thebench.bio #1 Ecosystem for the Bio-Tech Industry
All Clinical Trial Design questions

Question 12 of 20 Fundamentals Practical

How should inclusion and exclusion criteria be chosen?

Short answer

Each criterion should earn its place by protecting participants or serving the study objective. Define the population the drug is meant to treat, then exclude only people put at unreasonable risk or who would make the answer uninterpretable, using criteria specific enough not to exclude more than necessary.

Why

Eligibility criteria decide who the results apply to. Criteria that are too narrow make enrollment slow and give results that may not hold in the patients who will actually use the drug. Criteria that are too broad can expose vulnerable people to risk or dilute the effect.

FDA's guidance on enhancing participation in clinical trials, revised in December 2025, asks sponsors to examine each exclusion and drop or modify it if it is not needed for safety or the objective. It gives a clear example: if only advanced heart failure carries unreasonable risk, exclude NYHA class III and IV rather than all heart failure. It also warns against copying restrictive phase 2 exclusions into phase 3, which has a different purpose.

E6(R3) adds that participant selection should represent the populations the drug is intended to benefit, while accepting that early phase or bioequivalence studies may need a uniform group. Every criterion must also be checkable at screening with data the site actually has, or eligibility errors will follow.

How to work it through

  1. Write down the intended treated population and the trial's primary question.
  2. List each candidate criterion and state whether it serves safety, interpretability or neither.
  3. Remove criteria that serve neither, and replace broad exclusions with specific, measurable thresholds.
  4. Check each criterion can be verified at screening from records the site holds.
  5. Estimate the effect of the criteria on the eligible pool using registry or electronic health record data.
  6. Revisit criteria as safety data from other studies accumulate.

An example

A draft phase 3 protocol for a new oral anticoagulant excludes 'any renal impairment' and 'age over 75'. Hospital data show these two criteria remove 38 percent of otherwise eligible patients, and older patients are a large share of future users. The team replaces them with creatinine clearance below 30 mL per minute, where exposure data from a renal impairment study show unacceptable accumulation, and removes the age cap.

The projected eligible pool rises by about a third and results now cover the older patients most likely to receive the drug.

Where the answer stops holding

Broader criteria add heterogeneity, which can reduce power if the effect varies across the population. Some exclusions remain essential, for example known drug interactions or pregnancy where reproductive toxicity is not yet studied. Real world data used to size the eligible pool may be incomplete or coded differently from protocol definitions.

A strong answer covers

  • Each criterion justified by safety or the objective
  • Specific thresholds instead of blanket exclusions
  • Representativeness of the intended treated population
  • Do not copy phase 2 restrictions into phase 3 by default
  • Criteria verifiable at screening

Common mistakes

  • Copying eligibility criteria from a previous protocol without review.
  • Excluding whole groups such as older adults without a safety reason.
  • Writing criteria that the site cannot verify from its records.