Randomized comparison of two interventions to enhance understanding during the informed consent process for research

Abstract

Background/Aims:

Many investigators have tested interventions to improve research participant understanding of information shared during the informed consent process, using a variety of methods and with mixed results. A valid criticism of most consent research is that studies are often conducted in simulated research settings rather than ongoing clinical studies. The present study rigorously tested two simple and easily adoptable strategies for presenting key consent information to participants eligible to enroll in six actual clinical trials (i.e. six parent studies).

Methods:

In collaboration with the study team from each parent study, we developed two consent interventions: a fact sheet and an interview style video. The content of each of the intervention was based on the information shared in the consent form approved for each parent study. Participants were randomized to the standard consent process, or to one of two interventions. Once exposed to the assigned consent mode participants were asked to complete an assessment of understanding. The study was powered to determine whether those exposed to the fact sheet or video performed better on the consent assessment compared to those exposed to the standard consent. We also assessed participant satisfaction with the consent process.

Results:

A total of 284 participants were randomized to one of the three consent arms. Assessments of understanding were completed with a total of 273 participants from July 2017 to April 2019. Participants exposed to the video had better understanding scores compared to those exposed to the standard consent form process (p-value = .020. Participants were more satisfied with the video when compared to the standard consent. Participants who received the fact sheet did not achieve higher overall understanding or satisfaction scores when compared to the standard consent process.

Conclusions:

This randomized study of two novel consent interventions across six different clinical trials demonstrated a statistically significant difference in participant understanding based on overall scores among those exposed to the video intervention compared to those exposed to the standard consent.

Introduction

Many investigators have tested interventions to improve research participant understanding of information shared during the informed consent process, using a variety of methods and with mixed results.^1,2,3 There is evidence that: a) making consent material simpler and shorter^{4,5,6,7,8,9,10,11,12} and b) increasing dialogue^13,14,15,16 can improve understanding, while the utility of video and computer-based interventions is not yet clear.^{11,17,18,19,20,21,22,23} Of note, video presentations tested previously were designed to present the same consent material to participants as longer consent forms, but using video as the delivery medium. In addition, two recent studies indicate that video presentations to promote research literacy in advance of presenting consent information have been successful in Uganda and South Africa.^{24, 25} We are unaware of studies testing the effectiveness of videos delivering streamlined consent information, with smaller chunks of key information selected for emphasis.

A valid criticism of most consent research is that studies are often conducted in simulated research settings rather than ongoing clinical studies.¹ Simulated research may overestimate the value of specific enhancements by minimizing the impact of other influences on real patients’ understanding of research.^26,27,28,29 On the other hand, investigators running busy research programs may be hesitant to collaborate; they may think informed consent research collaborations carry added burdens: collecting additional data, confusing participants by enrolling them in a consent study and a clinical study, and slowing recruitment.³⁰ We are aware of only a handful of studies that have tested consent interventions in actual clinical studies.^{23,31,32,33,34,35,36}

Conscientious investigators need consent interventions that are relatively simple to create and easy to administer that can improve participants’ understanding beyond what is achieved with our existing lengthy and legalistic process. The present study rigorously tested two simple and easily adoptable strategies for presenting key information: a fact sheet and an interview-style video. The two interventions were compared to a standard consent form using a three parallel arm randomized trial in the context of six actual clinical studies, with the primary goal of assessing understanding when compared to the standard consent process in different clinical research settings. Our secondary goal was to assess participant satisfaction with the consent process. We hypothesized that either intervention would result in better understanding and expected more satisfaction when compared to the standard consent process.

