Ethics, Communication & Professionalism

Research ethics: IRB and informed consent for research

Clinical Overview and When to Suspect Research Ethics Issues

— 45 CFR 46 (Common Rule) governs federally funded human subjects research; revised 2018 ("Revised Common Rule")

— 21 CFR 50, 56 govern FDA-regulated research (drugs, devices, biologics)

— ICH-GCP E6(R2) is the international Good Clinical Practice standard

— Is it research? = systematic investigation designed to develop or contribute to generalizable knowledge

— Does it involve human subjects? = living individuals from whom data are obtained through intervention/interaction, or identifiable private information/biospecimens

— Pure QI projects intended only for local improvement

— Case reports of ≤3 patients (institution-dependent)

— Public health surveillance authorized by a public health authority

— Classroom educational exercises

— Journalism, biography, oral history

— Investigator is also the treating physician and pressures enrollment

— Consent obtained by a junior team member without proper delegation

— Protocol deviation involving a vulnerable subject (prisoner, child, cognitively impaired)

— Adverse event not reported within required window

— Investigator with undisclosed financial conflict of interest

— Data published without IRB approval of the protocol

Board pearl: If a vignette says "we want to share our findings at a national conference" or "publish in a peer-reviewed journal," the activity is presumed generalizable and therefore research — IRB review is required before data collection begins, not retroactively. Retroactive IRB approval is not permitted under the Common Rule.

Key distinction: QI = improve local care; Research = create generalizable knowledge. Intent at project initiation determines classification.

Core framework: Modern US research ethics rests on three Belmont principles — respect for persons (autonomy, informed consent), beneficence (favorable risk-benefit), and justice (fair subject selection and benefit distribution)

Regulatory backbone:

When does an activity require IRB review? Ask two questions:

Activities NOT requiring full IRB review (Step 3 favorites):

When to suspect a research-ethics violation in a vignette:

Presentation Patterns and Key History — Historical Context Shaping Modern Rules

— Response to Nazi physician experiments

— Introduced voluntary consent as absolutely essential

— Subject must have legal capacity, free power of choice, sufficient knowledge

— World Medical Association

— Distinguished therapeutic vs non-therapeutic research

— Required independent committee review — direct ancestor of the IRB

— 600 Black men in Alabama, 399 with latent syphilis, deceived and denied penicillin after it became standard in 1947

— Led directly to the National Research Act (1974) → creation of IRBs and the Belmont Report (1979)

— 2018 changes: single IRB for multisite studies, broad consent for future biospecimen use, new exempt categories, shorter consent forms with key information up front

— "Subjects not told they had a disease" → Tuskegee parallel → deception/consent failure

— "Institutionalized children" → Willowbrook parallel → vulnerable population

— "Tissue used without permission for commercial purposes" → HeLa/Moore v. Regents parallel → biospecimen consent

Board pearl: Moore v. Regents of UC (1990) held that patients do not retain property rights in excised cells used for research, but physicians have a fiduciary duty to disclose financial interests during consent. The exam favors the disclosure obligation.

Step 3 management: When a vignette describes potential historical-style abuse, the correct first step is almost always halt enrollment and notify the IRB — never "complete the study then report."

Why Step 3 tests this: Landmark abuses produced each layer of current regulation; recognizing the pattern often unlocks the answer

Nuremberg Code (1947):

Declaration of Helsinki (1964, multiple revisions):

Tuskegee Syphilis Study (1932–1972):

Willowbrook hepatitis studies, Jewish Chronic Disease Hospital case, Henrietta Lacks (HeLa cells): prompted scrutiny of vulnerable populations and biospecimen consent

Common Rule (1991), Revised Common Rule (2018):

Vignette cues pointing to a specific era/violation:

Physical Exam Findings — Anatomy of a Valid Informed Consent

— Capacity — subject can understand, appreciate, reason, and express a choice

— Voluntariness — free from coercion, undue influence, or manipulation

— Disclosure + Understanding — material information communicated in lay language (typically 8th-grade reading level)

— That the activity is research and participation is voluntary

— Purpose, expected duration, procedures

— Reasonably foreseeable risks/discomforts

— Reasonably expected benefits

— Appropriate alternative procedures or treatments

— Confidentiality protections and limits

— Compensation/medical treatment if injury occurs (for >minimal-risk research)

