Experimentation on Human Subjects – Comprehensive Study Notes

Context & Evolution of Human-Subject Research

Centuries of medical practice = conservative, innovation ≈ malpractice risk
$19^{th}-20^{th}$ c. rise of “properly scientific” medicine → experimentation viewed as both
- scientific driving force
- ethical imperative (duty to advance care)
But successes “not unalloyed” → repeated ethical crises when researchers forgot that
participants = persons w/ equal moral worth

High-Profile Scandals That Shaped U.S. Policy

Common pattern across scandals
- scientifically weak or redundant design
- deception/coercion of marginalized groups
- subjects denied existing care
- public outrage ⇒ regulatory change

1. Nazi Concentration-Camp Experiments (1933-45)

Freezing, high-altitude, gangrene, typhus vaccine trials, bone transplants
Featured deliberate killing/torture for data
Nuremberg Doctors’ Trial (1947) ⇒ 7 death sentences; judges articulate 10-point Nuremberg Code
- Principle #1: “Voluntary consent … absolutely essential”

2. Jewish Chronic Disease Hospital (Brooklyn, 1963)

22 debilitated elderly injected w/ live cancer cells by Dr. Chester Southam
- Purpose: test if delayed rejection due to cancer vs. debility
- No disclosure of “live cancer cells”; relied on paternalistic claim of “no risk”
Three junior physicians refused → resignations → Board of Regents hearing (1965)
- Southam & Dr. Mandel censured; licenses suspended (stayed) for “fraud, deceit & unprofessional conduct”
- Case crystallized informed-consent & competence debates

3. Willowbrook Hepatitis Studies (NY, 1956-71)

Mentally-retarded children intentionally infected w/ hepatitis; some pre-given gamma-globulin
Researchers claimed: infection inevitable in overcrowded wards; offered “milder strain” + data
Ethical charges
- Coercive admission policy (research participation = bed)
- Deceptive consent forms
- State failed duty to improve conditions rather than exploit them as laboratory

4. Tuskegee Syphilis Experiment (1932-72)

>400 poor Black men w/ advanced syphilis in Macon Cty., AL; told they had “bad blood”
Denied penicillin post-1943; autopsy stipends from Milbank Fund used as inducements
Peter Buxtun (PHS employee) whistle-blows (1972) → press exposé, Senate hearings, class-action suit $\,($ 1973)$
Outcomes
- 1974: National Research Act → National Commission → Belmont Report (1979)
- Long-term mistrust of govt./medicine in African-American communities; cited in HIV outreach resistance

Regulatory & Institutional Milestones

1947 Nuremberg Code (10 principles)
1966 U.S. Surgeon General PHS guidelines → local peer review of federally funded studies
1974 National Research Act → Institutional Review Boards (IRBs) mandated; National Commission formed
1979 Belmont Report articulates 3 governing principles (respect, beneficence, justice)
1990s Common Rule (45 CFR 46) harmonizes federal requirements

1980s–90s Paradigm Shift: From Protectionism to Inclusion

Earlier assumptions: research = risky; slow progress acceptable; vulnerable groups (women, minorities, children) should be shielded
AIDS & women’s-health activists + conservative libertarians:
- Viewed research ACCESS as benefit, paternalistic barriers as injustice
- Led to FDA “parallel-track” & accelerated approvals; NIH Revitalization Act (1993) mandates inclusion of women & minorities

Core Ethical Frameworks

Nuremberg Code (Selected Points)

Voluntary consent w/out coercion
Yield fruitful results unprocurable by other means
Avoid unnecessary suffering
Risk ≤ humanitarian importance of problem
9–10. Subject or scientist may terminate experiment upon risk emergence

Belmont Report Principles

Respect for Persons
- Autonomy + protection for diminished autonomy (children, cognitively impaired, prisoners)
- Operationalized via informed consent (information, comprehension, voluntariness)
Beneficence
- \text{Rule 1: }\text{Do no harm}
- \text{Rule 2: }\text{Maximize benefits; minimize risks}
- Requires systematic risk/benefit assessment
Justice
- Fair distribution of burdens/benefits; beware routine use of “convenient” disadvantaged populations
- Individual justice (no cherry-picking “good” patients); social justice (avoid macro-level exploitation)

