Table 1 Evidence for comprehension in informed assent processes.
From: Improving assent in health research: a rapid systematic review
Authors | Study design | Intervention & Control | Outcome | Findings | Critical appraisal |
|---|---|---|---|---|---|
Abramovitch et al. (1995) [8] | Non-RCT Participants: 177 healthy children (7-12 years) Description: 3 sub-studies on memory, hearing loss and personality Simulated scenario | Control: Standard descriptions of the sub-studies (n=131) Intervention: Standard descriptions + probing questions after descriptions of each study (n=46) | Measurement: Understanding of purpose, good things, and bad things Time point: after the description of all 3 studies | With probing questions, children understood purpose and good things better than without interspersed questions; no effect on understanding of bad things | Quality Rating: 2 Incomplete outcome reporting Non-comparable data provided Recruitment methods and selection unclear |
Adcock et al. (2012) [9] | Crossover-RCT Participants: 217 school children (7-11 years) Description: studies on blood pressure and gastroesophageal reflux Simulated scenario | Control: Standard assent form with 2 pages in paragraph form (n=190) Intervention: KidSent assent booklet with 16 pages with sentences and pictures (n=195) First, participants read one of the documents; 3 days later, they read the other document | Measurement: Understanding of study purpose, risks, procedures, and right to withdraw Time point: immediately after reading the respective document | Mean of correct answers for standard form 78.5% and for KidSent booklet 71.8%; significant difference between both groups Other results: more children in KidSent booklet group had perfect scores (34,7%) than in standard form group (22,1%); most children stated that they understood the KidSent booklet better | Quality Rating: 1 2 different studies were covered by the control and intervention form; one might have been more difficult to understand Randomization methods partly unclear |
Annett et al. (2017) [10] | RCT Participants: 64 healthy and ill adolescents (12-17 years) Description: clinical trial on asthma Real scenario | Control: Standard assent process with adolescent and parent together (n=34) Intervention: Separate assent process for adolescent and parent in different rooms (n=34) | Measurement: Understanding of asthma trial medicines, research process, rights and privileges, and risks and benefits Time point: immediately after assent process | In knowledge about risks and benefits, minors over 15 years scored better when assent was separate Younger children showed no difference Other results: Parents of older minors also showed better understanding when assent was separate; 15-17-year-olds scored better in asthma medicine than 12-14-year-olds | Quality Rating: 1 Incomplete outcome reporting Values for understanding between intervention and control not provided No description of randomization process |
Barnett et al. (2005) [11] | RCT Participants: 374 school children (9-11 years, first language English) Description: in 7 schools; study how to convey concept of RCTs Simulated scenario | Control: Standard block text format (n=123) Interventions: (1) Question and answer (Q&A) format (n=126) and (2) Story presentation (n=124) | Measurement: Understanding of randomization, safety and effectiveness, voluntariness, and avenue of redress Time point: immediately after reading information | Mean of correct answers for block text 70.4%, for Q&A 66.9%, and for story 64.2% Other results: Significant difference for amount of participants that answered all questions per topic correct (story presentation scored best) | Quality Rating: 1 No statistics provided for comparison of mean scores No description of randomization process |
Blake et al. (2015) [12] | RCT Participants: 120 adolescents (15-17 years, English-speaking, from youth serving agencies) Description: hypothetical HIV vaccine trial Simulated scenario | Control: Standard paper assent (n=31) Interventions: (1) Standard paper assent with interspersed questions (n=29) and (2) Web-based assent with interspersed questions, videos, and clip arts (n=60) | Measurement: Understanding of assent content Time point: immediately after assent process | Mean of correct answers for paper assent 74.8%, for paper assent with questions 81.8%, and for web-based assent 78.1%; no significant differences Other results: No significant difference for amount of participants that answered at least 80% correct (paper assent with questions best) | Quality Rating: 1 Less interaction with researcher in web-based assent Randomization methods stated |
Chappuy et al. (2008) [13] | Retrospective interviews Participants: 29 ill children (HIV or cancer, 8,5-18 years) Description: participants recently recruited for other clinical trials Real scenario | Linking personal and disease factors and understanding of study processes | Measurement: Understanding of study purpose, protocol design and procedures, risks, direct and indirect benefits, right to withdraw, duration, alternatives, voluntariness Time point: after recruitment for respective clinical trial | Mean of correct answers when trial recruitment took place more than 7 days after diagnosis 46% and when it took place earlier 20.8%; significant difference between both groups Other results: adolescents older than 14 years scored better than younger ones | Quality Rating: 3 No controlled groups Inclusion of several different clinical trials may have influenced outcomes Small sample, potential bias by group imbalances |