— Contacts for questions about the research, rights, and injury

— Statement that subject may withdraw at any time without penalty

— For applicable studies: notice that data will be posted on ClinicalTrials.gov

— Whether identifiable data/biospecimens may be used for future research

— Statement about commercial profit from biospecimens

— Whether clinically relevant results will be returned

Key distinction: Therapeutic misconception — subject believes research procedures are designed for their personal benefit (like clinical care). Investigators must explicitly counter this during consent. A vignette where a cancer patient says "I'm in the trial because Dr. X said it will cure me" signals therapeutic misconception requiring clarification before enrollment proceeds.

Valid research informed consent has three structural elements plus eight required disclosures under the Revised Common Rule

Three structural pillars:

"Key Information" section (2018 Revised Common Rule): must appear at the beginning of the consent document and concisely summarize:

Additional required elements:

Signature requirements: Subject (or LAR) signs and dates; investigator obtaining consent signs; copy given to subject

Process, not paperwork: Consent is an ongoing dialogue; re-consent is required if new risks emerge or protocol changes materially

Diagnostic Workup — IRB Review Levels and Categories

— Normal educational practices in established settings

— Anonymous surveys/interviews with adults (limits if sensitive topics or identifiable)

— Secondary research on de-identified existing data

— Benign behavioral interventions with adult subjects

— Taste/food quality evaluation

— Conducted by IRB chair or designated experienced reviewer (not full board)

— Examples: blood draws within volume limits, non-invasive data collection, prospective collection of biospecimens by noninvasive means, research on existing records, minor changes to previously approved research

— "Minimal risk" = probability and magnitude of harm not greater than those encountered in daily life or routine physical/psychological exams

— Research exceeds minimal risk

— Involves vulnerable populations (prisoners always require full board)

— Doesn't fit expedited categories

— Quorum required (majority of members, including at least one nonscientist)

— ≥5 members with varying backgrounds

— At least one scientist, one nonscientist, one member not affiliated with the institution

— Diversity in race, gender, cultural backgrounds

— Members with conflicts must recuse from voting

Board pearl: A study modification involving a new drug, new population, or new risk generally bumps an expedited study back to full board review — investigators cannot continue under the old approval. The exam loves "investigator added a new arm without IRB notification" as a regulatory violation.

Step 3 management: When asked who determines whether a project is exempt, the answer is the IRB, never the investigator or the department chair, even when criteria seem obviously met.

Three IRB review pathways, determined by risk level and research category:

Exempt review (still requires IRB determination — investigator cannot self-declare):

Expedited review — minimal risk research in defined categories:

Full board review — required when:

IRB composition requirements:

Diagnostic Workup — Special Documentation and Waivers

— Minimal risk

— Waiver will not adversely affect rights and welfare of subjects

— Research could not practicably be carried out without waiver

— Subjects provided with pertinent information after participation (when appropriate)

— For identifiable data/biospecimens: research could not practicably be carried out without using such information

— Only record linking subject to research would be the consent form AND principal risk is breach of confidentiality (subject choice); OR

— Research is minimal risk and involves no procedures requiring written consent outside research context (e.g., telephone survey)

— For life-threatening conditions where consent is not feasible

— Examples: cardiac arrest trials, severe TBI resuscitation

— Requires community consultation and public disclosure before study begins

— Independent data monitoring committee required

— Family member must be approached for consent if feasible within therapeutic window

— Children ≥7 generally provide assent (affirmative agreement)

— Both parents' permission required if research >minimal risk with no direct benefit

— Mature minor and emancipated minor doctrines vary by state

CCS pearl: Emergency research without prospective consent is one of the narrowest exceptions in US regulation. If a stem describes researchers enrolling unconscious trauma patients in a randomized resuscitation trial without prior community consultation, the correct answer is that the study violates 21 CFR 50.24 and must be halted.