Methods

The institutional Vice-Dean for research, assisted by a member of the IRB staff, identified clinical trials enrolling at our institution during the study period as potential studies for collaboration. Once candidate studies were identified, we reached out to the PIs with a one-page summary of our study. Ultimately, six studies were eligible, due to a combination of their enrolling sufficient numbers of participants and the principal investigator being willing to collaborate. Eligible participants for our consent study were English speaking adults 18 years or older who were being invited to join one of the six collaborating studies. We requested and received IRB approval to pre-randomize participants of these six collaborating studies to one of our three study arms; that is, the IRB approved the use of any of the three consent approaches to enroll participants in their respective clinical trial. Participants undergoing consent for these six studies were randomized into one of the three informed consent approaches created with their study-specific information. After informed consent, participants provided verbal consent to undergo an assessment of their understanding, using the Consent Understanding Evaluation - Refined (CUE-R) (See Appendix). The original CUE was developed and tested in previous work conducted by our team.³⁵ Additional details about CUE-R are provided below. After participants in the factsheet and video consent arms completed CUE-R, they received the standard consent form, a commitment made to the IRB as a condition of pre-randomization (Figure 1).

Figure 1:

Study Design

The primary study protocol (including the CUE-R) was approved by the [institutional review board] with which the investigators implementing the study were affiliated. A data and safety monitoring board was established for timely review of study results. In coordination with collaborating PIs, amendments to their previously approved studies were submitted to the [name of institutional review board] to revise their informed consent process such that they could use either the original consent form, the fact sheet, or the video (a script of the video was submitted). In some cases, the IRB asked to review the CUE-R. Amendments were reviewed and approved prior to data collection.

Interventions

Two consent interventions were developed: a fact sheet and an interview-style consent video. Both experimental interventions build on principles from relevant theories of learning: define a limited set of the most important learning goals and eliminate all other information, present information in discrete “chunks”, use plain language, and link information to specific learning goals.³⁷ We created the fact sheets and videos based on these four precepts for each of the six collaborating studies. The traditional consent form served as the control arm. Summary of the transformation from standard consent language to fact sheet and video script is documented in Table 1.

Table 1:

Page and Word Count of Standard Consent Form and Bulleted Fact Sheet, Length of Video

Study	Standard Informed Consent Form		Fact Sheet		Percent decrease in words	Video length (min)
	Pages	Words	Pages	Words
1	11	4742	5	1264	73%	9:20
2	12	4911	6	1768	64%	11:29
3	7	2612	4	979	63%	5:11
4	16	6303	8	2881	54%	20:54
5	9	3349	5	1496	55%	8:12
6	7	2833	4	1156	59%	7:57

Fact Sheet

Written consent summaries unique to each collaborating study were developed.³⁵ In creating these fact sheets, the priority was to identify, in conjunction with the collaborating PI and study team, the key elements of each study that researchers thought a participant should understand. All six fact sheets used similar section headings and standardized language relating to generic information, where appropriate (e.g., statements of no direct benefit, or informational risks). Each concluded with highlighted text boxes that summarized key points and flagged key responsibilities of both the study team and participant.

Video

The second consent intervention was an original, scripted, interview-style consent video, designed to be viewed on a tablet computer. The “interview” was conducted between a prospective participant (played by an actor) and the actual PI of the collaborating study. All video content was in question-answer format and ended with both the PI and study participant reciting brief summaries of their respective responsibilities. Video content very closely mirrored that of the fact sheets in order to essentially isolate the effect of the mode of information delivery.

Assessment of Understanding and Satisfaction

For this study, we refined and added satisfaction questions to an assessment tool we used in a prior study.³⁵ The original CUE was influenced heavily by others’ prior work.^{38,39,40,41,42} The master version of the CUE-R (a refined version of the original CUE) includes nine open-ended and 28 close-ended questions across six domains.

Of the nine open-ended questions all are related to key elements of understanding. Six of the nine questions were asked of every participant and cover what the study is about, what will happen if they join the study, what are the good things/bad things that might result if they join, and whether they can say no or decide to leave the study. Those asked to join a study that included randomization (n=1 trial) where asked two additional questions, whether they would be randomized and why and those asked to join a randomized study that included a placebo (n=2 trials) were also asked one additional question about what the term placebo meant to them.

Of the 14 closed-ended questions meant to assess participant understanding, seven were the same across all collaborating studies, while seven others were contextually modified to assess the same type of information (e.g., length of the study) for each individual study. All participants were asked at least 5 close-ended questions. As with the open ended questions those asked to join a randomized trial or randomized trial that included a placebo were asked up to 9 additional questions regarding their understanding. The up to 9 open-ended questions and up to 14 close-ended questions regarding participant understanding were combined to create the overall score of understanding (the number correct/the number asked).

