Table 1 RCTs of opt-in and opt-out consent
Study population and purpose | Parents in a health district of the UK were asked for consent for inclusion of low birth-weight infants on a register for the purpose of monitoring disability in children[26] | Mothers in the US were asked for consent for inclusion of infants at high risk to participate in a clinical trial of primary follow-up care[27] | Angina patients in two general practices in the UK were asked for consent to be involved in clinical research[28] | Patients aged 50-74 years in a general practice in Australia were asked to consent to testing decision aids for the screening of colorectal cancer[29] | Cancer patients in the Netherlands who had undergone primary surgery were asked for consent for the storage of excised tissue for future research purposes[30] |
|---|---|---|---|---|---|
Sample size randomised (n) | Opt-in: 39 Opt-out: 30 | Opt-in: 32 Opt-out: 25 (3 were excluded as they did not receive the allocated intervention) | Opt-in: 252 Opt-out: 258 | Opt-in: 92 Opt-out: 60 | Opt-in: 60 'Opt-out plus': 73 Control group (standard opt-out): 131 |
Mode of invitation | Verbal information, letter and reply slip given by a nurse prior to an infant's discharge from hospital | Verbal information and reply form given by a nurse within 24-48 hours of delivery. The opt-out form was shortened to include only specific disclosures that are appropriate for low risk research | Letter, information leaflet and reply card sent from a doctor | Letter sent from a doctor (plus reply card for the opt-in arm only) | Verbal information, specific information leaflet and reply form given by a doctor/nurse. The control group was only given a routine hospital leaflet and did not receive verbal information |
Mode of response | Reply-paid slip | Reply form was collected from the mother | Reply card or telephone | Telephone or email (or reply-paid card for the opt-in arm) | Reply-paid form. The control group leaflet instructed patients to opt out by informing their doctor |
Reminder letter | No | No | After two weeks for the opt-in arm only | No | No |
Time to respond | Not stated | Prior to discharge from hospital. Once a mother reached a decision, an interview occurred within the next 24 hours (usually 2 hours) | Opt-in: Not stated Opt-out: patients could opt out verbally when telephoned after two weeks | Not stated | One month |
Participation rate | Opt-in: 79% Opt-out: 97% | Opt-in: 75% Opt-out: 91% | Opt-in: 48% Opt-out: 59% | Opt-in: 51% Opt-out: 90% | Opt-in: 85% 'Opt-out plus': 97% Standard opt-out: 100% |
Recruitment rate | Not applicable | Face-to-face interview Opt-in: 81% Opt-out: 82% | Clinic attendance Opt-in: 38% Opt-out: 50% | Telephone survey Opt-in: 47% Opt-out: 67% | Postal and telephone survey Opt-in: 93% and 52% 'Opt-out plus': 93% and 51% Standard opt-out: 88% and 47% |
Evidence of selection bias | Not stated | Modest differences were found. Subjects recruited in the opt-in arm were older, more likely to be married and undergo a vaginal delivery than subjects in the opt-out arm | Subjects recruited in the opt-in arm were healthier and had less risk factors for coronary disease than subjects in the opt-out arm | Subjects recruited in the opt-in arm were more likely to prefer an active role in decision making than subjects in the opt-out arm | Subjects recruited in the opt-in arm were similar in age, sex, education and type of cancer to the 'opt-out plus' arm. The control group was similar, except that women were over-represented |
Design flaws | Small sample size, non-random allocation and no mention of whether blinding was used | Small sample size and the collection of reply forms is resource-intensive and impracticable on a large scale | None evident | Small sample size and non-parallel design | Small sample size and no mention of whether blinding was used |