AN INDEPENDENT PUBLIC HEALTH READING ROOM
Places. People.
Possibilities.

Why Some Trials Have a Run-In Period Before Randomisation

A clinical research schedule on a clipboard with indistinct text and an empty consultation chair.
AI-created conceptual illustration of a fictional setting; not a record of people or outcomes.

The first participant in a clinical trial does not always move straight from consent to random assignment. Some studies include a run-in period: an initial phase before randomisation that may involve treatment, a placebo or other specified procedures.

For someone reading a trial report, this phase raises a useful question. Did the people eventually randomised differ from the people who first entered the study, and what happened between those two points?

Read the purpose before judging the design

A published methodological study of run-in periods describes different purposes and reporting practices. Such periods may be used to assess adherence or tolerability, among other objectives. The presence of a run-in is not, by itself, proof that a trial is unreliable. Its consequences depend on what the researchers did and how they accounted for it.

Look for the duration, the intervention or procedures involved, and the rules for continuing to randomisation. Do not assume a run-in, an eligibility screening visit and a washout period mean the same thing. A report should explain its own sequence.

Follow the people who did not continue

Consider a fictional study in which 100 people enter a run-in and 80 are subsequently randomised. The treatment comparison concerns those 80 randomised participants. The other 20 still matter when asking how the study population was formed.

Reasons for not continuing could have very different implications. A participant who withdraws because of scheduling difficulties is not the same as one excluded after an adverse effect or failure to meet an adherence requirement. Those are illustrative possibilities, not a description of any particular trial.

In the report’s flow diagram and methods, ask how many people entered this phase, how many left, and whether the reasons are reported clearly. If information is absent, describe it as missing rather than inventing a motive for exclusion.

A further question is how well the randomised group represents the people for whom the results may later be discussed. Selecting participants able to tolerate or follow a regimen can affect that interpretation. It does not make every conclusion invalid, but it changes the boundaries of the evidence.

This is a way to read research, not a reason to alter prescribed treatment or a trial protocol. If you are considering participation, the study team can explain whether there is a run-in, what it involves and what support is available if you decide not to continue.