical criterion. Such an illustration depends on assumptions and does not describe what was observed in the trial. The posted planning assumptions use a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a stated fraction of the observed effect. For time-to-event illustrations, the total number of events is the planning quantity; for other outcomes, the quantity is participants per arm.

Those planning assumptions are not the trial’s own design justification. The registry record does not report a sample-size calculation, the effect size used to design the trial, or a rationale for the number enrolled. A planning illustration based on an observed effect is conditional on that effect and should not be described as observed power. It also cannot repair the lack of a reported primary estimate or establish that the trial had adequate precision.

Why might over-dispersion matter?

In general, over-dispersion occurs when variation in observed data exceeds what a statistical model assumes. For count outcomes, for example, a model that assumes a particular relationship between the mean and variance may understate uncertainty if the data are more variable than assumed. Methods that allow additional variation can produce more appropriate uncertainty estimates. This is a general statistical consideration, not a claim that over-dispersion occurred in this study. The registry does not identify a count model or report an over-dispersion assessment.

More broadly, no detailed model, covariate set, test, confidence-interval construction method, or handling of repeated measurements is reported in the registry record. Statistical interpretation must therefore remain limited to the design and outcome definitions that are stated, rather than assuming methods that are not documented.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists five outcome measures and two analyses, but does not report how many hypotheses were tested, which outcome was assigned to each analysis, or whether any testing hierarchy was used. It does not describe an adjustment for multiple comparisons, a primary time point among the several listed months, or how multiple HPV types were handled in analysis. When many outcomes, groups or time points are examined, the chance of at least one apparently notable result can increase. In general, a prespecified testing strategy or an appropriate adjustment can help address that issue. Whether such steps were used is not reported in the registry record.

The primary outcome description identifies a laboratory-tested subset, but the subset’s size and the method used to select participants are not reported. The number of missing samples, reasons for missingness, and handling of missing measurements at each assessment are also not reported in the registry record. Missing data can affect both precision and the representativeness of an estimate, particularly when availability of samples is related to participant characteristics or outcomes. The record does not describe imputation, complete-case analysis, or sensitivity analyses.

Likewise, the analysis population is not defined. The registry does not say whether analyses included all enrolled participants, only those who received particular doses, or only those with tested samples. It does not report how participants who received fewer doses than planned were classified beyond describing two default groups as incomplete-dose patterns. The trial’s sample-size justification and interim-analysis plans are not reported in the registry record. These gaps prevent assessment of whether the posted analysis followed a prespecified plan or how robust findings might be to alternative analysis choices.

11. 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: abortion/gynecological conditions (6.2%), neonatal death (1.0%), preterm delivery (0.8%), hospitalization (0.6%).

The registry lists serious-event counts, other adverse-event counts and deaths for each group. For the three-dose group, the serious-event field records 125 affected among 4348 at risk, the other-event field records 637 among 4348 at risk, and deaths are listed as 30. The corresponding entries are 135, 614 and 22 for the two-dose group, with 4979 at risk; 292, 614 and 38 for two doses by default, with 3452 at risk; 356, 296 and 45 for the single-dose group, with 4950 at risk; and 0, 0 and 9 for the unvaccinated group, with 4646 at risk.

These are descriptive registry counts. They should not be read as randomized comparisons or as evidence that vaccination caused or prevented an event. The groups were not randomly allocated, and the record does not report comparative safety analyses, event definitions, follow-up time for these counts, or whether individuals could contribute to more than one event category. The listed serious-event terms include abortion/gynecological conditions at 6.2%, neonatal death at 1.0%, preterm delivery at 0.8%, and hospitalization at 0.6%. The denominator and calculation basis for these percentages are not reported in the registry record. The counts and terms alone do not establish comparative risk or causality.

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

The record means that investigators set out to examine antibody responses under different quadrivalent HPV vaccine dose patterns, with a stated hypothesis that a two-dose regimen would offer similar immunogenicity and protection to a three-dose regimen. It documents the groups, the antibody assay, and several planned measurement times. It also records secondary outcomes concerning infection by other high-risk HPV types and cervical neoplasia associated with non-included HPV types.

It does not provide the numerical primary antibody findings. Without those estimates and their confidence intervals, the record does not show how antibody measurements compared across dose groups or how precisely any group-level values were estimated. It also does not report numerical results for the secondary clinical outcomes. Consequently, the record cannot establish from its posted results whether dose patterns had similar antibody responses or whether they were associated with different infection or neoplasia outcomes.

