Controlled Human Malaria Infection Model for Evaluation of Transmission-blocking Interventions - Study 2: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Controlled Human Malaria Infection Model for Evaluation of Transmission-blocking Interventions - Study 2 (NCT03454048), a randomized na study. The primary outcome is Frequency of Adverse Events in the CHMI-trans Model.

2. Clinical Question

The study was designed to establish a controlled human malaria infection transmission model for evaluating how interventions may affect malaria parasite transmission. Its central scientific question was whether the model could assess gametocyte development and infectiousness to Anopheles mosquitoes after different challenge routes and treatment regimens. The registry describes this as a model-development study, rather than a trial testing a vaccine or biologic.

These questions have different roles. Adverse-event frequency describes events observed during the study period, while mosquito infectiousness addresses whether the challenge model can generate a measurable transmission-related outcome. The registry does not report a prespecified hypothesis comparing the treatment regimens for the primary endpoint, nor does it report a formal statistical analysis plan. Accordingly, the posted results should be read as descriptive evidence about outcomes recorded in this study, not as a definitive treatment-effect evaluation.

3. Trial Design

This was a randomized, open-label, parallel study at a single center. The registry lists the study as completed, with enrollment of 24, a start date of 2018-05-07, and primary completion on 2018-11-01. The phase is listed as NA. Randomization is recorded, but the sequence-generation method, allocation concealment, and assignment procedures are not reported in the registry record. Masking is listed as none.

The study used two challenge cohorts. Cohort A received a controlled human malaria infection delivered by five Pf-infected mosquitoes. Cohort B received a blood-stage challenge by intravenous injection of approximately 2,800 Pf-infected erythrocytes. In both cohorts, the group descriptions begin with a single oral subcurative low dose of piperaquine, identified as 480 mg, and describe a subsequent treatment using either piperaquine or sulfadoxine pyrimethamine. The registry descriptions of when the second treatment was given are cut off; the conditions and timing are not reported in the registry record.

The four posted groups are Group 1 (Cohort A) LD-PIP/LD-PIP2/PIP, Group 2 (Cohort A) LD-PIP/LD-PIP2/SP, Group 3 (Cohort B) LD-PIP/LD-PIP2/PIP, and Group 4 (Cohort B) LD-PIP/LD-PIP2/SP. Each group has a posted size of 6. This structure places challenge route and subsequent regimen within the study design, but the registry record does not describe a formal factorial analysis or state that the study was powered to separate these effects.

The study's purpose was to establish a model for evaluating transmission-blocking interventions, including vaccines, biologics such as monoclonal antibodies, and drugs. Those intervention classes describe the intended future use of the model; they are not all interventions tested in this study. The listed study interventions are piperaquine, sulfadoxine pyrimethamine, atovaquone proguanil, and malaria challenge infections. The registry does not report detailed eligibility criteria, recruitment procedures, or the rationale for the chosen enrollment.

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 Frequency of Adverse Events in the CHMI-trans Model, measured over up to day 51 after challenge infection.

The designated primary endpoint is the frequency of adverse events in the CHMI-trans model, recorded up to day 51 after challenge infection. It is expressed as a number of adverse events rather than as a continuous measurement or a participant-level binary outcome. This endpoint is used to describe the events recorded during the challenge and follow-up period, providing a safety-oriented account of the model's conduct.

The posted arm-level values are 95 for Group 1, 95 for Group 2, 107 for Group 3, and 52 for Group 4. These are presented as event counts. They should not be read as the number of participants with an event: the registry gives no participant-level breakdown for this primary outcome, and multiple events may be recorded for a participant. Nor does the record provide an uncertainty interval or a statistical comparison for these counts. The data therefore describe the posted event totals by group, but do not establish whether differences between groups are systematic or attributable to challenge route or treatment regimen.

Adverse-event frequency does not directly answer the model's transmission question. The registry also identifies mosquito infectiousness and several gametocyte-related outcomes as secondary measures. Those outcomes concern the intended biological and transmission-related performance of the model, whereas the primary endpoint records adverse events. Keeping the endpoints distinct is important: a count of adverse events is not a measure of mosquito infectiousness, and neither endpoint alone characterizes every aspect of the study.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The posted information combines descriptive group summaries with one unadjusted comparison for a secondary outcome. The registry does not report a complete statistical analysis plan, so the methods should be interpreted at the level documented rather than as evidence of analyses that may have been planned or conducted elsewhere.