Coors et al. (2016) [14] | Non-RCT Participants: 76 healthy and ill adolescents (substance use disorder, 14-17 years, no intellectual deficiency) Description: biobanking and genomics study in several stages Real scenario | Control: Standard risk information Intervention: Standard risk information + additional information on 7 previously identified salient risks | Measurement: Understanding of risks Time point: immediately after assent process | In patients, the additional information on salient risks improved scores significantly In healthy adolescents, there was no significant difference | Quality Rating: 2 Incomplete outcome reporting No description of allocation of participants; numbers per group unclear |
Friedman et al. (2016) [15] | RCT Participants: 568 healthy adolescents (14-17 years, male only, gay or bisexual) Description: survey on online behaviour of gay youth Real scenario | Control: Study information (n=186) Interventions: (1) Study information + requirement to answer 2 questions correctly (n=187) and (2) Study information + requirement to answer 7 questions correctly (n=195) | Measurement: Understanding of risks and voluntariness Time point: immediately after survey | Mean of correct answers for information without questions 63%, with 2 questions 92.5%, and with 7 questions 93%; significant difference between conditions with and without questions Other results: assent significantly rarer completed when questions interspersed | Quality Rating: 1 Questions to assess understanding at the end are the same as used in the intervention Online study with high number of dropouts Male participants only Randomization methods stated |
Interventional study Participants: 101 school children (10-13 years) Description: comic about characteristics of research studies Simulated scenario | Intervention: Comic strip with information on medical research (n=101) | Measurement: Understanding of 8 research aspects Time point: after reading the comic strip | Mean of correct answers for comic strip 83.0%; best score for side effects, worst score for anonymity Other results: survey on user satisfaction | Quality Rating: 4 No control group Recruitment methods and selection unclear Dropouts not described | |
Lally et al. (2014) [17] | RCT Participants: 120 adolescents (16-19 years old, male/female who have sex with men) Description: consent and brochures on characteristics of an HIV vaccine trial Simulated scenario | Control: Standard informed consent (n=42) Interventions: (1) Informed consent with 1-sided supplemental information (presentation of pertinent facts) (n=39) and (2) Informed consent with 2-sided supplemental information (n=39) (presentation of common misconceptions and rebuttal with factual information) | Measurement: Understanding of randomization, interpretation of side effects, and unproven efficacy (part of consent and intervention brochures); understanding of non-brochure topics Time point: immediately after reading the information | Mean of correct answers for consent alone 72.1%, for consent + 1-sided information 78.6%, and for consent + 2-sided information 80.2%; significant difference between consent only and consent + 2-sided information for randomization and side effects Other results: no significant differences for topics not covered by the supplemental brochures | Quality Rating: 1 Some participants are older than 18 years Randomization methods stated 5-point Likert-type response scale potentially inappropriate for understanding items |
Lee et al. (2013) [18] | Interventional study Participants: 123 adolescents (12-17 years) Description: study on Hepatitis B vaccination in youth Real scenario | Intervention: Simplified assent form with every day, non-medical language and supporting graphs in a Q&A format | Measurement: Understanding of procedure, randomization, future benefits, blinding, direct benefit, voluntariness Time point: immediately after reading the form | Mean of correct answers for the simplified assent form 85.8% Other results: 56.1% answered all questions correctly | Quality Rating: 4 No control group Dropouts not described |
Mayne et al. (2017) [19] | Case series with intervention Participants: 2 children (3 years) Description: Story of a toymaker who makes science toys Simulated scenario | Intervention: Interactive nonfiction narrative (powerpoint with photos, clip arts, active buttons) on touch computer; concepts: dialogic reading, sustained shared thinking, cycle telling and retelling | Measurement: Understanding of research purpose and context, participatory rights, and consent Time points: 1 week before, 2 and 9 weeks after outreach | Understanding of the basic research concepts improved or stayed high after presentation of the interactive narrative | Quality Rating: 4 No control group Only 2 selected participants Incomplete outcome data due to erratic interest of participants |
Miranda et al. (2017) [20] | Interventional study Participants: 42 hospitalized children (5-10 years, clinically stable) Description: study on vulnerability during illness and hospitalization Real scenario | Intervention: Illustrated booklet (text, images, illustrations for colouring) | Measurement: Understanding of research proposal Time point: during application of booklet | All children understood the research proposal Other results: Children wanted the booklet to be able to colour it | Quality Rating: 4 No control group Understanding was assessed only by “observations by researcher” Inconclusive outcome reporting |
