| Is consent required? | Rationales provided | ||
| Principles/policies* | Role of identifiability in influencing decision | Access by outside research assistant | |
| Yes (n = 14) | Respect for individuals (14)REB policy (2)Consent is practicable (3) | Concern over identifiability (10)No concern (3)Not raised (1) | All 14 expressed concern. For 7, this was either the reason or an important factor for requiring consent |
| “It depends” (n = 3) | – | Pivotal role (3). Consent required if needed full postal code, birth date, ethnic origin. Would drop requirement if reduced data specificity (eg, age instead of birth date) | Not raised in the rationale |
| No (n = 10) | Minimal risk (7)REB policy (4)Province allows (1)Consent impracticable (1) | Identifiability is an important criterion, but not an issue here, because data are anonymous (6)Identifiability not relevant in consent decision (4) | Not raised in the rationale |
| Notification and opt-out (n = 2) | Precedent of other case (1)No rationale (1) | – | Not raised in the rationale |
NB One site was dropped from the analyses, as the opinion at that site was that the scenario did not represent research but quality assurance.
*Totals may add up to greater than the denominator, as more than one rationale may have been cited.