When Does a Clinical Trial Need a Washout Period?

How Long Should Researchers Wait Between Treatments?

You are designing a randomized trial in which participants will receive more than one intervention. After completing the first treatment, they will begin another. But what happens if the first intervention continues to influence their outcomes after treatment has stopped?

Perhaps a medication remains active in the body. A behavioral intervention may produce lasting habits, while an educational intervention may leave participants with knowledge or skills they cannot simply forget.

Should the trial include a washout period between interventions? If so, how long should it last, and how can researchers determine whether the earlier treatment's effects have sufficiently diminished?

The answer matters because residual treatment effects can undermine the interpretation of subsequent treatment comparisons, particularly in crossover designs.

Use a Washout When Residual Treatment Effects Could Compromise the Comparison

Why Washout Periods Matter and How They Should Be Designed

What Is a Washout Period?

A washout period is a planned interval during which a previous intervention is discontinued or withheld to allow its residual effects to diminish before a subsequent intervention or outcome assessment.

It is most commonly discussed in crossover trials, where participants receive different treatments in successive periods.

For example, participants may receive treatment A during the first period, complete a washout interval, and then receive treatment B. Another randomized sequence may reverse the order.

The washout aims to reduce the possibility that outcomes observed during treatment B are influenced by treatment A.

However, a washout period does not guarantee that participants return to their original condition. Whether the interval is sufficient depends on the intervention, outcome, and underlying process being studied.

What Is a Carryover Effect?

A carryover effect occurs when an earlier treatment influences outcomes in a later treatment period.

Carryover may result from residual medication, persistent physiological changes, acquired skills, altered behavior, or other lasting consequences of the previous intervention.

Suppose a participant receives treatment A followed by treatment B. If the effect of A persists during the B period, the observed outcome under B may reflect both treatments rather than B alone.

This complicates estimation of the treatment effect because the outcome depends partly on treatment history.

Carryover should be distinguished from period effects. A period effect arises from differences associated with the timing of measurements, such as disease progression, seasonal variation, or changes in participants' circumstances.

Randomizing treatment sequences can help distribute order-related influences, but it does not automatically eliminate carryover.

The consequences of carryover effects in crossover trials should therefore be considered before choosing the design.

When Is Washout Particularly Important in Crossover Trials?

In a conventional crossover trial, participants receive more than one intervention in a randomized sequence. Each participant can contribute observations under multiple treatment conditions.

This arrangement can improve statistical efficiency when within-person comparisons are appropriate and outcomes are sufficiently stable or reversible.

However, crossover designs depend on assumptions about the relationship between treatment periods. Residual effects from an earlier intervention can interfere with the intended comparison.

Washout is particularly relevant when treatment effects persist beyond the active administration period but are expected to diminish within a feasible interval.

For instance, a medication may remain in the body after dosing stops, or a temporary physiological response may continue for several days.

By contrast, if an intervention produces permanent or long-lasting changes, extending the washout interval may not restore the conditions needed for a credible crossover comparison.

Researchers should establish whether a crossover trial is appropriate before assuming that washout can solve every treatment-history problem.

How Do Pharmacokinetic and Pharmacodynamic Effects Influence Washout?

For medication trials, washout planning commonly considers both pharmacokinetics and pharmacodynamics.

Pharmacokinetics concerns what the body does to a drug, including absorption, distribution, metabolism, and elimination.

Pharmacodynamics concerns what the drug does to the body, including its biological effects and their duration.

These processes are related but not interchangeable.

A drug may be largely eliminated from the bloodstream while its physiological effects continue. Conversely, measurable concentrations may persist even after the clinically relevant effect has diminished.

Therefore, the disappearance of a drug from circulation does not necessarily establish the disappearance of treatment effects.

A defensible washout interval should consider the duration of both drug exposure and relevant biological consequences, using evidence specific to the treatment and outcome.

How Is Drug Half-Life Used to Estimate Washout Duration?

The elimination half-life of a drug is the time required for its concentration or amount in the body to decrease by half during the relevant elimination phase.

Under a simplified first-order elimination model with a constant half-life, the remaining fraction after a specified number of half-lives can be calculated.

Five half-lives are sometimes used as a practical reference because approximately 96.875% of the modeled amount has been eliminated. This is not a universal washout rule.

Some drugs have active metabolites, nonlinear elimination, tissue accumulation, or prolonged pharmacodynamic effects. Patient characteristics and interactions with other medications can also influence elimination.

The amount remaining after several half-lives does not by itself establish that carryover is negligible for the outcome being measured.

Researchers should use drug-specific evidence and relevant clinical expertise rather than treating a fixed number of half-lives as sufficient in every trial.

Why Is Pharmacological Elimination Not the Same as Recovery?

Even when the active substance is no longer present at meaningful concentrations, the participant's physiological condition may not have returned to its pretreatment state.

For example, an intervention could change receptor regulation, enzyme activity, immune responses, or other biological processes that recover more slowly than the drug is eliminated.

Similarly, stopping a treatment may produce withdrawal effects or a rebound response. These effects can influence outcomes during the washout interval or subsequent treatment period.

Consequently, researchers should consider whether the relevant biological system has recovered sufficiently, not merely whether the previous treatment has been discontinued.

In some studies, repeated baseline measurements or predefined stabilization criteria may provide useful evidence about recovery. Nevertheless, apparent stabilization does not prove that all residual effects have disappeared.

Can Behavioral or Educational Interventions Be Washed Out?