Even if numerical antibody results were available, an antibody measurement is not itself a direct clinical outcome. The primary endpoint assesses immunogenicity, whereas the study’s stated protection hypothesis concerns persistent infection and cervical neoplasia. The relationship between an antibody measurement and those outcomes would need to be assessed with relevant clinical outcome data and an appropriate analysis; the registry record does not report those results.

The non-randomized design also means that comparisons among the dose groups may be influenced by differences between participants, and the record does not describe adjustment methods. The reported vaccination stoppage in April 2010 adds a potential context for dose completion and follow-up, but its specific effects are not reported. The findings that can be stated from this record are therefore about the study’s design and measurement plan, not a conclusion that one dose pattern is better, equivalent, or protective.

13. Why This Trial Matters Statistically

This study is statistically notable because its central question compares dose patterns, while its groups were not randomly allocated. That distinction is important for clinical researchers: a contrast between observed groups is not automatically an estimate of the causal effect of assigning one regimen rather than another. Differences in participant characteristics, dose completion or follow-up may accompany differences in dose history. Interpretation must account for the design rather than treating group labels as interchangeable randomized arms.

The endpoint also illustrates the distinction between a biological measurement and a clinical outcome. Antibody MFI is used as an immunogenicity measure, but the hypothesis extends to persistent HPV infection and cervical neoplasia. Statistical evidence about the antibody endpoint cannot be assumed to answer the clinical-outcome question without results linking the measurements to those outcomes. The record lists relevant secondary outcomes but does not report their numerical findings.

Finally, the record highlights why estimates and uncertainty matter as much as endpoint definitions. A geometric mean with a confidence interval can communicate both the measured level and its precision; an endpoint description without those numbers cannot show whether group estimates differ or how uncertain they are. The study’s multiple arms, time points and tested subset further make clear why transparent reporting of analysis populations, missing data and multiplicity procedures is important. Those details are not reported in the registry record, so the posted information supports a careful account of what was intended and measured, but not a definitive statistical answer to the dose-comparison question.

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 registry record does not report numerical values for the primary antibody endpoint, confidence-interval limits, or numerical findings for the listed secondary outcomes. It also does not report the detailed statistical methods used to analyze the endpoint. This prevents assessment of the effect estimates, their precision and the basis for any inferential claim.

Non-randomized allocation limits comparability between dose groups. Participant characteristics, baseline antibody measurements, adjustment variables and methods for addressing confounding are not reported. The record describes different dose patterns, including groups with incomplete doses, but does not explain how those patterns arose or how they were handled analytically. The stated stoppage of vaccination in April 2010 may be relevant to dose completion and follow-up, but its consequences are not described.

The primary outcome uses a laboratory-tested subset, yet the subset’s size, selection process and completeness are not reported. Missing measurements, reasons for missingness, and analysis populations are also not reported. Although several outcome time points are listed, the record does not specify the primary analysis time point or how repeated measurements and differing initial assessment months were addressed. The number of outcome measures and analyses is listed, but multiplicity procedures are not described.

The record’s status is unknown, and the primary outcome population description is truncated. Eligibility criteria, detailed participant characteristics, sample-size justification, and interim-analysis plans are not reported in the registry record. Together, these limitations mean that the registry entry alone is insufficient for evaluating the completeness, precision or robustness of the trial’s primary statistical findings.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Trial of Two Versus Three Doses of Human Papillomavirus (HPV) Vaccine in India — NCT00923702.
• Registry record last updated 2023-09-28; status last verified 2023-09.
• Results first posted 2022-10-20.
• 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

The registry describes a phase 4, non-randomized, unmasked parallel study in India with enrollment of 22729. It compares groups with differing quadrivalent HPV vaccine dose histories and an unvaccinated cohort. The primary outcome is a geometric-mean summary of MFI antibody measurements to HPV types 16, 18, 6 and 11, with a 95% confidence interval and assessments listed across months 7, 12, 18, 36 and 48. The primary outcome uses a laboratory-tested subset.

The study summary states that vaccination in HPV vaccine trials in India was stopped in April 2010 for reasons unrelated to the study. The record lists five outcome measures and two analyses, along with adverse-event counts by group. It does not report numerical antibody results, numerical secondary outcome findings, detailed statistical methods, missing-data handling, analysis populations or multiplicity procedures. The dose groups are not randomized, so they are not a like-for-like control comparison.