Reference
40 terms used in clinical research, each with what it means in practice and what it implies under FDA regulatory standards.
How far a substance has travelled through human testing.
Laboratory and animal work only — no humans have received the substance in a controlled study.
FDA implication
Cannot support any human efficacy or safety claim. An Investigational New Drug (IND) application must be cleared before human dosing may begin.
Source
FDA — Investigational New Drug (IND) Application →Published / last updated 16 September 2026 · fda.gov
First administration to humans, usually 20–100 participants, focused on safety, tolerability and how the body handles the substance.
FDA implication
Establishes a safe dose range only. Efficacy conclusions are not supported at this stage.
Source
FDA — The Drug Development Process, Step 3: Clinical Research →Published / last updated 4 January 2018 · fda.gov
Several hundred participants with the target condition; looks for a first efficacy signal and refines dosing.
FDA implication
Treated as preliminary evidence. Promising Phase 2 results frequently fail to replicate in Phase 3.
Source
FDA — The Drug Development Process, Step 3: Clinical Research →Published / last updated 4 January 2018 · fda.gov
Large confirmatory trials, often thousands of participants across multiple sites, comparing against placebo or standard care.
FDA implication
This is the evidence tier the FDA normally requires for approval — usually two adequate and well-controlled trials.
Source
FDA — The Drug Development Process, Step 3: Clinical Research →Published / last updated 4 January 2018 · fda.gov
Post-marketing studies and surveillance after approval, tracking rare harms and long-term outcomes in real-world use.
FDA implication
Can be required as a condition of approval; findings may trigger label changes, boxed warnings or withdrawal.
Source
FDA — Postmarket Requirements and Commitments →Published / last updated 9 August 2024 · fda.gov
Design decides how much a result can be trusted.
Participants are assigned to treatment or control by chance, so the groups differ only by the intervention.
FDA implication
The reference standard for an 'adequate and well-controlled investigation' in FDA regulations.
Source
FDA Guidance — Demonstrating Substantial Evidence of Effectiveness →Published / last updated 25 June 2026 · fda.gov
Neither participants nor investigators know who receives active treatment.
FDA implication
Expected for subjective outcomes such as pain, mood, energy or sleep quality, where expectation strongly biases reporting.
Source
FDA Guidance E10 — Choice of Control Group and Related Issues in Clinical Trials →Published / last updated 24 August 2018 · fda.gov
The comparison group receives an inactive product identical in appearance.
FDA implication
Required where withholding active treatment is ethical; otherwise an active comparator is used.
Source
FDA Guidance E10 — Choice of Control Group and Related Issues in Clinical Trials →Published / last updated 24 August 2018 · fda.gov
Everyone knows who receives what.
FDA implication
Accepted for safety follow-up and some oncology settings, but weak for subjective efficacy endpoints.
Source
FDA Guidance E10 — Choice of Control Group and Related Issues in Clinical Trials →Published / last updated 24 August 2018 · fda.gov
Each participant receives both treatment and control in sequence, acting as their own control.
FDA implication
Efficient, but only valid when the condition is stable and the effect washes out between periods.
Source
FDA Guidance E9 — Statistical Principles for Clinical Trials →Published / last updated 24 August 2018 · fda.gov
Investigators observe outcomes without assigning treatment, following groups over time.
FDA implication
Shows association, not causation. Cannot on its own support a therapeutic claim.
Source
FDA Guidance — Demonstrating Substantial Evidence of Effectiveness →Published / last updated 25 June 2026 · fda.gov
Statistical pooling of several trials to produce one combined estimate.
FDA implication
Only as reliable as the trials pooled; heterogeneity and publication bias can distort the result.
Source
FDA Guidance — Demonstrating Substantial Evidence of Effectiveness →Published / last updated 25 June 2026 · fda.gov
What was actually measured — and whether it matters clinically.
The single main outcome the trial was designed and powered to detect.
FDA implication
Only the pre-specified primary endpoint normally supports an approval claim.
Source
FDA Guidance — Multiple Endpoints in Clinical Trials →Published / last updated 16 April 2024 · fda.gov
Additional outcomes measured alongside the primary one.
FDA implication
Supportive at best; significant secondary results after a failed primary endpoint are treated as exploratory.
Source
FDA Guidance — Multiple Endpoints in Clinical Trials →Published / last updated 16 April 2024 · fda.gov
A lab or imaging marker (cholesterol, bone density, IGF-1) used as a stand-in for a clinical outcome.
FDA implication
Accepted for accelerated approval only when reasonably likely to predict clinical benefit; confirmatory trials are then required.
Source
FDA — Surrogate Endpoint Resources for Drug and Biologic Development →Published / last updated 24 July 2018 · fda.gov
Several events counted together, for example heart attack, stroke and cardiovascular death.
FDA implication
A positive composite can be driven by its least serious component — always read the breakdown.
Source
FDA Guidance — Multiple Endpoints in Clinical Trials →Published / last updated 16 April 2024 · fda.gov
Outcome reported directly by the participant, such as fatigue or pain.
FDA implication
Accepted when collected with a validated instrument; blinding matters a great deal here.
Source
FDA Guidance — Patient-Reported Outcome Measures in Medical Product Development →Published / last updated 17 October 2019 · fda.gov
The numbers that decide whether a finding is real and meaningful.
The probability of seeing a result this extreme if the treatment truly had no effect; p < 0.05 is the common threshold.
FDA implication
Statistical significance is not clinical importance, and a significant p-value on an unplanned analysis carries little weight.
Source
FDA Guidance E9 — Statistical Principles for Clinical Trials →Published / last updated 24 August 2018 · fda.gov
The range of effect sizes compatible with the data, usually at 95%.