Waiver of informed consent — IRB may waive ALL of the following must be met:

Waiver of documentation of consent (oral consent permitted, no signature):

Emergency research exception (21 CFR 50.24 — "exception from informed consent" / EFIC):

Short form consent: when subject doesn't speak English well — short form in their language plus full English form with witness/translator

Assent (children) + parental permission:

Risk Stratification — Vulnerable Populations and Subpart Protections

— Subpart B — pregnant women, fetuses, neonates

— Subpart C — prisoners

— Subpart D — children

— (Subpart E governs IRB registration)

— IRB must include a prisoner representative

— Research limited to: study of incarceration/criminal behavior; prisons as institutions; conditions affecting prisoners as a class; research on practices likely to improve prisoner health/well-being

— Cannot enroll prisoners in general drug trials without specific justification

— If a subject becomes incarcerated mid-study, investigator must notify IRB promptly

— §46.404: minimal risk

— §46.405: greater than minimal risk but prospect of direct benefit

— §46.406: minor increase over minimal risk, no direct benefit, but generalizable knowledge about subject's condition

— §46.407: not otherwise approvable — requires Secretary of HHS panel review

— Risk to fetus must be minimal OR research holds prospect of direct benefit to woman/fetus

— Father's consent required only for research with no maternal benefit and >minimal fetal risk (with exceptions for unavailability, rape, incest)

— Capacity must be assessed at time of consent

— Legally authorized representative (LAR) provides surrogate consent

— Subject's assent still sought when possible

— Advance research directives encouraged

Board pearl: A college professor recruiting their own students is undue influence even without explicit coercion. Same for physicians recruiting their own patients or employers recruiting employees — IRBs require firewall procedures (independent recruiter, opt-in rather than opt-out, no impact on grades/care/employment disclosed in writing).

45 CFR 46 has additional subparts offering extra protections:

Prisoners (Subpart C):

Children (Subpart D) — categories of permissible research:

Pregnant women/fetuses (Subpart B):

Cognitively impaired adults:

Pharmacotherapy — Conflict of Interest and Financial Disclosure

— "Significant Financial Interest" = >$5,000 in publicly traded entity, any equity in non-public entity, any IP rights with expected income, or sponsored travel

— Must be disclosed to institution before application submission and within 30 days of acquiring new interest

— Institution determines if it constitutes a Financial Conflict of Interest (FCOI) and reports to PHS funding agency

— Public disclosure on ClinicalTrials.gov and in publications/consent forms

— Independent monitoring of research

— Modification of research plan

— Disqualification from participation

— Divestiture of significant financial interest

— Severance of relationships creating conflict

— Sponsor cannot restrict publication of negative results (most journals require this in author agreements; ICMJE policy)

— Per-subject payments to investigators must be reasonable and not contingent on enrollment volume in a way that creates undue influence

— "Finder's fees" for referring patients to studies are generally prohibited

— Compensation for time/inconvenience is acceptable

— Cannot be so large as to constitute undue influence (especially for low-income or vulnerable subjects)

— Should not be contingent on completing the study in a way that penalizes withdrawal

Step 3 management: A physician-investigator who owns equity in the drug sponsor and is enrolling their own clinic patients should disclose the financial interest to the IRB, disclose it in the consent form, and arrange for an independent investigator to obtain consent. "Stop the research" is too aggressive; "do nothing" is wrong; disclosure plus management is the answer.

Financial COI is heavily tested because it threatens objectivity, subject welfare, and public trust

Investigator-level COI thresholds (PHS regulations, 42 CFR 50 Subpart F):

Management strategies (institution may impose):

Industry sponsorship issues:

Subject payments:

Sunshine Act (Open Payments): physicians' industry payments publicly reported

Procedures — IRB Operations, Continuing Review, and Reporting

— Initial approval typically valid up to 1 year (Revised Common Rule eliminated mandatory annual continuing review for some minimal-risk studies, but most IRBs retain it)

— Continuing review before expiration required for greater-than-minimal-risk studies

— Lapse in approval = must stop all research activities including data analysis, except as needed to protect already-enrolled subjects

— Any change to protocol, consent form, recruitment materials, or investigator team requires IRB approval before implementation

— Exception: changes necessary to eliminate apparent immediate hazard to subjects may be made first, then reported within required timeframe (usually 5 working days)

— Unanticipated problems involving risks to subjects or others (UPIRTSOs) — typically within 10 working days

— Serious adverse events — within 10 working days

— Protocol deviations affecting subject safety/data integrity

— Subject complaints

— New information affecting risk-benefit

— Breach of confidentiality

— Required for most NIH-funded clinical trials, especially Phase III and high-risk Phase I/II

— Independent of investigators and sponsors

— Can recommend early stopping for efficacy, futility, or safety

— Form FDA 483 issued for observations

— Warning Letters for significant noncompliance

— Clinical hold can halt IND research

CCS pearl: When a vignette mentions a serious unanticipated adverse event in a research subject, the immediate next step is to (1) provide necessary clinical care, (2) report to IRB and sponsor within required timeframe, and (3) assess whether consent form requires modification to disclose the new risk to current and future subjects. Continuing enrollment without IRB notification is a major violation.