Some behavioral effects diminish after an intervention ends, but others may persist because participants acquire knowledge, habits, or skills.

Consider a trial comparing two approaches to teaching statistical reasoning. Students first receive method A, followed by method B.

Even if several weeks pass between the two approaches, students cannot be assumed to forget what they learned from method A. Their later performance may reflect accumulated knowledge rather than the effect of method B alone.

This creates a fundamental difficulty for crossover designs involving durable learning.

A washout interval may reduce temporary influences such as fatigue or short-lived priming, but it cannot reliably reverse acquired competence.

Similar concerns arise with psychotherapy, professional training, organizational interventions, and behavior-change programs.

When the intervention is intended to produce lasting change, a parallel-group design may be more appropriate than crossover.

When Does a Trial Need Washout Before Randomization?

Washout is not limited to intervals between treatment periods. Some trials require participants to discontinue a previous intervention before randomization so that prior treatment does not interfere with the experimental comparison.

For example, participants may have been using a medication that interacts with the investigational treatment or influences the primary outcome.

A pre-randomization washout may help establish a more comparable starting condition, but it also raises ethical and methodological concerns.

Withdrawing an effective treatment can expose participants to deterioration, withdrawal symptoms, or other harms. It may also exclude people unable to tolerate discontinuation, changing the population entering the randomized trial.

The procedure should therefore be justified through the scientific objectives, safety evidence, and applicable ethics requirements.

A pre-randomization washout can be part of a run-in period, but the two terms should not be treated as synonyms.

How Long Should a Washout Period Be?

Washout duration should be based on the expected persistence of the relevant treatment effects and the amount of residual influence that would be acceptable for the planned comparison.

Researchers should consider the intervention's mechanism, available pharmacokinetic or pharmacodynamic evidence, the outcome being measured, and the natural course of the condition.

For nonpharmacological interventions, relevant considerations may include the duration of temporary behavioral effects, fatigue, learning, adaptation, or other responses.

A longer washout is not automatically better. Prolonged intervals can increase dropout, extend study duration, expose participants to untreated symptoms, or introduce additional changes unrelated to treatment.

The appropriate interval is therefore a balance between reducing carryover and preserving participant safety and study feasibility.

The detailed determination of washout period duration should follow evidence about the specific intervention rather than a generic calendar interval.

Can Researchers Test Whether Washout Was Sufficient?

Researchers may examine pre-period measurements, treatment concentrations, physiological indicators, or other evidence relevant to residual treatment effects.

However, a nonsignificant statistical test for carryover does not establish that carryover is absent. Such tests may have limited power, particularly in small crossover trials.

Using a preliminary carryover test to decide whether to analyze both treatment periods or only the first period can also create statistical problems.

A stronger approach is to justify washout prospectively using scientific evidence, specify the assumptions required for the crossover comparison, and plan appropriate sensitivity analyses when uncertainty remains.

If meaningful carryover is likely and cannot be adequately controlled, the design itself may need reconsideration.

Does Every Crossover Trial Need a Washout?

Not necessarily. A separate washout interval may be unnecessary when treatment effects are sufficiently short-lived, residual influence is negligible for the intended outcome, or the design includes other justified procedures that address treatment history.

However, the absence of a washout should be justified rather than assumed harmless.

Some crossover trials incorporate an initial period after treatment switching during which outcomes are not included in the principal analysis. Others may use treatment periods long enough for relevant transient effects to diminish before measurement.

These approaches require explicit assumptions and do not automatically resolve persistent carryover.

The decision should reflect the intervention and the intended estimand, not simply the convenience of a shorter protocol.

When Is Washout Inappropriate or Unsafe?

A washout may be inappropriate when discontinuing treatment creates unacceptable risks or when residual effects cannot reasonably disappear within a feasible period.

For example, withholding effective treatment from participants with a serious condition may lead to deterioration. In other circumstances, stopping medication may cause withdrawal or rebound effects that compromise both safety and interpretation.

Interventions intended to produce permanent change also pose a different problem: waiting longer does not necessarily restore the participant's earlier condition.

In these circumstances, a parallel-group trial may offer a more credible and ethically defensible comparison.

Planning Washout Between Two Treatments in a Crossover Trial

The calculation does not establish that 50 hours is a clinically sufficient washout. It only estimates elimination under simplified assumptions.

Suppose the treatment's biological effect continues for several days after the drug is largely eliminated. In that case, a 50-hour interval could still permit meaningful carryover.

Conversely, if the relevant effects resolve rapidly and discontinuation is safe, a shorter interval might be scientifically defensible.

The correct duration depends on the evidence concerning both residual exposure and residual effects. It cannot be determined from elimination half-life alone.

Common Misconceptions About Washout Periods

How to Decide Whether Washout Is Necessary

Begin by identifying whether the proposed trial exposes participants to successive interventions or requires discontinuation of previous treatment before the experimental comparison.

Then assess whether earlier exposure could influence later outcomes and whether those effects are reversible within a practical and ethically acceptable interval.

When washout is included, prespecify its duration, scientific rationale, treatment discontinuation procedures, safety monitoring, and any criteria for beginning the next period.

Where appropriate, collect evidence of recovery or stabilization before the subsequent intervention. Such evidence may support the washout rationale but should not be interpreted as proof that all residual effects have disappeared.

Finally, ensure that the analysis reflects the treatment sequence, period structure, and assumptions about carryover.

Before Adding a Washout Period to Your Trial

Questions About Washout Periods in Randomized Trials

Washout Must Address Residual Effects, Not Simply Separate Treatments

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.