What does a count outcome represent?

The primary endpoint is the frequency of adverse events, reported as a number. A count records how many events were observed, not necessarily how many people experienced at least one event. If one participant can have more than one event, an event count and a participant count answer different questions. The registry reports arm-level counts but does not report the distribution of events across participants or person-level results. Without that information, the posted totals cannot indicate whether events were concentrated among a few participants or spread across a group.

Count outcomes can vary more than would be expected under a simple Poisson model when observations differ in their underlying event rates or when events cluster within participants. This is called over-dispersion. As a general statistical point, when over-dispersion is present, a model that assumes less variability can produce confidence intervals that are too narrow and p-values that overstate the evidence. The registry does not report a count model, a dispersion estimate, or a participant-level analysis for the primary endpoint. No model-specific inferential conclusion can therefore be attributed to these posted counts.

What does a risk difference mean?

For mosquito infectiousness, the record reports a binary participant outcome and a risk difference. A risk is the proportion of participants with the outcome in a group. A risk difference expresses the difference between the group risks on the same scale; it is not a relative risk or an odds ratio. The posted estimate is 0.75, with Cohort A recorded at 0 of 12 and Cohort B at 9 of 12. The confidence interval from 0.505 to 0.995 describes the range of values compatible with the estimate under the method used to construct it. The interval is not a range of individual participant outcomes and does not guarantee that the true effect lies within it.

The unadjusted p-value is <0.001. In general, a p-value evaluates how compatible the observed data are with a specified null model, conditional on the test assumptions. It is not the probability that there is no difference, nor does it measure the importance of a result. The registry does not identify the specific test used or provide its assumptions, so no particular test procedure should be inferred from the p-value alone.

Why does the comparison structure matter?

The mosquito-infectiousness comparison is Cohort B against Cohort A, with Cohort A as the reference group in the posted analysis. The cohorts received different challenge approaches: infected-mosquito challenge in Cohort A and blood-stage challenge in Cohort B. The study also includes regimen groups within each cohort. A simple two-cohort comparison does not automatically account for those regimen assignments, nor does it disentangle all possible effects of challenge route and treatment. The record does not report a regression model or adjusted estimate for this outcome.

The study is randomized, parallel, and open label, but the registry does not report the randomization method or explain whether cohort assignment itself was randomized. That distinction matters: random allocation can support causal interpretation for the factor actually randomized, but a randomized trial label alone does not establish that every contrast between groups is randomized. No causal claim about challenge route should therefore be inferred from the label without additional design details.

How should planning illustrations be read?

Any sample-size calculations accompanying the results are planning illustrations based on the observed effect, not observed power and not proof that the original study had a particular probability of detecting an effect. The stated illustration assumptions use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is expressed as a fraction of the observed effect. Such calculations address a hypothetical future design under those assumptions. They do not change the uncertainty in the study's observed result, and they do not supply the study's original sample-size justification, which is not reported in the registry record.

As a general principle, a planning calculation based on an observed effect may be optimistic when the observed effect is imprecise or unusually large. It is best treated as a design aid rather than as evidence that the current trial was adequately powered. The registry does not report a prospective sample-size calculation, the assumptions used for the original enrollment target, or an interim-analysis plan.

What is not specified?

The registry record does not report the missing-data method, the definition of analysis populations, an intention-to-treat or per-protocol approach, or sensitivity analyses. It also does not identify the statistical test behind the mosquito-infectiousness p-value, the confidence-interval construction method, or a model for the primary counts. These omissions do not invalidate the posted summaries, but they restrict the extent to which the analyses can be reproduced or evaluated against a prespecified plan.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists multiple secondary measures spanning gametocyte density, area under the curve, commitment, sex ratio, and mosquito infectiousness. When several outcomes or comparisons are examined, the chance of obtaining at least one small p-value can increase. Statistical methods can address this multiplicity, for example by adjusting significance thresholds or controlling an error rate, but no such strategy is reported in the registry record.

