Effects of interventions to improve recruitment
| Intervention Reference ID | Participants recruited | Risk ratio (95% CI) | Absolute difference (%)* | Heterogeneity | Risk of bias† (studies) | |||
|---|---|---|---|---|---|---|---|---|
| Intervention | Comparator | χ2 | p Value | I2 (%) | ||||
| Trial design | ||||||||
| Open vs blind design16 32 | 1161/2339 | 1029/2494 | 1.22 (1.09 to 1.36) | 9 | 2.74 | 0.10 | 64 | A│A (2) |
| Active comparator vs placebo59‡ | 65/218 | 85/218 | 0.76 (0.59 to 0.99) | −9 | – | – | – | C (1) |
| Patient preference vs conventional RCT18 | 90/135 | 97/138 | 0.95 (0.81 to 1.11) | −4 | – | – | – | A (1) |
| Obtaining consent | ||||||||
| Consent process Opt-out vs opt-in55 | 40/60 | 44/92 | 1.39 (1.06 to 1.84) | 19 | – | – | – | A (1) |
| Consent to experimental vs usual48‡, 50‡ | 1087/1320 | 920/1136 | 1.01 (0.98 to 1.05) | 1 | 0.42 | 0.51 | 0 | B│C (2) |
| Consent to standard vs usual48‡, 50‡ | 325/623 | 920/1136 | 0.76 (0.49 to 1.17) | −19 | 14.74 | <0.001 | 93 | B│C (2) |
| Refusers choose treatment vs usual50‡ | 482/607 | 836/985 | 0.94 (0.89 to 0.98) | −5 | – | – | – | C (1) |
| Physician modified consent vs usual48‡ | 91/150 | 84/151 | 1.09 (0.90 to 1.32) | 5 | – | – | – | B (1) |
| Participant modified consent vs usual48‡ | 85/150 | 84/151 | 1.02 (0.83 to 1.24) | 1 | – | – | – | B (1) |
| Consent form Researcher read vs participant read56 | 27/51 | 25/53 | 1.12 (0.76 to 1.65) | 6 | – | – | – | C (1) |
| Approach to participants | ||||||||
| Delivery of information Full video presentation+Q&A vs standard info+brief video+Q&A35‡ | 64/93 | 66/95 | 0.99 (0.82 to 1.20) | −1 | – | – | – | B (1) |
| Video presentation+written information vs written only60 | 26/42 | 17/48 | 1.75 (1.11 to 2.74) | 26 | – | – | – | B (1) |
| Audiovisual information on trials vs standard21 22 33 | 88/247 | 82/248 | 1.20 (0.75 to 1.91) | 7 | 4.00 | 0.14 | 50 | B│C│A (3) |
| Interactive computer presentation vs paper-based information36‡ | 23/30 | 17/30 | 1.35 (0.93 to 1.96) | 20 | – | – | – | C (1) |
| Interactive computer presentation vs audio-taped information44‡ | 31/50 | 21/50 | 1.48 (1.00 to 2.18) | 20 | – | – | – | B (1) |
| Verbal educational session+information brochure vs brochure only45 | 178/232 | 147/218 | 1.14 (1.01 to 1.28) | 9 | – | – | – | C (1) |
| Supplementing information Booklet on trials+standard information vs standard information only23‡, 34 | 27/53 | 25/57 | 1.18 (0.64 to 2.18) | 8 | 2.38 | 0.12 | 58 | C│C (2) |
| Study questionnaire with invitation vs invitation only31 37 | 333/1483 | 281/1470 | 1.14 (0.77 to 1.64)§ | 3 | 4.41 | 0.04 | 77 | A│A (2) |
| Newspaper article+study information vs study information only51 | 73/2243 | 71/2245 | 1.03 (0.75 to 1.42) | 0 | – | – | – | B (1) |
| Favourable article+information vs standard article+information51 | 57/1374 | 54/1371 | 1.05 (0.73 to 1.52) | 0 | – | – | – | B (1) |
| Framing and content Treatment described as working ‘twice as fast’ vs ‘half as fast’52‡ | 35/52 | 20/48 | 1.62 (1.10 to 2.37) | 26 | – | – | – | B (1) |
| Trial of treatment described as new vs treatment described as standard38‡ | 43/64 | 50/60 | 0.81 (0.66 to 0.99) | −16 | – | – | – | C (1) |
| Information emphasising pain involved vs standard information54‡ | 18/51 | 30/47 | 0.55 (0.36 to 0.85) | −29 | – | – | – | B (1) |
| Information emphasising risk involved vs standard information54‡ | 13/50 | 30/47 | 0.41 (0.24 to 0.68) | −38 | – | – | – | B (1) |
| Negative framing vs neutral framing of side-effects/survival43‡ | 20/30 | 23/30 | 0.87 (0.63 to 1.20) | −10 | – | – | – | B (1) |
| Positive framing vs neutral framing of side-effects/survival43‡ | 18/30 | 23/30 | 0.78 (0.55 to 1.11) | −17 | – | – | – | B (1) |
| Total information disclosure vs standard disclosure53 | 27/29 | 23/28 | 1.13 (0.93 to 1.38) | 11 | – | – | – | A (1) |
| Less detailed information on risk and benefits vs more detailed information27‡ | 4/10 | 3/9 | 1.20 (0.36 to 3.97) | 7 | – | – | – | B (1) |
| Information leaflet+verbal explanation vs information leaflet only27‡ | 10/18 | 7/19 | 1.51 (0.73 to 3.10) | 19 | – | – | – | B (1) |
| Telephone contact Telephone reminder vs no reminder31 49 | 165/536 | 117/522 | 1.66 (1.03 to 2.46)§ | 15 | 2.44 | 0.12 | 59 | A│A (2) |
| SMS messages (inc quotes) vs no SMS messages26 | 17/405 | 0/406 | 35.09 (2.12 to 581.48) | 4 | – | – | – | A (1) |
| Eligibility screening Enhanced recruitment (inc African American interviewer) vs standard24 | 78/3079 | 95/3297 | 0.88 (0.65 to 1.18) | 0 | – | – | – | B (1) |
| Enhanced recruitment+baseline data by telephone vs standard24 | 87/3075 | 95/3297 | 0.98 (0.74 to 1.31) | 0 | – | – | – | B (1) |
| Enhanced recruitment+baseline data face-to-face vs standard24 | 116/2949 | 95/3297 | 1.37 (1.05 to 1.78) | 1 | – | – | – | B (1) |
| Researcher-administered screening questionnaire vs standard paper based29‡ | 42/78 | 76/141 | 1.00 (0.77 to 1.29) | 0 | – | – | – | C (1) |
| Electronic screening questionnaire vs standard paper based29‡ | 69/151 | 76/141 | 0.85 (0.67 to 1.07) | −8 | – | – | – | C (1) |
| Telephone screening vs face-to-face screening20‡ | 64/99 | 190/370 | 1.26 (1.06 to 1.50) | 13 | – | – | – | C (1) |
| Eligibility screening by senior investigator vs screening by research assistant46 | 28/162 | 22/185 | 1.45 (0.87 to 2.44) | 5 | – | – | – | C (1) |
| Financial incentives | ||||||||
| Cash incentive+study information vs information only26 | 13/246 | 1/245 | 12.95 (1.71 to 98.21) | 5 | – | – | – | A (1) |
‡Recruitment to a hypothetical trial.
*Absolute difference between the intervention and comparator groups (for multistudy interventions, this was calculated using the risk ratio and average comparator group risk).
†Risk of bias: A, low; B, moderate; C, high.
§Analysed as an OR and converted to a risk ratio using the average comparator group risk.