Satisfaction with the consent process was assessed through a set of four questions about: whether they read/paid attention to the form/fact sheet/video (i.e. mode), how easy it was to understand their mode of consent, how complicated their mode was, and whether they would recommend that researchers use the mode they were exposed to in the future, each gauged with a 5-point Likert scale. A composite satisfaction score was calculated.

Additional domains (unrelated to the understanding score and satisfaction) in the closed ended portion included: participant opinions of research in general and the consent process and their motivation for joining the clinical study. The CUE-R also included nine demographic questions.

Our primary hypothesis was that exposure to the fact sheet or video would result in better understanding compared to standard consent (perfect score of 1.0, or 100%). Our secondary hypothesis was that participants would be more satisfied with the fact sheet or video when compared to standard consent (perfect score of 1 on each item, 5 total; scaled from 0–100 with 0 being most satisfied, 100 being least satisfied, and lower score indicating more satisfaction).

After going through one of the three consent approaches, potential participants were asked about their willingness to be interviewed by phone for 15–20 minutes “by a colleague interested in how well we explain our research studies to patients.” Participants were free to refuse participation in the assessment. Participants orally consented to complete the CUE-R assessment interview. After completion of the interview, consent study staff further disclosed to participants the fact of randomization to different consent modalities and the reason for pre-randomization. Participants received a $20 gift card for participating in the CUE-R assessment.

Randomization and Blinding

Participants were randomly assigned with equal allocation within each collaborating study to one of the three consent arms. Block randomization (with fixed blocks of three) was used to assign participants to one of the three arms. Randomization assignments were generated by the study biostatistician using standard statistical software and delivered to collaborating study staff who delivered the assigned mode. Masking of assignment to participants and outcome assessment staff was impossible; but the section of the assessment dedicated to questions about the mode of consent delivered came at the end of the assessment. In addition, all assessment interviews were audio recorded and reviewed for evidence of overt bias and quality assurance.

Sample size

The primary outcome of interest was a comparison of the mean proportion correct on understanding items for those exposed to the fact sheet or video compared to the mean proportion correct for those exposed to the standard consent form, respectively. We anticipated being able to detect a mean improvement in the proportion of correct scores of 0.103 (or 10.3%, e.g., a mean score of 0.73 in the standard arm and a mean score of 0.833 for those exposed to either novel intervention), with a power of 80% in a two-sided test with an alpha of 0.025 (to account for the two primary testing contrasts for the primary outcome) with a sample of 300 participants and a conservatively assumed standard deviation (SD) of 0.23 for this outcome in the population. We based this expected difference on what we believed to be a meaningful distance as well as on findings from our previous study that indicated a trend toward a 10% difference in proportion of correct scores when comparing understanding based on the standard form versus an intervention to improve understanding.³⁵ We did not power the study to detect a pre-specified difference in satisfaction but with this sample size, we had 80% power to detect an effect size of 0.45 (i.e., 0.45 SD) for the satisfaction outcome comparing each intervention arm to the standard consent arm using a two-sided test with an alpha of 0.025.

Data Entry and Analysis

CUE-R responses were collected in real time by consent study staff and entered into a Qualtrics (Provo, UT 2019) version of CUE-R. The assessment was audio recorded to facilitate verbatim transcription of responses to open-ended questions, as well as to support data checks for quality assurance. Data were exported to Excel and open-ended responses coded. Responses to open-ended questions were scored on a scale of 0 (incorrect) to 1.0 (correct). Some responses were given half credit. The consent study team coded the responses without knowledge of the intervention to which the participant was randomized. Responses were initially coded by one study team member and verified by another. Discrepancies were discussed and resolved as a team. The percent correct knowledge scores and the scaled satisfaction score were calculated and analyzed using SAS statistical software version 9.4 (Cary, NC 2018). The primary contrasts of fact sheet and video vs standard consent were each evaluated using a 2-sided test with alpha of 0.025 for the primary outcomes using analysis of variance. For other comparisons, we reported estimates and corresponding 95% confidence intervals. General linear models (GLM) were used to explore understanding and satisfaction associated with sociodemographic variables independent of intervention effects, and potential effect modification by sociodemographic variables were explored by including appropriate cross-product interaction terms in the GLM models.