IRB approval is time-limited:

Amendments and modifications:

Reportable events (timeframes vary by IRB but commonly):

Data and Safety Monitoring Boards (DSMB):

FDA inspections and audits:

Special Populations — Capacity Assessment and Surrogate Consent

— Understanding — grasp relevant information

— Appreciation — apply information to one's own situation

— Reasoning — manipulate information rationally to compare options

— Expressing a choice — communicate consistent preference

— Dementia, severe psychiatric illness, intellectual disability, intoxication, acute medical illness

— Assessment tools: MacCAT-CR (MacArthur Competence Assessment Tool for Clinical Research)

— Independent capacity assessor recommended for high-risk studies

— Court-appointed guardian with research authority

— Durable power of attorney for healthcare (if research authority specified)

— Spouse → adult child → parent → adult sibling → other relative → close friend

— Surrogates should consider subject's prior expressed wishes and best interests

— Many states restrict surrogate consent for greater-than-minimal-risk research without direct benefit

— Subject's dissent or objection (verbal or behavioral) generally overrides surrogate consent

— Polypharmacy, sensory impairment, and slower processing speed may require modified consent process (large print, extra time, teach-back)

— Age alone does not impair capacity

Key distinction: Capacity is a clinical determination made by any physician at the bedside; competence is a legal determination made by a court. The exam tests this distinction frequently. A patient adjudicated incompetent by a court cannot provide research consent regardless of clinical presentation; their court-appointed guardian must consent within the scope of guardianship authority.

Capacity is decision-specific — a subject may have capacity to consent to a blood draw but not to a Phase I oncology trial

Four elements of decisional capacity (Appelbaum criteria):

Cognitive impairment populations:

Hierarchy of surrogate decision-makers (varies by state):

Limitations on surrogate research consent:

Older adults:

Special Populations — International, Community-Based, and Genomic Research

— Must comply with both US regulations AND host country requirements

— Declaration of Helsinki and CIOMS guidelines are key international frameworks

— Post-trial access to interventions proven beneficial — ethical obligation per Helsinki

— Standard of care comparator: must offer best proven intervention, with limited exceptions (CIOMS Guideline 4)

— Equivalent IRB oversight in host country required

— Community advisory boards engaged from study design through dissemination

— Tribal IRBs have sovereign authority over research involving Native American communities — federal IRB approval does not substitute

— Example: Havasupai Tribe case — DNA collected for diabetes research was used for schizophrenia, migration, and inbreeding studies without consent → $700,000 settlement and return of samples

— GINA (Genetic Information Nondiscrimination Act, 2008) prohibits health insurance and employment discrimination based on genetic information (does NOT cover life, disability, long-term care insurance)

— Certificates of Confidentiality protect against compelled disclosure of identifiable research data

— Return of individual genetic results: required if results are analytically valid, clinically actionable, and subject wants them (per consent)

— Broad consent under Revised Common Rule allows future unspecified research on biospecimens with appropriate disclosures

Board pearl: Research on de-identified stored biospecimens collected for clinical purposes generally does not require new consent under the Revised Common Rule (may qualify as exempt or not human subjects research), BUT institutional policy and tribal/community consent may still apply. The Havasupai case illustrates that legal sufficiency ≠ ethical sufficiency.

