Subjective vs objective endpoints (s.5.7)

PAAB code section 5.7 states:

Comparative claims of efficacy and safety generally require support of evidence from head-to-head well-designed, adequately controlled, blinded, randomized clinical studies. Open-label studies are generally not considered to be a high level of evidence and are not acceptable if subjective end-points are included in the study. Comparative claims should be relevant to current medical opinion and practice.

PAAB frequently receives questions relating to the terms “subjective” and “objective” in this section of the Code.

The term “subjective endpoints” refers to variables whose values are vulnerable to systematic bias from EITHER i or ii:

• The subject’s awareness of his/her group assignment
• Explanation: In such cases, the subject’s behaviors, attitudes, and expectations can inadvertently influence the outcome in a systematic way. Some examples:
• The individual’s experience of pain may be influenced by knowledge of which analgesic he/she is taking
• Knowledge of which therapy the patient has been assigned to may influence how he/she completes a Patient Reported Outcomes questionnaire

• ii. The researcher’s knowledge of the subject’s group assignment
• Explanation: In such cases, the researcher’s behaviors, attitudes, and expectations can inadvertently impact the outcome in a systematic way. Some examples:
• May impact how instructions are provided to the subject (tone, encouragement, expressed optimism…)
• May impact how the investigator interprets diagnostic imaging/scans

Blinding is not required in cases where the variable is vulnerable NEITHER to subject knowledge of group assignment NOR researcher knowledge of the subject’s group assignment (e.g. overall survival rate or blood cholesterol levels measured using a digital meter). These are examples of objective endpoints.

In cases where the variable is vulnerable only to one of the party’s knowledge of the subject’s group assignment, a methodology blinding only the vulnerable party will satisfy s5.7. For example, in some cancer studies, only the assessor interpreting the scans for tumor presence and/or size are blinded.

Unsure about whether a study meets the blinding provisions discussed in PAAB 5.7? Submit an opinion to PAAB with your specific query. See the fee schedule on our website for details. Comparative claims of efficacy and safety generally require support of evidence from head-to-head well-designed, adequately controlled, blinded, randomized clinical studies. Open-label studies are generally not considered to be a high level of evidence and are not acceptable if subjective end-points are included in the study. Comparative claims should be relevant to current medical opinion and practice.