Seven thousand infants. That is the number of newborns who were assigned to the placebo arm of a hepatitis B vaccine trial in Guinea-Bissau, a country where approximately 20% of mothers unknowingly carry the virus and can transmit it to their children at birth. The researchers did not withhold an experimental treatment. They withheld a vaccine with decades of confirmed efficacy, one the WHO recommends for every newborn on Earth, to study questions about its "broader health effects." The ethics of that design are not ambiguous. The trial was wrong, and it should not resume.

When "We Need More Data" Becomes a Cover Story

\p>The concept the trial violates is called equipoise: the genuine scientific uncertainty that must exist before you can ethically randomize patients. Think of equipoise as a scale. When the evidence tilts heavily to one side, you cannot pretend the scale is balanced just because you want to weigh something new. For the hepatitis B birth dose, the scale has not been balanced for decades. Paul Offit, director of the Vaccine Education Center at Children's Hospital of Philadelphia, put it plainly: existing evidence shows the vaccine prevents mother-to-baby transmission with minimal side effects. There is no legitimate equipoise here. The CDC's defense that this is a "gold-standard, independent study" collapses the moment you ask what scientific uncertainty justifies the placebo arm.

The funding trail makes this worse. A $1.6 million no-bid CDC contract went to the University of Southern Denmark's Bandim Health Project in December 2025. No-bid contracts already carry a transparency problem; no-bid contracts funding research that no U.S. institutional review board would approve carry a different problem entirely. Jessica Malaty Rivera of Defend Public Health named it directly: "No ethical board in the U.S. would ever approve of something like this. Of course they would go to a place where they can get away with it." That sentence should disturb anyone who thinks international research standards mean something.

The local ethics approval was compromised in ways that are almost too specific to be believed. The Guinea-Bissau committee used the signature of a president who had already resigned. Only 4 of the 6 required members were active at the time of approval. SDU's own Dean of Health Sciences, Ole Skøtt, acknowledged "issues relating to conflicts of interest" in the local committee's approval process. The Bandim Health Project, which operates inside Guinea-Bissau, reportedly influences the very government bodies that were supposed to provide independent oversight.

The Pattern That Makes This So Familiar

I will grant the opposing argument one fair point: understanding the full immunological effects of co-administered vaccines in low-income settings is a legitimate scientific question. Studies examining non-specific vaccine effects have produced genuinely interesting findings. But "interesting question" does not license any methodology. You cannot study whether aspirin prevents headaches by giving half your participants a migraine. The question is real; the design is still wrong.

The comparison to Tuskegee, raised by the Genetic Literacy Project in March, deserves careful handling. The parallel is not perfect. Penicillin was withheld from Black men with syphilis for decades after it was confirmed as curative. This trial ran for months before suspension. The scale differs. But the structural logic is identical: researchers decided that men and infants in vulnerable, politically unstable populations could bear a risk that Western ethics boards would never impose on Western patients.

Guinea-Bissau's new health minister suspended the trial in February 2026 following a November coup. The University of Southern Denmark placed it on full hold on March 19, 2026. Neither action resolves the underlying problem. Danish law only requires notification if research procedures occur in Denmark. Congress should require that any federally funded trial abroad meet U.S. IRB standards, full stop. The gap in that requirement is exactly what this trial exploited, and it will be exploited again unless someone closes it.