Thank you for taking the time to write such a detailed response. I’m replying directly and point-by-point, because while I appreciate the care with which you’ve articulated your position, I remain fundamentally unconvinced that the ethical and scientific case you’re advancing holds.

1) On what the trial is actually comparing

I am clear that this is not “HBV0 vs no newborn vaccines.” That has never been the claim. The ethical question is narrower and sharper: whether it is acceptable to randomize newborns to not receive a hepatitis B birth dose in a high-prevalence setting where the birth dose is globally recommended as standard prevention. The presence of BCG and OPV in both arms does not neutralize that concern. It simply changes the framing.

2) On “overall health effects” and the manufacture of equipoise

You argue that because no trials have specifically examined early all-cause morbidity/mortality outcomes for BCG+OPV versus BCG+OPV+HBV0, there is genuine uncertainty that justifies randomization. I disagree.

Absence of data on a speculative harm mechanism is not the same as equipoise when weighed against well-established, biologically coherent, and epidemiologically robust benefits of HBV birth dosing for preventing perinatal infection and downstream chronic disease. The hypothesis that an aluminum-adjuvanted, non-live vaccine meaningfully attenuates early survival benefits conferred by live vaccines remains unproven, contested, and context-dependent. Elevating that hypothesis to a level sufficient to justify withholding a proven preventive intervention is not scientific neutrality, it is a value judgment about which risks matter more, and to whom.

3) On non-specific effects of BCG and OPV

I do not dispute that BCG and OPV may have beneficial non-specific effects, nor that these effects have been observed in multiple settings. What remains unresolved and this is critical, is whether those findings are sufficiently consistent, causal, and generalizable to justify delaying or denying another WHO-recommended intervention at birth. The leap from “live vaccines may reduce early mortality” to “therefore we should test withholding HBV0” is not compelled by the data. It is an interpretive choice with negative downstream consequences in a study like yours for the children who do not receive the intervention.

4) On the ACIP rapid review

Low certainty or inconclusive evidence in a rapid review does not constitute evidence of harm, nor does it establish equipoise for withholding a vaccine. It reflects limitations in the existing literature, not a signal that the intervention is unsafe. Using that uncertainty to justify an individual-level randomized delay of HBV0 in African newborns would be ethically untenable in many other settings and that asymmetry matters.

5) On “Africa needs RCTs”

I agree entirely that Africa needs more randomized trials. But JUSTICE in research is not satisfied by conducting trials in Africa; it requires careful scrutiny of which questions are asked and how they are answered. “We already know it works elsewhere” is indeed an insufficient argument in many domains. But neither is “we do not know everything about possible interactions” a sufficient reason to expose African newborns to delayed protection from a virus that causes lifelong morbidity and mortality disproportionately on the continent. This is actually the foreign gaze in action. a problematic phenomenon in global health research. I recommend strongly reading Dr. Seye Abimola's seminal writing on this.

6) On endpoints and interpretability

A composite endpoint of mortality and severe infectious morbidity in the first 42 days is highly vulnerable to contextual effects like care-seeking behavior, admission thresholds, surveillance intensity, and post-hoc definitional flexibility. When paired with a study team that is understandably invested in a particular biological narrative about vaccine interactions, this raises real concerns about interpretability, even with DSMB oversight.

7) On safeguards and “net benefit”

Ethics approvals, consent processes, insurance coverage, and improved weekend vaccination access are baseline requirements, not ethical offsets. The argument that half of participants receive HBV0 while the other half receives what they “would probably have received anyway” effectively concedes the core concern: the trial design preserves existing gaps in protection rather than correcting them. You are exploiting scarcity and vulnerability of the context to conduct your study which in itself is unethical and deeply problematic. If trial infrastructure can improve timely delivery of BCG and OPV, it could just as readily improve delivery of HBV0 for all newborns.

8) On pre-implementation randomization

Pre-implementation periods are not ethical blank slates. They do not automatically justify individual-level randomization to non-receipt of a recommended intervention, particularly when the population involved bears a disproportionate burden of the disease in question. Equipoise cannot be defined solely by uncertainty in ancillary outcomes while sidelining well-established primary benefits.

At the core, my concern is not about your intentions, nor about the importance of studying vaccine interactions. It is about a recurring pattern in global health research: uncertainty is tolerated, even valorized, when it delays interventions for African populations, while far higher evidentiary thresholds are implicitly applied elsewhere. That asymmetry, more than any single methodological detail is what I find ethically and morally troubling.

I’m open to continued discussion about appropriate evidentiary thresholds and alternative study designs that could answer important biological questions without withholding a globally recommended birth-dose intervention. But I do not agree that the current design meets that bar. It will cause real irreversible harm to vulnerable children.

Best regards,

Boghuma