The mosquito-infectiousness comparison has an unadjusted p-value of <0.001. It should be understood as the result for that individual comparison, without assuming that it has been adjusted for other outcomes or analyses. The registry does not report a hierarchy of endpoints, a procedure for selecting outcomes for formal testing, or a prespecified multiplicity plan. The primary outcome is posted as descriptive counts with no inferential comparison reported.

Missing-data handling is not reported in the registry record. The posted infectiousness outcome gives group sizes and participant counts, but the record does not describe whether any participants had missing assessments, how missing observations were handled, or whether all randomized participants were included in the analysis. The primary event counts likewise lack an accompanying account of follow-up completeness or event ascertainment. These gaps mean that the posted denominators and counts cannot be used to infer a missing-data strategy.

Analysis populations are also not reported in the registry record. It does not specify whether results represent all randomized participants, participants who received challenge, participants completing follow-up, or another population. The record does not provide an intention-to-treat definition, a per-protocol definition, or sensitivity analyses. Although randomization is listed, the analysis population and handling of protocol deviations are necessary to understand how randomized assignment relates to the reported estimates.

The record's available result is therefore best read as an unadjusted secondary comparison with incomplete information about analysis populations, missing observations, and multiplicity. These limitations affect the strength and reproducibility of statistical interpretation, but do not alter the values recorded for the posted outcome.

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 non-serious adverse event terms

The serious adverse-event summary reports zero affected participants in each of the four groups, with 6 at risk in each group. The registry also records zero deaths in every group. These figures describe the serious-event entries posted for the study; they do not establish that serious risks are absent. In a small study, an event that was not observed can still be possible, and the registry does not provide uncertainty intervals for these safety figures.

The separate count of other adverse events indicates 6 affected participants in each group, with 6 at risk in each. These participant-level safety entries should not be conflated with the primary endpoint's adverse-event counts of 95, 95, 107, and 52. The primary outcome is an event count, whereas the safety summary distinguishes participants affected by serious or other adverse events. The measures therefore have different units and should not be compared as though they described the same quantity.

The registry does not provide serious-event terms, severity grades, relatedness assessments, or detailed timing in the safety summary. The absence of serious-event entries in the posted record is a descriptive observation within this study, not a guarantee about safety in other settings or populations.

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

The result supports a limited statistical conclusion: in the posted mosquito-infectiousness comparison, Cohort B had more participants recorded as infectious for mosquitoes through DFA than Cohort A, and the estimated risk difference is 0.75 with a confidence interval from 0.505 to 0.995. The unadjusted p-value is <0.001. In this model-development study, that pattern indicates that the two challenge cohorts produced different recorded infectiousness outcomes.

It does not establish that one drug regimen blocks transmission, because the reported comparison is between challenge cohorts rather than a regimen-specific comparison. It also does not establish a causal effect of challenge route on its own: the registry does not report whether cohort assignment was randomized, how the analysis accounted for regimen groups, or whether other design features were adjusted. The study's randomized designation should not be extended automatically to a contrast whose assignment mechanism is not described.

The result does not show that a vaccine, monoclonal antibody, or other biologic prevents transmission. Those intervention classes are part of the model's intended future application, not interventions evaluated in the posted comparison. Nor do the primary adverse-event counts answer the transmission question; they describe event totals without a reported statistical comparison or participant-level distribution.

The findings are specific to the study's controlled human infection setting and the outcomes as recorded through the stated follow-up. The registry does not report demographic characteristics or detailed eligibility information, which further limits assessment of generalizability. The result should not be taken as a prediction of infectiousness or treatment response in other populations or settings.

In practical statistical terms, the posted outcome provides evidence that the model produced different mosquito-infectiousness observations across the two cohorts. Precision is represented by the interval, while the study design and reporting gaps limit causal and external interpretation. Results for the other listed secondary outcomes are not reported in the registry record, so they cannot be used to build a broader claim about gametocyte biology or transmission blocking.

