The Foster Children, the HIV Trials and the Records Investigators Couldn’t See

The history of New York City foster children in HIV research contains a problem that cannot be resolved by choosing between two sweeping narratives. An investigation found failures in oversight and documentation. It also failed to substantiate some of the gravest accusations. And investigators could not examine every record they needed.

In 2005, the city’s Administration for Children’s Services commissioned the Vera Institute of Justice to review what had happened. After examining child-welfare files for 796 children, Vera identified 532 who had participated in 88 HIV/AIDS clinical trials or observational studies. Its report was published in 2009. Vera’s report overview

That distinction between types of research matters. Observational research follows participants; an intervention trial tests a treatment or other intervention. Calling every enrollment an experimental drug exposure obscures what each child actually experienced.

The medical question and the consent question

One study in the history was ACTG-218, also called PACTG 218. Its published report describes a randomized, blinded trial of recombinant HIV envelope proteins in children already infected with HIV. The investigators examined safety and immune responses. These were protein-based vaccine candidates, not mRNA vaccines. The original 1998 study

The researchers reported no adverse events attributed to the study vaccines and no significant difference in disease progression between the groups. They found immune responses, but an immune response is not the same as a demonstrated clinical benefit. Those findings concern the study population and design; they do not answer who was authorized to enroll a particular foster child. Published results

A trial can ask a medically legitimate question while its consent arrangements deserve scrutiny. Conversely, a failure in consent documentation does not establish that the intervention caused a particular injury. Investigating one issue should not erase the other.

A policy change opened a route to enrollment

Vera describes an initial decision against approving foster-child participation in PACTG 218. Following a 1994 change to the city’s joint-guardianship policy, contract foster-care agencies could authorize some enrollments without prior city approval. Vera identified three foster children enrolled in PACTG 218 through that provision. Final report, printed pages 152–155

This is more specific than a claim that researchers simply ignored a ban. It identifies a change in the process and the authority to consent. It also poses a serious governance question: when a public institution declines approval, what safeguards should govern a different institution’s ability to authorize the same participation?

The question concerns the children’s protections, not just the signatures on a form. A record should make it possible to reconstruct which authority applied, what was explained and why participation was considered appropriate for the child.

What the deaths do—and do not—show

Vera reported that 80 participating children died while in foster care; 25 died while enrolled in medication trials. Its medical reviewers did not identify a death directly caused by trial medication. The executive summary also described serious toxicities in some children, including severe anemia and reduced liver function, and documented treatment adjustments. Vera’s executive summary

Those findings cannot responsibly be compressed into a claim that experimental drugs killed 80 children. They also do not amount to proof that every treatment was harmless. Death during a period of care, death during a trial and death caused by an intervention are different categories.

The review’s limits matter to any causal conclusion. A retrospective investigation can only assess what the available evidence permits. Its inability to establish a particular harm is not permission to invent that harm, and its incomplete access is not permission to declare every concern resolved.

Missing records were an accountability failure

A resolution introduced in the New York City Council in 2009 described consent forms as missing in roughly one-fifth of cases and identified failures in required review procedures. It sought access to medical and clinical-trial records. The legislative history shows that the resolution was introduced and referred to committee, then filed at the end of the session; it was not adopted. Resolution 2192-2009

Missing consent documentation is not identical to proof that consent never occurred. It means that the records examined did not establish it. That is still consequential where the subjects are children whose decisions are being made by adults and institutions responsible for their welfare.

Vera said the New York State Department of Health declined requests to authorize access to medical and research records. It recommended either permitting access or conducting a state investigation. That restriction limited the review’s ability to resolve outstanding questions. Executive summary, records-access recommendation

The report describes the access problem at the time of its investigation. This article does not establish that the same restrictions remain in effect today.

Where Fauci fits in the record

Anthony Fauci directed the National Institute of Allergy and Infectious Diseases from 1984 to 2022 and oversaw a research portfolio that included HIV/AIDS. That is the documented institutional context for his connection to the era. NIAID biography

The sources reviewed here do not establish that he personally selected these foster children, authorized their individual enrollments or caused their deaths. His leadership role is a legitimate subject of institutional history. It cannot substitute for evidence of a specific personal act.

Accountability becomes clearer when the questions are assigned to the decisions at issue: research sponsorship, protocol approval, guardianship, consent, treatment monitoring and preservation of records. Reducing that chain to one famous name can obscure the responsibilities the documents actually reveal.

The lasting lesson is demanding but concrete. Vulnerable participants need access to potentially beneficial research and protections that can be demonstrated afterward. The New York review shows why both matter—and why neither unsupported accusations nor missing paperwork should be allowed to end the inquiry.

DOCUMENTATION

SOURCES & RECEIPTS

Sources checked September 26, 2026. Links open in a new tab; some publishers restrict access.

1. The Experiences of New York City Foster Children in HIV/AIDS Clinical TrialsVera Institute of Justice • report overview, 2009VIEW SOURCE →
2. Safety and immunogenicity of HIV recombinant envelope vaccines in HIV-infected infants and childrenLambert and colleagues • ACTG-218, 1998; PMID 9859958VIEW SOURCE →
3. Final report, with appendicesVera Institute • especially printed pages 101 and 152–155VIEW SOURCE →
4. Executive summaryVera Institute • findings and limitations, 2009VIEW SOURCE →
5. Resolution 2192-2009: records access proposalNew York City Council • introduced September 30, 2009; filed at session endVIEW SOURCE →
6. Anthony S. Fauci, M.D.NIAID • former director biographyVIEW SOURCE →