Implementation Evaluation

In addition to the data collection described above, we conducted interviews with study coordinators involved in the delivery of the consent interventions. We utilized a structured interview tool, adapted from exit interviews conducted under our previous consent research.³⁵ Topics covered included: overall burden, study logistics, experience with each consent intervention, and whether they would recommend use of either consent approach in the future. Participation was optional and oral consent was obtained.

Results

A total of 284 participants were randomized to one of the three consent arms. Eleven individuals were exposed to the consent interventions but chose to not complete the CUE-R assessment (3 assigned to standard consent; 5 to fact sheet; 3 to video). We completed assessments with a total of 273 participants from July 2017 to April 2019. We stopped short of our target recruitment of 300 participants given the available financial and human resources, as well as the expected rate of accrual across remaining collaborating studies. While our intent was to enroll 60 participants from each of five collaborating studies, the actual enrollment across studies varied from five to 100 participants from the six studies (see Table 2).

Table 2:

Accrual across Collaborating Studies

Study	Study Title (ClinicalTrials.gov Registry Number)	Intervention Assignment			Total n (%)
		Standard n (%)	Fact Sheet n (%)	Video n (%)
1	Randomized Controlled Trial of An Internet Cessation Program Plus Online Social Network for HIV + Smokers (NCT02781090)	30 (30)	35 (35)	35 (35)	100 (100)
2	Enhancing Medication-based Analgesia in Humans (NCT02901275)	16 (33)	15 (31)	18 (36)	49 (100)
3	NanoSpectrometer Biomarker Discovery and Confirmation Study (NCT03275688)	10 (36)	9 (32)	9 (32)	28 (100)
4	Clinical Trial of Air Cleaners to Improve Indoor Air Quality and COPD Health (NCT02236858)	11 (38)	9 (31)	9 (31)	29 (100)
5	A Randomized Controlled Trial for Prevention of Venous Thromboembolism following Radical Prostatectomy (NCT03006562)	21 (34)	20 (32)	21 (34)	62(100)
6	Clinical performance study plan for Icare ic200 Tonometer (n/a)	1 (20)	2 (40)	2 (40)	5 (100)
	Total	89 (32.6)	90 (32.9)	94 (34.4)	273 (100)

Demographics

Sixty-two percent (n=169) of participants were male. The mean age of the sample was 54 (median 56) years. Almost half (n=127; 47%) of the participants identified as African-American; 42% (n=114) identified as being White. Almost half of participants indicated they were in good health (n=130; 48%), the other two quarters in excellent (n=74; 27%) or fair health (n=63; 23%). Thirty-nine percent (n=106) reported having completed college or more and 14% (n=38) did not graduate from High School. Sixty-three percent (n=171) reported being unemployed (including retired). One quarter of participants (n=67) reported an annual income of more than $100,000 and just under half (n=120; 44%) reported an annual income of less than $25,000. Almost all individuals reported having health insurance (n= 259; 95%) and more than half of those (n=166; 64%) reported they had public insurance. More than half reported they had enrolled in research in the past (n=158, 58%). See Table 3 for details by study arm.

Table 3:

Demographics by Intervention Assignment

	Standard n=89		Fact Sheet n=90		Video n=94		Overall n=273
Gender	n	%	n	%	n	%	n	%
Female	32	36.0	36	40.0	34	36.2	102	37.4
Male	57	64.0	52	57.8	60	63.8	169	61.9
Other	0	0.0	2	2.2	0	0.0	2	0.7
Race	n	%	n	%	n	%	n	%
White	37	41.5	36	40.0	41	43.6	114	41.7
Black	39	43.8	45	50.0	43	45.7	127	46.5
Asian	2	2.3	2	2.2	3	3.2	7	2.6
Native American Indian or Alaska Native	3	3.4	0	0.0	1	1.1	4	1.5
Other	3	3.3	2	2.2	2	2.2	7	2.7
Don’t Know	2	2.3	4	4.4	0	0.0	6	2.2
Ethnicity	n	%	n	%	n	%	n	%
Hispanic	4	4.5	2	2.2	4	4.3	10	3.6
Non-Hispanic	85	95.5	88	97.7	90	95.7	263	96.3
Education	n	%	n	%	n	%	n	%
Less than High School	11	12.4	13	14.4	14	14.9	38	14
12th Grade/GED	24	27	20	22.2	25	26.6	69	25.3
Associate/Technical Degree/Some College	15	16.9	22	24.4	23	24.4	60	22
College Degree	20	22.5	18	20.0	16	17.2	54	19.8
Some Grad School	2	2.3	1	1.1	2	2.1	5	1.8
Grad School Degree	17	19.1	16	17.8	14	14.9	47	17.2
Employment Status	n	%	n	%	n	%	n	%
Full-time	22	24.7	27	30.0	25	26.6	74	27.1
Part-time	10	11.2	8	8.9	10	10.6	28	10.3
Not employed	57	64.0	55	61.1	59	62.8	171	62.6
Household Income	n	%	n	%	n	%	n	%
Less than $25,000	41	46.1	38	42.2	41	43.6	120	44.0
Between $25,000 and $49,999	13	14.6	12	13.3	18	19.2	43	15,8
Between $50,000 and $74,999	6	6.7	9	10.0	8	8.5	23	8.4
Between $75,000 and $99,999	6	6.7	2	2.2	2	2,1	10	3.7
More than $100,000	19	21.4	26	28.9	22	23.4	67	24.5
Refused	4	4.5	3	3.3	3	3.2	10	3.7
Health Status	n	%	n	%	n	%	n	%
Excellent	23	25.8	26	28.9	25	26.6	74	27.1
Good	48	53.9	43	47.8	39	41.5	130	47.6
Fair	16	18.0	17	18.9	30	31.9	63	23.1
Poor	2	2.3	4	4.4	0	0.0	6	2.2
Type of Health Insurance	n	%	n	%	n	%	n	%
Public	57	64.0	56	62.2	53	56.4	166	60.8
Private	29	32.6	30	33.3	34	36.2	93	34.1
None	3	3.4	4	4.4	7	7.5	14	5.1
Participated in Research in Past	n	%	n	%	n	%	n	%
Yes	53	58.4	50	55.6	56	59.6	158	57.9
No	36	40.5	37	41.1	38	40.4	111	40.7
Don’t know	1	1.1	3	3.3	0	0.0	4	1.5

Primary Outcome: Understanding

The primary outcome (understanding) was evaluated via score comparisons on relevant items between those exposed to the fact sheet and standard consent process and between the video and standard consent process. We compared the adjusted mean scores (adjusted for variation in enrollment across the six collaborating studies) on the overall percent correct score but also analyzed the open-ended questions and close-ended questions separately. Participants who received the fact sheet did not achieve higher overall understanding scores when compared to the standard consent group (p = 0.49), and there was no difference between these two arms for open-ended or close-ended assessments (Table 4).

Table 4:

Understanding – Three Comparisons

Questions	Standard Consent Adjusted Mean(SE)	Fact Sheet Adjusted Mean(SE)	Video Adjusted Mean(SE)	Model-based Mean Difference (95% CI)	P-value
Fact Sheet vs. Standard Consent
Overall	77.7 (1.7)	79.0 (1.6)		1.3 (−2.4, 5.2)	0.49
Open Ended	75.1 (2.1)	76.2 (2.0)		1.2 (−3.5, 5.9)	0.62
Close Ended	79.38 (1.9)	80.8 (1.9)		1.5 (−2.9, 5.8)	0.66
Video v. Standard Consent
Overall	77.7 (1.7)		82.0 (1.6)	4.4 (0.6, 8.2)	0.02
Open Ended	75.1 (2.1)		81.4 (2.0)	6.3 (1.6, 10.9)	0.008
Close Ended	79.3 (1.9)		82.4 (1.8)	3.1 (−1.2, 7.4)	0.16
Video vs. Fact Sheet (exploratory)
Overall		79.0 (1.6)	82.0 (1.6)	3.1 (−.7, 6.8)	0.11
Open Ended		76.2 (2.0)	81.4 (2.0)	5.1 (0.5, 9.7)	0.03
Close Ended		80.8 (1.8)	82.4 (1.8)	1.6 (−2.6, 5.9)	0.45