International research:

Community-based participatory research (CBPR):

Genomic and biospecimen research:

Pediatric international research: "standard of care" debates particularly fraught (e.g., perinatal HIV trials)

Complications and Adverse Outcomes — Research Misconduct

— Fabrication — making up data or results

— Falsification — manipulating materials, equipment, processes, or omitting data such that research isn't accurately represented

— Plagiarism — appropriation of another's ideas, processes, results, words without giving appropriate credit

— Honest error and differences of opinion are NOT misconduct

— Report to Research Integrity Officer (RIO) at institution

— Inquiry phase (preliminary, ~60 days)

— Investigation phase if warranted (~120 days)

— Findings reported to Office of Research Integrity (ORI) for PHS-funded research

— Sanctions: debarment from federal funding, retraction of publications, termination of employment, license actions

— Good-faith reporters protected from retaliation

— Retaliation itself is reportable misconduct

— ICMJE criteria — all four must be met: substantial contribution to design/data; drafting or critical revision; final approval; accountability

— Ghost authorship (real contributor omitted) and gift/honorary authorship (listed without contribution) are unethical

— Guest authorship by senior figures who didn't contribute = misconduct

Step 3 management: A trainee who suspects a PI of fabricating data should first contact the institutional Research Integrity Officer, not the press, not social media, not the PI directly. Going to ORI directly is also acceptable but RIO is the standard first step. Discussing concerns with the PI privately is not recommended for true suspected misconduct.

Research misconduct (federal definition, 42 CFR 93): Fabrication, Falsification, or Plagiarism (FFP) in proposing, performing, reviewing research, or reporting results

Process when misconduct suspected:

Whistleblower protections:

Authorship issues:

Duplicate publication and salami slicing — submitting the same data or thinly sliced data to multiple journals — violates most journal policies

Data sharing and reproducibility: NIH Data Management and Sharing Policy (effective 2023) requires plans for sharing scientific data

When to Escalate — Stopping Rules and IRB Suspension Authority

— Must report suspension/termination to OHRP, FDA (if applicable), institutional officials, and the funding agency

— Continuation poses unreasonable risk

— Subject withdraws (must respect immediately for already-enrolled but not yet randomized; for randomized subjects, continue collecting safety data if subject consents)

— Sponsor terminates the study

— Efficacy — overwhelming benefit demonstrated (e.g., HIV antiretroviral trials)

— Futility — no realistic chance of demonstrating benefit

— Harm — unacceptable safety signal (e.g., VIGOR study rofecoxib cardiovascular signal)

— Pre-specified statistical stopping rules (O'Brien-Fleming, Pocock boundaries) reduce risk of false-positive early stops

— IND safety reports to FDA within 15 days for serious and unexpected suspected adverse reactions

— 7-day report for unexpected fatal or life-threatening events

— IRB must be notified; study cannot continue without an approved PI

— Transition plan for enrolled subjects required

CCS pearl: If during a CCS-style vignette you see an unexpected death in a research subject potentially related to study intervention, the immediate actions are: (1) stabilize/treat clinically as appropriate, (2) notify IRB and sponsor within mandated timeframe (typically same/next business day for deaths), (3) suspend enrollment pending review, and (4) assess need to inform currently enrolled subjects so they can reconsider participation.

IRB has authority to suspend or terminate approval if research is not being conducted in accordance with IRB requirements or has been associated with unexpected serious harm to subjects

Investigator-initiated stopping: required when

DSMB-recommended early stopping:

Clinical hold (FDA): can halt or restrict an IND-regulated trial for safety concerns, deficient protocol, or unqualified investigator

For Aggregate Safety reporting:

When an investigator leaves the institution:

Key Differentials — Research vs Quality Improvement vs Public Health

— Research: generalizable knowledge, hypothesis testing, pre-specified intervention, often randomization, publication intent at outset

— QI: local improvement of care processes, iterative PDSA cycles, uses established evidence, no formal hypothesis testing, internal dissemination

— Publication intent alone does NOT make a QI project research — but a project designed from the start to produce generalizable knowledge IS research even if framed as QI

— When unsure: consult IRB for a "not human subjects research" determination — this is itself the correct answer on the exam

— Surveillance authorized by public health authority for disease control is generally not research

— Outbreak investigations may transition to research if generalizable knowledge becomes the goal — requires IRB

— "Off-label" prescribing for individual patient benefit is clinical care, not research

— Systematic study of an innovative technique becomes research

— Surgical innovation guidelines (IDEAL framework) suggest IRB review for novel procedures

— Resident case logs, M&M conferences = education

— Studying resident performance for publication = research

Board pearl: When in doubt, the safest exam answer is "submit to the IRB for determination." The IRB — not the investigator, department chair, or hospital attorney — has the authority to classify an activity. This answer is correct on the vast majority of "is this research?" stems.