Randomized Clinical Trials (RCTs) – Ethical Tensions

Scientific rationale: randomization removes bias → reliable data; long-term patient benefit
Critiques (Hellman & Hellman; Markman)
- Randomization may override physician’s fiduciary duty to individual
- Kantian concern: treating present patient as means to future ends
Concept of Equipoise (Freedman)
- Ethical RCT only when honest professional disagreement exists
- Personal “strong hunch” ≠ sufficient to stop trial; community uncertainty is key

Vulnerability & “Less-Than-Best” Standard of Care Debates

Problem: Poor settings lack infrastructure to deliver best available therapy; should trials test cheaper/shorter regimens?
Case 1 AZT “short-course” mother-to-child HIV trials (Africa/Asia, 1990s)
- Compared low-dose AZT to placebo despite known efficacy of 076 protocol (reduces transmission 30\% \to 8\%)
- Critics (Angell; Lurie & Wolfe): placebo unethical, mirrors Tuskegee; sponsors must provide global standard
- Defenders (Crouch & Arras): placebo acceptable if equipoise exists locally & trial addresses host-community need; but must plan post-trial access
Case 2 Kennedy-Krieger Lead Paint Study (Baltimore, 1990s)
- Low-income families housed in partially abated units to test cheaper lead remediation
- Maryland court likened child exposure to non-therapeutic research w/ > minimal risk; sparked debate on “minimal risk” definition & surrogate consent

Pediatric Research – Protection vs Access

>50\% of U.S. pediatric drug use historically “off-label”; FDA 1997 list shows millions of annual prescriptions without pediatric data
1994 onward: FDA & NIH rules → mandatory pediatric inclusion or justified exclusion
Wendler’s 4-prong test for trials offering < best care: addresses host need; design indispensable; net community benefit; participants not worse off than no-trial world

Applications of Belmont Principles (Consent, Risk/Benefit, Subject Selection)

Informed Consent Elements
- Complete, tailored information (allowing for necessary, limited deception → debrief)
- Comprehension checks; adapt to language, maturity, cognition
- Voluntariness: free of coercion (threats) & undue influence (excessive rewards or authority pressure)
Risk/Benefit Assessment
- Identify probability & magnitude of physical, psychological, social, economic harms
- Rule-outs: Brutal treatment never justified; risks must be necessary & minimized
Justice in Recruitment
- Avoid routine targeting of easily manipulable groups (e.g., prisoners, economically disadvantaged)
- Ensure that those who bear research burdens can realistically share in its fruits

Selected Numerical / Statistical References (in …)

1{,}600 $children infected w/ HIV daily world-wide;$ 90\% in developing nations
AZT 076 regimen ↓ vertical transmission: 30\% \rightarrow 8\%
FDA (1997) outpatient pediatric prescriptions lacking adequate labeling: Albuterol (1{,}626{,}000 $< age 12), Phenergan ($ 663{,}000 < age 2), etc.

Condensed Tuskegee Timeline (key entries)

1932 PHS begins non-therapeutic study; men not told they have syphilis
1933-34 $Spinal taps; addition of ~$ 200 control subjects
1942-43 PHS blocks draftees’ treatment; penicillin introduced elsewhere
1969 CDC panel (all but 1 tied to study) votes to continue
1972 Buxtun leaks to press → study halted
1997$$ President Clinton issues formal government apology

Lexicon & Key Concepts

Equipoise: state of genuine professional uncertainty → ethical pre-req for RCT
Placebo-controlled trial: arm receiving inert agent; ethical only if no proven therapy OR compelling methodological need + ancillary protections
“Standard of Care” debate: de jure (what expert community knows is effective) vs de facto (what is actually available locally)
Minimal Risk (federal definition): harm/discomfort ≈ daily life or routine exams of average healthy child
Therapeutic vs Non-Therapeutic research: primary intent = patient benefit vs knowledge generation; boundaries often blurred