Murphy et al. (2007) [21] | RCT Participants: 187 healthy adolescents (15-19 years, male/female/ transgender, at risk for HIV, English-speaking) Description: study on HIV vaccination Simulated scenario | Control: HIVNET standard assent form (n=94) Intervention: Based on HIVNET version, but reorganized, simplified text, implementation of illustrations (n=93) | Measurement: Understanding of study details including procedures, benefits and risks Time point: immediately after assent process | Mean of correct answers for standard version 71.7% and for illustrative version with simplified text 80.5%; significant difference between both groups Other results: understanding of procedures and benefits was also significantly better in the intervention group; illustrative version with simplified text contained fewer words, fewer words per sentence, less passive voice, and had higher reading ease | Quality Rating: 1 No description of randomization process No indication of standard deviations Simplified text and illustrations are tested together Some participants are older than 18 years |
O’Lonergan and Forster-Harwood (2011) [22] | RCT Participants: 170 children (11-14 years, no deficits in cognition, hearing, or vision, did not undergo procedures yet) together with parents Description: study involving common procedures in paediatrics (DXA and abdominal ultrasound) Simulated scenario | Control: Standard permission and assent process (n=87) Intervention: Multimedia process in Microsoft PowerPoint with same text like standard process but with hyperlinks to videos and voice-overs (n=83) | Measurement: Understanding of essential elements of the permission and assent process Time point: immediately after assent process | Mean of points for correct answers for standard process 44% and for multimedia process 51.2%; significant difference between both groups for total score, study procedures, and risks Other results: parents also scored significantly better with multimedia process; all participants overestimated their comprehension | Quality Rating: 1 No description of randomization process Incomplete outcome reporting (answers to some questions were not presented individually) |
Tait et al. (2007) [23] | RCT Participants: 190 hospitalized children (7-17 years, no cognitive impairment, no emergent illness) Description: study on postoperative nausea and vomiting Simulated scenario | Control: Standard form including verbal explanation (n=95) Intervention: Modified form with improved readability and processability as well as use of bullets, bolding, increased font size, and pictures (also including verbal explanation) (n=95) | Measurement: Understanding of purpose of study, protocol, risks, direct and indirect benefits, alternatives, voluntariness, and freedom to withdraw Time point: immediately after assent process | Mean of points for correct answers for standard form 60.4% and for modified form 68.5%; significant difference between both groups Other results: differences between groups were higher in younger children; most children preferred modified form; all children overestimated their comprehension | Quality Rating: 1 No description of randomization process Large number declined participation; possibly selection bias of highly motivated children Assessors were blinded |
Tait et al. (2012) [24] | Before and after study Participants: 4 children (8-14 years, from waiting room in hospital) Description: pilot study; trial on asthma Simulated scenario | Intervention: 3D modelled avatars present a dialogue between a child and a doctor in an interactive program | Measurement: Pre- and post-intervention understanding of clinical trial, randomization, placebo, and blinded study; post-intervention understanding of elements of the study Time point: directly before and after using the program | Correct descriptions of the 4 terms from pre- to post-intervention: 25% to 50%, 0% to 0%, 0% to 50%, and 25% to 50%; mean of points for correct answers about elements of the study 61.7% | Quality Rating: 4 No control group Only 4 participants |
Tait et al. (2015) [25] | RCT Participants: 135 children (10-17 years, attendants of a paediatric clinic, no cognitive impairments, English-speaking) Description: study on general aspects of trials Simulated scenario | Control: Standard paper form (text only) (n=68) Intervention: Interactive iPad program in written and visual formats together with voice-over and interactive exercises with corrective feedback (content identical to standard form) (n=67) | Measurement: Understanding of clinical trial, participation, protocol, randomization, placebo, blinding, double-blinding, effectiveness, and informed consent Time point: immediately after reading the information | Mean of points for correct answers for standard form 49.2% and for interactive program 64.7%; significant difference between both groups Other results: most children preferred the interactive program over the standard form | Quality Rating: 1 Randomization methods stated Assessors were blinded |
Ulph et al. (2009) [26] | Cross-sectional study Participants: 106 school children (7-11 years) Description: study on methods to convey probabilities in a cup game Simulated scenario | 6 different formats were tested in all participants: (1) verbal labels (rare) (2) percentages (1%) (3) pie charts (4) proportions as words (1 in 100) (5) proportions as notation (1:100) (6) mixed format | Measurement: 3 trials to choose the highest probability shown for each format Time point: during the game | Mean of points for correct answers was highest for pie charts (90%), followed by verbal labels, percentages (79%), proportions as words (64%), proportions as notation (62.7%), and mixed format (43%) | Quality Rating: 4 Game may not represent complexity of medical research Only understanding of probability was tested Incomplete outcome reporting |