Research vs QI — most commonly tested distinction:

Research vs Public Health Practice/Surveillance:

Research vs Innovative Clinical Care:

Research vs Educational Activities:

Program evaluation — similar to QI; generally not research if scope is local

Key Differentials — Other Ethical Frameworks Overlapping Research

— Clinical: best interest of individual patient, fiduciary duty

— Research: scientific question, generalizable knowledge, individual patient may not benefit

— Equipoise — genuine uncertainty about which arm is better — is the ethical justification for randomization

— Clinical equipoise (community of experts disagree) is the operative standard

— Misconception: subject misunderstands research is for generalizable knowledge, not personal care

— Optimism: subject understands but hopes for personal benefit (acceptable)

— HIPAA Privacy Rule governs use of Protected Health Information by covered entities

— Research uses of PHI generally require authorization (separate from consent) OR waiver by IRB/Privacy Board

— Limited data sets and de-identified data have different requirements

— A study can be exempt under Common Rule but still require HIPAA authorization

— FDA-regulated research (drugs, devices, biologics) has additional requirements

— FDA does not recognize all Common Rule exempt categories

— Both apply concurrently for many studies

Key distinction: Consent (Common Rule) ≠ Authorization (HIPAA). Consent governs the research relationship; authorization governs the use of PHI. A single form may combine both, but the elements are distinct and an IRB can waive one without waiving the other.

Clinical care ethics vs research ethics: distinct frameworks that sometimes conflict

Therapeutic misconception vs therapeutic optimism:

HIPAA vs Common Rule:

FDA regulations vs Common Rule:

Animal research (IACUC) vs human research (IRB): separate committees, separate frameworks (3Rs: Replace, Reduce, Refine)

Embryonic stem cell research, gene editing, gain-of-function research: additional oversight layers (ESCRO, IBC, DURC policies)

Secondary Prevention — Post-Study Obligations and Dissemination

— Return of clinically relevant individual results when analytically valid and actionable

— Post-trial access to interventions proven beneficial (Declaration of Helsinki) — particularly important in international/resource-limited settings

— Notification of study results in lay summary form (EU Clinical Trials Regulation mandates this; US increasingly expects it)

— Continued safety monitoring if delayed adverse effects possible

— ClinicalTrials.gov registration required for "applicable clinical trials" (most Phase II–IV interventional studies of FDA-regulated products) before first subject enrolled

— Results reporting within 1 year of primary completion date (FDAAA 801)

— Failure to register/report → civil monetary penalties up to ~$10,000+/day; loss of NIH funding

— CONSORT (RCTs), STROBE (observational), PRISMA (systematic reviews) reporting guidelines

— Ethical obligation — failure to publish negative data biases the literature (publication bias)

— Investigators must publish regardless of sponsor preference

— NIH DMS Policy (2023): data management and sharing plan required for all NIH-funded research generating scientific data

— De-identified individual participant data sharing increasingly expected

Step 3 management: A pharmaceutical sponsor pressures the PI not to publish negative trial results. The correct action is to publish the results in accordance with the pre-specified analysis plan and ICMJE/journal disclosure requirements. Most contracts post-2005 explicitly preserve publication rights; suppression of negative findings constitutes research misconduct or breach of professional ethics.

Post-study obligations to subjects:

Publication and dissemination requirements:

Publication of negative results:

Data sharing:

Biospecimen and data retention: governed by IRB-approved protocol and consent; destruction or de-identification at study close if specified

Follow-Up, Monitoring Parameters, and Investigator Training

— CITI Program (Collaborative Institutional Training Initiative) — standard for human subjects protections training

— Required before IRB approval; renewal typically every 3 years

— Good Clinical Practice (GCP) training — NIH requires for clinical trial investigators

— Specific training for vulnerable populations, biospecimen research, international research as applicable

— Site monitoring visits (sponsor or CRO) — verify source documents, consent forms, drug accountability

— Internal audits by institution

— FDA Bioresearch Monitoring (BIMO) inspections — investigators, IRBs, sponsors, monitors

