Spring Intro 2023

16/02/2023

Blinding

• “Blinding” means that the identity of study treatment remains unknown even after randomisation

• In a double-blind study neither the investigator nor the patient knows which treatment group they have been assigned to

• In a single-blind study one party is unaware of treatment assignment while the other party knows

• In an open-label trial both parties know which treatment has been assigned

Note: Some studies are randomised, even though they are open-label

The Organisation for Professionals in Regulatory Affairs

47

Bias

• Bias can be conscious or subconscious • It is essential to minimise the potential for bias that might affect the conclusions drawn from a clinical trial

• Randomised • Double-blind • Controlled • Prospectively analysed (pre-planned) • Hard endpoints (survival, plasma drug concentrations) Low potential for bias

• No randomisation • Open / single-blind • Not controlled • Retrospectively analysed (“post-hoc”) • Soft endpoints (quality of life, perception of pain) High potential for bias !

The Organisation for Professionals in Regulatory Affairs

48