Fourteen Thousand Newborns and a Six-Week Lottery
The Guinea-Bissau vaccine trial is enrolling 14,000 newborn infants in a randomized experiment that withholds the hepatitis B birth dose — a vaccine recommended by every major international health body — from half its participants for six weeks. Of those 14,000 infants, 7,000 receive the vaccine at birth. The remaining 7,000 wait. Hepatitis B can survive on a dry surface for seven days.
That delay is not a technicality. The hepatitis B birth dose exists precisely because the neonatal period is the single most dangerous window for acquiring a chronic infection. A newborn who encounters the virus does not fight it the way an older child or adult does. The younger the exposure, the higher the probability of chronic infection — and chronic infection is what quietly seeds liver cancer and cirrhosis decades later.
Hepatitis B is not a fragile or reclusive virus. It survives on surfaces for up to seven days and can transmit through contact that would barely register as a risk for most pathogens. The birth dose was designed specifically to close that gap before the world has a chance to reach the baby.
Withholding it for six weeks does not pause the clock. It simply removes the shield while the clock keeps running.
The Hypothesis That Started This: When Vaccines Do Things Nobody Planned
A vaccine trains your immune system to recognize one enemy. That is the standard story, and for most vaccines it is accurate enough. But in the 1980s, working in Guinea-Bissau's impoverished Bandim district, Danish researcher Dr. Peter Aaby noticed something the standard story could not explain: children who received the measles vaccine seemed to be dying less often from everything, not just measles. This observation gave birth to a field of inquiry called non-specific effects, or NSEs — the idea that some vaccines reshape the immune system's general alertness, reducing all-cause mortality beyond the pathogen they were designed to fight.
Dr. Aaby and his longtime collaborator Dr. Christine Stabell Benn at the University of Southern Denmark have built decades of research on this premise. Their specific claim, relevant to the current trial, is that live vaccines may lower overall mortality while inactivated vaccines, like the hepatitis B birth dose, might paradoxically raise it in certain contexts. The hypothesis is genuinely interesting. Science has surprised us with stranger things.
The problem is that "interesting hypothesis" and "confirmed finding" are different categories. A 2025 independent review of 13 previous trials by these same researchers found no credible evidence of non-specific effects and raised pointed concerns about data selection — which results were included, which were not. That is not a minor procedural complaint. Selective data is the quiet mechanism by which a hypothesis that should be abandoned instead accumulates the appearance of support.
An Unusual Coalition: Who Is Funding the Hepatitis B Birth Dose Study
Follow the money and the picture gets stranger. The US Centers for Disease Control and Prevention has allocated $1.6 million in federal grant funding to the trial — taxpayer dollars flowing to a study that withholds a proven vaccine from African newborns. That number is not a typo. At roughly $0.22 per hepatitis B dose, $1.6 million could vaccinate more than seven million infants. Instead, it is funding a randomized experiment on fourteen thousand of them.
The CDC grant sits alongside private backing from two foundations: the Pershing Square Foundation and the Bluebell Foundation. Pershing Square was co-founded by billionaire hedge fund manager Bill Ackman. What unites a federal public health agency, a Wall Street philanthropist, and a small European foundation around a single trial in Guinea-Bissau is not immediately obvious.
Robert F. Kennedy Jr., serving as US Secretary of Health and Human Services, has described the trial explicitly as a prototype for his vaccine safety research agenda. A sitting US Health Secretary is pointing to a foreign study, on newborns in one of the world's poorest countries, as the model for how his department intends to interrogate vaccines going forward. This is stated policy — and it tells you something important about what this trial actually is: less a scientific expedition into immunology, more an opening move in a domestic political argument about vaccines, played out at real cost, far from home.
The Ethics Committee Walks Out
Nine months before the October approval, the trial was already dead. In January 2026, an eruption of international criticism forced a suspension, and for a brief window it seemed the scientific community's alarm had worked. Then the University of Southern Denmark's ethics committee reconvened, reviewed a revised protocol, and voted yes.
Four of its members immediately resigned.
One of them, Stig Børsen Hansen, chose to say exactly why. "You are now enrolling newborns in a lottery to test a hypothesis," he told the Guardian. The Helsinki Declaration, the foundational document of research ethics since 1964, is unambiguous on one point: populations who cannot give informed consent require extra protection, not reduced protection. A newborn cannot consent. A newborn cannot weigh risk. What the revised protocol offered them was a draw of the cards.
The WHO issued a formal statement of significant concern. The Africa CDC had already opposed the study on ethical grounds before the suspension, its objections quiet but on record. These were the two largest global health oversight bodies on earth, and both had said: stop.
A sitting US Health Secretary is pointing to a foreign study, on newborns in one of the world's poorest countries, as the model for how his department intends to interrogate vaccines going forward.
The revised protocol added one new safeguard — mothers would be tested for hepatitis B, and infants born to positive mothers would be vaccinated immediately and excluded from the trial. It was a genuine improvement. Hansen's point was that it was not enough.
The Five-Percent Crack in the Safeguard
When international pressure forced a pause in January 2026, the researchers went back to work on the protocol. The revised version added a screening step: test every mother for hepatitis B, exclude any infant whose mother tests positive, and vaccinate that infant immediately. Health Minister Quinhin Nantote greenlit the revised study, and the trial moved forward. On paper, the protection looks clean.
The problem lives in the test itself.
The hepatitis B screening tool used in the trial carries an estimated five-percent false negative rate — meaning roughly one in twenty infected mothers will be classified as safe, their infants assigned to the unvaccinated control group without a flag. Compare this to a fire door with a five-percent chance of staying locked during a fire: acceptable in a risk-benefit calculation, perhaps, but not when you are the family on the other side. The revised protocol does not eliminate the danger of mother-to-child transmission; it filters most of it out and lets a remainder through.
What makes that remainder harder to justify is a single number: a hepatitis B birth dose costs twenty-two cents. Not twenty-two dollars. Twenty-two cents. Any argument rooted in resource constraints collapses against that figure. The crack in the safeguard is not a matter of what is possible. It is a matter of what was chosen.
What We Are Actually Testing
Guinea-Bissau plans to introduce universal birth-dose vaccination by 2028. That two-year window is not a fixed feature of the country's public health landscape; it is a scheduled construction zone. The researchers' argument, stripped to its skeleton, is this: since the policy does not yet exist, withholding the vaccine withholds nothing.
Critics have a word for that logic. The Tuskegee comparison, made by several global health ethicists, is uncomfortable to hear and structurally difficult to dismiss. In both cases, a vulnerable population in a resource-limited setting received less than the available standard of care, so that researchers from wealthier institutions could gather data they could not legally collect at home. The parallel is not perfect. It never is. But the architecture is recognizable.
The honest question buried under all of this is whether a country's infrastructure gap is an ethical research opportunity or an ethical boundary. The Guinea-Bissau vaccine trial has not answered that question. It has simply chosen a side — and fourteen thousand newborns are living with the consequences of that choice.