— Adverse event diaries, lab monitoring, ECGs, imaging per protocol

— Pregnancy testing in studies of reproductive-age subjects exposed to investigational agents

— Withdrawal/safety follow-up after intervention completion

— Trainees should be listed as study personnel on IRB applications

— Cannot obtain consent without delegation by PI and appropriate training

— Research mentors are responsible for trainee conduct

— Signed consent forms retained ≥3 years after study completion (FDA: ≥2 years after marketing application approval or discontinuance)

— Source documents, case report forms, drug accountability logs

CCS pearl: A resident wants to do a chart review project for a poster. Step 3 management: (1) Determine if it's research (likely yes if generalizable knowledge intended), (2) Submit for IRB determination/approval before accessing charts, (3) Complete required human subjects training (CITI), (4) Obtain HIPAA waiver of authorization if not consenting individual patients, (5) Be listed on the IRB protocol under a faculty PI.

Required investigator training:

Ongoing monitoring during study conduct:

Subject monitoring parameters depending on study:

IRB member education: ongoing training in regulatory updates, ethical frameworks

Mentorship and trainee research:

Documentation requirements:

Ethical, Legal, and Patient Safety Considerations

— Informed consent edge case 1: A non-English-speaking patient is approached for a trial. Correct: use IRB-approved translated consent OR short form in subject's language plus full English form, witnessed by a qualified interpreter who signs. Ad hoc translation by family member is not acceptable for research consent.

— Edge case 2: Patient in the ED is unconscious from severe sepsis; sepsis trial available with EFIC approval. Correct: enroll under emergency exception, attempt to contact family within therapeutic window, re-consent the patient as soon as capacity returns; subject may withdraw without penalty.

— Edge case 3: Adolescent age 16 wants to enroll in a contraception study; parent refuses. State-dependent — mature minor doctrines and emancipated minor status may permit independent consent for reproductive health research; IRB determines acceptable approach in advance.

— Child abuse, elder abuse, intimate partner violence disclosed during research interview = mandatory reporting still applies; this must be disclosed in consent form as a limit of confidentiality

— Communicable disease reporting (e.g., new HIV diagnosis) — same

— Imminent harm to self or others — Tarasoff-style duty applies

Board pearl: A Certificate of Confidentiality protects research data from subpoena but does not protect against the investigator's own mandatory reporting obligations for child abuse, elder abuse, or threats of imminent harm. Both must be disclosed during consent.

Step 3-specific scenarios:

Mandatory reporting intersections:

Certificates of Confidentiality: protect against compelled disclosure in legal proceedings but do not override mandatory reporting laws

Transition-of-care risk: when a research subject is hospitalized at another institution, the home study team must be notified to coordinate safety reporting and ensure investigational product is handled correctly (especially for unblinding emergencies)

Unblinding for clinical emergency: must be possible 24/7; documented in protocol; investigator should unblind only when knowledge of assignment will change clinical management

Compensation for research injury: investigators must clarify in consent who pays — sponsor, institution, or subject's insurance; "no compensation available" must be explicitly stated when true

High-Yield Associations and Rapid-Fire Clinical Facts

Key distinction: OHRP enforces Common Rule (HHS-funded); FDA enforces 21 CFR (regulated products); ORI handles research misconduct in PHS-funded research. Different agencies, sometimes overlapping jurisdiction.

Belmont principles → Respect for persons, Beneficence, Justice (1979)

National Research Act (1974) → created IRBs

Common Rule = 45 CFR 46 (Revised 2018)

FDA research regs = 21 CFR 50 (consent), 56 (IRBs), 312 (IND), 812 (IDE)

Tuskegee → 40 years, untreated syphilis → Belmont Report

Willowbrook → hepatitis in institutionalized children → vulnerable populations

Henrietta Lacks / HeLa → biospecimen consent issues

Havasupai → tribal sovereignty, secondary use of DNA

Moore v. Regents → no property right in cells; duty to disclose financial COI

Nuremberg Code (1947) → voluntary consent essential

Declaration of Helsinki → WMA, periodically revised

GINA (2008) → genetic discrimination in health insurance & employment (NOT life/disability/LTC)

