Loading
Whether patients and doctors know which treatment is being given. Double-blind: neither knows (a placebo hides it). Open-label: both do, unavoidable for surgery or radiotherapy but a source of bias when judging progression and symptoms.
Blinding protects subjective outcomes (progression on scans, symptom scores, decisions to stop or switch) from expectation bias. Many oncology trials are open-label because the comparators differ in route or schedule; they compensate with blinded independent central review of imaging and hard endpoints such as overall survival. Placebo-controlled designs are standard for maintenance and adjuvant trials (adding a pill or nothing). Unblinding at interim analysis, or effective unblinding through side effects (alopecia, rash), can compromise a trial, and open-label designs tend to show larger investigator-assessed PFS effects than blinded review.
Shares Blinded trial, Intention-to-treat (ITT) and per-protocol analysis, N-of-1 trial, Stratified randomisation, allocation concealment and minimisation.
Shares ACT IV, Control arm and comparator (investigator's choice), Single-arm trial.
Shares N-of-1 trial, Control arm and comparator (investigator's choice), Clinical equipoise.
Shares N-of-1 trial, Control arm and comparator (investigator's choice), Add-Aspirin.
Shares Clinical equipoise, Single-arm trial.
Shares Clinical equipoise, Add-Aspirin, Placebo.
Shares Intention-to-treat (ITT) and per-protocol analysis, ACT IV, Control arm and comparator (investigator's choice), Single-arm trial.
Shares Clinical equipoise, Single-arm trial.