N-of-1 Trial
Single-Patient N-of-1 Randomized Controlled Trial · Also known as: single-patient RCT, n=1 trial, individual RCT, crossover n-of-1
An N-of-1 trial is a single-patient randomized controlled trial in which a patient alternates between treatment A and treatment B (or active drug and placebo) in repeated, randomized cross-over periods. Developed systematically in the 1990s–2010s by Kravitz, Duan, and Vohra, N-of-1 trials enable personalized medicine by determining which treatment works best for that specific individual, avoiding the assumption that population-average effects apply to all patients. They are ideal for chronic conditions with variable outcomes and heterogeneous treatment response.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use N-of-1 trials when: (1) patient has a chronic condition with variable outcomes (pain, fatigue, asthma, migraine, mood disorder), (2) two or more reasonable treatment options exist, but optimal choice for this patient is unclear, (3) patient-reported outcomes are primary (e.g., pain relief, symptom control rather than biomarkers), (4) treatments are reversible and short-acting (no long-term irreversible effects), (5) individual tailoring matters more than population-level evidence (e.g., a treatment average helps population but may harm this patient), (6) traditional RCTs show population benefit but patient is non-responder (seeking personalized evidence), (7) treatments have variable cost or side effects, and patient needs to know which is worth it for them specifically.
Strengths & limitations
- Personalized evidence: directly answers 'Does this treatment work for ME?' rather than average population effect.
- Randomized comparison: despite single patient, randomization eliminates bias and placebo effect.
- Reversibility and fast cycle time: multiple treatment cycles in weeks/months (vs years for population trials) allow quick turnaround.
- Ethical: minimizes prolonged exposure to ineffective treatments; if one treatment is clearly superior, can stop trial early and adopt it.
- Cost-effective: small sample (n=1), minimal infrastructure; can be conducted in clinical practice, integrated into routine care.
- Very small sample: statistical power is low, detecting true treatment differences requires large effect sizes or many cycles. Wide confidence intervals are common.
- Not generalizable: results apply only to the individual, not other patients, even with identical diagnoses.
- Carry-over effects: prior treatment may influence outcomes in subsequent period, biasing comparison. Requires careful washout and design.
- Outcome measurement bias: outcome fluctuates naturally; distinguishing treatment effect from natural variation requires many cycles, extending trial duration.
- Requires motivated patient: patient must adhere, record outcomes diligently, tolerate uncertainty during trial. Loss of engagement is common.
Frequently asked
How many cycles does an N-of-1 trial need?
Typically 2–6 cycles, depending on outcome variability and treatment effect size. If outcome is stable and treatment effect is large, 2–3 cycles suffice. If outcome is highly variable or effect is small, 4–6+ cycles provide better discrimination. General rule: continue until a clear pattern emerges or you've completed pre-specified number of cycles. Statistical power increases with more cycles (more data points per treatment). However, each cycle takes time (1–4 weeks), so very long trials (10+ cycles) become impractical. Balance need for power against patient burden and practical feasibility.
What is the difference between N-of-1 trials and crossover trials?
Crossover trials typically enroll many patients (n>1) who each cross over between treatments in repeated periods. N-of-1 trials enroll a single patient (n=1) in a randomized crossover design. The analysis differs: crossover trials pool data across patients, computing population-average treatment effects. N-of-1 trials analyze the individual, computing individual-specific effects. Both use randomized crossover to eliminate bias. N-of-1 is the extreme case of crossover design.
What is a washout period, and when is it necessary?
A washout period is a gap (days or weeks) between treatment cycles where patient receives neither treatment, allowing residual drug or effects to clear. Necessary if prior treatment has lasting effects that would bias comparison in the next cycle (e.g., some drugs accumulate or have long half-lives). Example: if Drug A is given Week 1–2, washout Week 3, Drug B Week 4–5, the washout prevents Drug A's lingering effects from inflating Drug B's outcome. Washout duration depends on treatment half-life and onset of action. Many interventions (non-reversible procedures, structural changes) do not require washout; conversely, some drugs (SSRIs, some biologics) require extended washout. Consult literature and pharmacology.
Can I use N-of-1 trials for all chronic conditions?
No. N-of-1 works best for conditions with variable, reversible outcomes. Ideal: symptoms that fluctuate (pain, fatigue, mood), can be measured frequently, and where treatments are quickly reversible. Poor fits: conditions with stable, invariant symptoms; treatments with permanent effects (surgery); outcomes only measurable long-term (mortality). Also unsuitable when one treatment is clearly superior (unethical to randomize a patient to inferior treatment). Ensure treatments are reasonable, reversible, and patient is genuinely uncertain which is best.
Sources
- Gabler, N. B., Duan, N., Vohra, S., & Kravitz, R. L. (2011). N-of-1 trials in the medical literature: a systematic review. Medical Care, 49(8), 761–768. DOI: 10.1097/mlr.0b013e318215d90d ↗
- Kravitz, R. L., Duan, N., & Eslick, I. (2010). Evidence-based medicine, heterogeneity of treatment effects, and the trouble with averages. The Milbank Quarterly, 88(4), 503–520. link ↗
- Vohra, S., Shamseer, L., Sampson, M., Barrowman, N., Yap, B., Uleryk, E., ... & Moher, D. (2015). CONSORT extension for reporting N-of-1 trials (CENT) 2015: explanation and elaboration. BMJ Open, 5(7), e007838. DOI: 10.1136/bmj.h1793 ↗
How to cite this page
ScholarGate. (2026, June 4). Single-Patient N-of-1 Randomized Controlled Trial. ScholarGate. https://scholargate.app/en/clinical-research/n-of-1-trial
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Pragmatic Clinical TrialClinical Research↔ compare
- Randomized Controlled TrialExperimental design↔ compare
- Real-World Evidence StudiesClinical Research↔ compare