Module 14 2022

21/03/2022

Techniques to avoid bias - randomization

• Assignment of devices or no/other treatment to patients by chance to produce device groups with similar distribution of prognostic factors, both known and unknown. • Avoids possible bias in the selection and allocation of patients to device groups • The randomisation schedule: • is a sequential list of devices, or corresponding codes, by patient number • should be filed securely to ensure that blindness is properly maintained • is accessible in an emergency to break the blind for any patient

The Organisation for Professionals in Regulatory Affairs

Statistical Considerations in Medical Device Clinical Investigations

17

Techniques to avoid bias - blinding

• Blinding is accomplished by coding the interventions so that one or more parties do not know what is being administered • In a single-blind trial, only patients are blinded and in a double-blind trial, patients and all others directly involved in the study. • Blinding is often difficult or impossible in a clinical investigation of a medical device. • Care must be exercised to minimize bias, e.g. by ensuring that the evaluator is blinded to the assignment of patients to a particular intervention or control group.

The Organisation for Professionals in Regulatory Affairs

Statistical Considerations in Medical Device Clinical Investigations

18

9