FDA implication
More informative than a p-value; a wide interval signals an imprecise, low-confidence estimate.
Source
FDA Guidance E9 — Statistical Principles for Clinical Trials →Published / last updated 24 August 2018 · fda.gov
Relative reduction is a proportion (a 25% drop); absolute reduction is the real-world difference (from 4% to 3%).
FDA implication
Promotional use of relative figures alone is considered misleading; absolute numbers must be available.
Source
FDA Guidance E9 — Statistical Principles for Clinical Trials →Published / last updated 24 August 2018 · fda.gov
How many people must be treated for one to benefit.
FDA implication
Not a regulatory requirement, but the most practical way to explain the size of a benefit.
Source
FDA Guidance E9 — Statistical Principles for Clinical Trials →Published / last updated 24 August 2018 · fda.gov
The trial's ability to detect a real effect; commonly set at 80–90% before enrolment.
FDA implication
An underpowered trial that finds nothing has not shown the treatment fails — it simply could not tell.
Source
FDA Guidance — Non-Inferiority Clinical Trials to Establish Effectiveness →Published / last updated 24 August 2018 · fda.gov
All randomised participants are analysed in their assigned group, whether or not they completed treatment.
FDA implication
The preferred primary analysis; per-protocol analyses alone tend to overstate benefit.
Source
FDA Guidance E6(R3) — Good Clinical Practice →Published / last updated 8 September 2025 · fda.gov
How harms are recorded and acted upon.
Any unfavourable occurrence during the trial, whether or not caused by the treatment.
FDA implication
Must be reported regardless of suspected cause; causality is judged separately.
Source
FDA — MedWatch: Safety Information and Adverse Event Reporting Program →Published / last updated 10 October 2026 · fda.gov
Death, a life-threatening event, hospitalisation, lasting disability or a birth defect.
FDA implication
Subject to expedited reporting to the FDA; can pause or stop a trial.
Source
FDA — MedWatch: Safety Information and Adverse Event Reporting Program →Published / last updated 10 October 2026 · fda.gov
A situation where the substance should not be used at all.
FDA implication
The strongest labelling restriction, reserved for risks that clearly outweigh any benefit.
The most serious warning the FDA can require on a label.
FDA implication
Signals a risk of death or serious injury; often accompanied by a risk-management programme.
What a legal label does and does not promise about evidence.
Reviewed and authorised for a specific indication, dose form and population after confirmatory trials.
FDA implication
Applies only to the approved indication. Approval for weight loss says nothing about other uses.
Marketed under the 1994 Dietary Supplement Health and Education Act as a food, not a drug.
FDA implication
No pre-market approval of efficacy or safety. Manufacturers hold responsibility; the FDA acts mainly after harm is reported.
Wording such as 'supports joint health', permitted on supplement labels with a disclaimer.
FDA implication
Not FDA-evaluated. Claiming a substance treats, prevents or cures a disease makes it an unapproved drug.
Generally Recognised As Safe for a stated use in food.
FDA implication
A safety designation for a specific use and dose only — never a statement of efficacy.
Source
FDA — Current Good Manufacturing Practices for Food and Dietary Supplements →Published / last updated 24 April 2024 · fda.gov
Sold outside any approved medical or supplement channel, with no manufacturing oversight.
FDA implication
Illegal to market for human consumption. Purity, dose and sterility are unverified; many such substances are also banned in sport.
Source
FDA — New Dietary Ingredient (NDI) Notification Process →Published / last updated 11 June 2025 · fda.gov
Prescribing an approved drug for an indication, dose or population outside the label.
FDA implication
Legal for clinicians, but manufacturers may not promote it and the evidence base may be thin.
Source
FDA — The Drug Development Process, Step 4: FDA Drug Review →Published / last updated 4 January 2018 · fda.gov
Mixed by a pharmacy for an individual rather than manufactured commercially.
FDA implication
Not FDA-approved, reviewed or verified for potency; a recurring safety issue for peptide products.
Source
FDA — Human Drug Compounding: Laws and Policies →Published / last updated 17 December 2024 · fda.gov
What our own labels mean when you read a monograph.
Strong, consistent human evidence from multiple well-conducted trials or robust meta-analyses.
FDA implication
Typically corresponds to Phase 3 evidence or an equivalent trial base.
Source
FDA Guidance — Demonstrating Substantial Evidence of Effectiveness →Published / last updated 25 June 2026 · fda.gov
Moderate human evidence; results are promising but incomplete or mixed.
FDA implication
Usually Phase 2 or limited Phase 3 data.
Source
FDA Guidance — Demonstrating Substantial Evidence of Effectiveness →Published / last updated 25 June 2026 · fda.gov
Limited or conflicting human evidence, often small or short trials.
FDA implication
Not sufficient for a regulatory claim.
Source
FDA Guidance — Demonstrating Substantial Evidence of Effectiveness →Published / last updated 25 June 2026 · fda.gov
Little or no credible human evidence for the specific claim, or evidence pointing against it.
FDA implication
Any marketing claim at this level is unsupported.
Source
FDA Guidance — Demonstrating Substantial Evidence of Effectiveness →Published / last updated 25 June 2026 · fda.gov
Human studies exist but measured lab or hormone levels, not real outcomes such as strength, body composition or health events.
FDA implication
A surrogate marker change is not a demonstrated clinical benefit.
Source
FDA — Surrogate Endpoint Resources for Drug and Biologic Development →Published / last updated 24 July 2018 · fda.gov
No published human trial of any kind for this substance and claim; the claim is extrapolated from animal or cell work.
FDA implication
No regulatory or clinical standing whatsoever.
Source
FDA — Investigational New Drug (IND) Application →Published / last updated 16 September 2026 · fda.gov