13. Why This Trial Matters Statistically

This study is statistically notable because its central purpose is model development. A clinical challenge model must do more than record biological measurements: it needs to generate outcomes relevant to the intended research question. Mosquito infectiousness through DFA is directly connected to the proposed use of the model for evaluating transmission-blocking interventions, making the cohort comparison relevant to whether the model can produce a measurable transmission-related endpoint.

The design also illustrates why the exact comparison matters. Cohort A and Cohort B differed in challenge route, while each cohort included groups with different subsequent regimens. A result comparing the cohorts may be useful for understanding the model, but it is not automatically a clean estimate of a treatment effect or a route effect. Statistical interpretation depends on which factor was randomized, how assignments were made, and whether relevant design features were incorporated in the analysis. Those details are not reported in the registry record.

The distinction between an event count and a participant count is another important statistical feature. The primary endpoint reports adverse-event totals, while the safety summary separately reports participants affected. These quantities answer different questions, and combining them would obscure whether the focus is event burden or the number of people experiencing events. The record's descriptive primary results and its inferential secondary comparison also illustrate that not every endpoint has the same level of statistical support.

Finally, the study highlights the role of uncertainty and multiplicity in small model-development investigations. The mosquito-infectiousness estimate has a reported interval and an unadjusted p-value, but the record does not describe multiplicity control, missing-data handling, or the analysis population. Such limits do not erase the observed pattern; they define how far the evidence can reasonably be taken. Distinguishing model feasibility from intervention efficacy is central to a careful reading of these results.

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 contains limited information about participant characteristics, eligibility criteria, recruitment, and the basis for enrollment. Without those details, it is difficult to assess how representative the participants were or how broadly the findings might apply. The record lists a single-center study, which further anchors interpretation to the study setting rather than establishing performance across settings.

Important design details are absent. Randomization is recorded, but the randomization procedure and allocation concealment are not reported. It is also not clear from the available record whether assignment to challenge cohort was randomized. The study is open label, and the record does not describe outcome-assessor masking. These gaps matter particularly for interpreting comparisons between cohorts and for considering whether knowledge of assignment could affect outcome measurement or reporting.

The primary adverse-event endpoint is reported as group counts, without a participant-level distribution, uncertainty estimates, or a formal comparison. The registry does not specify how events were defined, classified, or assessed for relatedness. The safety summary provides affected-participant counts but not event terms or severity detail. Thus, the safety information is useful as a record of posted counts but does not support a detailed assessment of event patterns.

For the secondary infectiousness comparison, the record does not identify the statistical test or confidence-interval method, and it does not report adjustment for regimen or other design features. The result is unadjusted, and no multiplicity plan is documented. Missing-data procedures and analysis populations are also not reported. Numerical results for the other listed gametocyte outcomes are absent from the registry record, limiting assessment of whether the model's biological measures formed a coherent pattern.

The arm descriptions are incomplete where they describe conditions for the subsequent treatment. The registry therefore does not provide enough information to reconstruct all treatment timing and eligibility details. These limitations are specific to the available record and constrain reproducibility and interpretation; they should not be replaced with assumptions about the protocol.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Controlled Human Malaria Infection Model for Evaluation of Transmission-blocking Interventions - Study 2 — NCT03454048.
• Registry record last updated 2020-07-07; status last verified 2020-03.
• Results first posted 2020-03-31.
• 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.

18. Record Summary

NCT03454048 is a completed, randomized, open-label, parallel, single-center study intended to establish a controlled human malaria infection transmission model. Enrollment is listed as 24. Cohort A received mosquito-delivered challenge, and Cohort B received blood-stage challenge by intravenous injection. Groups within each cohort received regimens involving low-dose piperaquine followed by piperaquine or sulfadoxine pyrimethamine.

The primary endpoint was adverse-event frequency up to day 51 after challenge infection. Posted group counts were 95, 95, 107, and 52, without a reported inferential comparison. For the secondary mosquito-infectiousness outcome, the record reports 0 of 12 in Cohort A and 9 of 12 in Cohort B, a risk difference of 0.75 with a confidence interval from 0.505 to 0.995 and an unadjusted p-value of <0.001. The registry also lists gametocyte outcomes, but their numerical results are not reported in the registry record. Details of multiplicity, missing-data handling, analysis populations, and the statistical analysis plan are not reported.