HIPAA → PHI authorization required for research, waivable by IRB

FDAAA 801 → ClinicalTrials.gov registration & results reporting

Minimal risk = harm of daily life or routine exam

IRB minimum membership = 5, including nonscientist and unaffiliated member

Prisoner research → IRB must include prisoner representative

Children's assent typically ≥7 years

Pregnant women → Subpart B; Prisoners → Subpart C; Children → Subpart D

Therapeutic misconception → subject believes research = personal treatment

Clinical equipoise → ethical basis for randomization

Research misconduct = FFP (Fabrication, Falsification, Plagiarism); honest error excluded

ORI = Office of Research Integrity (PHS)

OHRP = Office for Human Research Protections (Common Rule)

ICMJE authorship = 4 criteria, all must be met

Certificates of Confidentiality → protect data from subpoena; do NOT override mandatory reporting

EFIC = Emergency exception from informed consent; 21 CFR 50.24

CITI training = standard human subjects training

DSMB → independent body for safety/efficacy monitoring in trials

Board Question Stem Patterns

— Stem describes resident analyzing 50 charts to present at national meeting → submit to IRB for determination before starting

— Medical student approaches ED patient at 3 AM with consent form → likely violations: undue circumstances, possible inadequate delegation, possible inability to assess capacity → stop, refer to attending PI, obtain consent in appropriate setting

— Investigator wants to enroll prisoners in general antihypertensive trial → not permitted under Subpart C unless one of the four permitted categories applies

— Physician owns stock in device company and enrolls own patients in device trial → disclose to IRB, disclose in consent, arrange independent investigator for enrollment

— Subject dies unexpectedly in trial → provide clinical care, report to IRB and sponsor in mandated timeframe, suspend enrollment pending review, consider re-consent of existing subjects

— Trainee notices PI altering data → report to institutional Research Integrity Officer

— Sponsor asks PI to delay publication of negative results indefinitely → publish per pre-specified analysis plan

— Cancer patient says "I'm in this trial because Dr. X said it will cure me" → clarify research purpose, reassess understanding, re-consent if needed

— Patient with mild dementia wants to enroll → formal capacity assessment for the specific decision; if lacking, surrogate consent + subject assent

— US researcher conducting trial in low-income country with placebo control when proven therapy exists elsewhere → ethically problematic; use best proven intervention as comparator

Step 3 management: When the answer choices include "consult IRB," "report to IRB," or "submit for IRB determination," that is almost always the correct answer over "proceed" or "discuss informally with colleague." The IRB is the institutional safety net for research ethics.

Pattern 1 — "Is this research?"

Pattern 2 — Consent process violation:

Pattern 3 — Vulnerable population:

Pattern 4 — Conflict of interest:

Pattern 5 — Adverse event:

Pattern 6 — Misconduct suspicion:

Pattern 7 — Sponsor pressure:

Pattern 8 — Therapeutic misconception:

Pattern 9 — Capacity:

Pattern 10 — International:

One-Line Recap

Research ethics on Step 3 reduces to a single rule: any systematic activity designed to produce generalizable knowledge involving human subjects requires prospective IRB review, valid informed consent (or an approved waiver), and ongoing oversight — with extra protections for vulnerable populations and mandatory reporting of adverse events, conflicts, and misconduct.

Board pearl: The most commonly missed answer is recognizing that IRB determination is required before an activity begins — not at the manuscript stage. The second most missed is that clinical equipoise, not personal physician uncertainty, justifies randomization. Master these two and you've cleared most research-ethics questions on test day.

The three Belmont principles — respect for persons (autonomy/consent), beneficence (favorable risk-benefit), justice (fair selection) — are the ethical foundation, operationalized through the Common Rule (45 CFR 46) and FDA regulations (21 CFR 50/56)

Valid consent requires capacity, voluntariness, and disclosure, with "key information" presented up front under the Revised Common Rule; waivers require minimal risk and impracticability, and emergency research (EFIC) requires community consultation

Vulnerable populations (Subparts B–D: pregnant women, prisoners, children) require additional protections; surrogate consent for cognitively impaired adults follows state hierarchies, and subject dissent generally overrides surrogate consent

When in doubt on the exam: submit to the IRB for determination, report adverse events promptly, disclose conflicts transparently, publish negative results, and never permit retroactive IRB approval — these reflexes will resolve the overwhelming majority of research ethics vignettes