Participants exposed to the video had statistically significant better understanding scores compared to those exposed to the standard consent form process (between arm difference 4.4%; p=.02). The higher overall understanding score observed in the video arm was mainly due to the significantly higher score participants in that arm had for the open-ended questions (between arm difference 6.3%, 95% CI: 1.6 to 10.9%, p = .008).

We also conducted an exploratory analysis comparing understanding scores for the fact sheet to those of the video and found a trend that those exposed to the video had higher understanding scores (p = .11) when compared to those exposed to the fact sheet. As with video versus standard comparison the higher overall understanding score observed in the video arm was mainly due to the significantly higher score participants in that arm had for the open-ended questions (between arm difference 5.1%, 95% CI: .5 to 9.7%, p = .03). (Table 4).

Secondary Outcome: Satisfaction

Satisfaction with the method of consent was high across all three arms of the study. For overall satisfaction, we found evidence to support that participants were more satisfied with the video compared to the standard form (between arm difference −6.9, 95% CI: −10.8 to −2.1, p<0.01; lower scaled scores represent greater satisfaction). We did not find evidence that participants were more satisfied with the fact sheet compared to the standard form (between arm difference −3.1, 95% CI: −7.1 to 0.8, p=0.12) (Table 5). We also conducted an exploratory comparison of satisfaction between the fact sheet and video and found no difference.

Table 5.

Satisfaction Scores^* - Three Comparisons

Arm Comparison	Adjusted Mean (SE)	Model-based Adjusted Mean Difference (95% CI)	p-value
Fact Sheet vs. Standard Consent	13.8 (1.7) vs. 17.0 (1.7)	−3.1 (−7.1, 0.8)	0.12
Video vs. Standard Consent	10.1 (1.7) vs. 17.0 (1.7)	−6.9 (−10.8, −2.9)	0.00074
Video vs. Fact Sheet	10.1 (1.7) vs. 13.8 (1.7)	−3.7 (−7.6, 0.2)	0.06

^*Range 0–100 with 0 being most satisfied, 100 being least satisfied, and lower score indicating more satisfaction.

Exploratory analyses:

A number of participant attributes were associated with understanding scores. These analyses were completed on the entire sample independent of their study arm. Those who were older (95% CI −.38 to −.09) ), those who were not white (95% CI −9.05 – 1.83), those who had no previous experience with research (95% CI 1.13 to 7.95) and those who agreed that research studies carry more risk than regular health care (95% CI −9.05 to 1.87) had lower understanding scores. Those who reported college or higher education attainment had higher scores (95% CI 4.43 to 11.68). Participants who were older were significantly more satisfied with fact sheet and video compared to the standard form (p = 0.03). No other sociodemographic variable was found to be associated with satisfaction. There were no interaction effects on understanding scores between participant attributes and the intervention to which they were randomized.

Evaluation of Implementation

All 11 clinical research coordinators eligible to participate in the evaluation interview agreed to participate. Study coordinators reported the time required for the research participant to complete our assessment was the greatest burden of working with our study. Three coordinators noted it interrupted the flow of the consent process and two mentioned access to a telephone, room, or person conducting the interview contributed to the burden.

Study coordinators reported that the fact sheet was the least burdensome of the interventions for them to deliver. Coordinators also noted the fact sheet appeared to save time, was easier to understand, and maintained attention. There was a range of positive responses on the usefulness of the video between coordinators including that it was helpful for visual learning, easy to understand, maintained attention, and required less effort/burden for the coordinator. On the other hand, they were concerned that participants were not as attentive to the video as compared to the fact sheet. The study coordinators reported that they believed the fact sheet to be the most likely intervention to improve understanding. Almost all of the coordinators (n=9) recommended that the fact sheet should be used in the future, alone or in combination with the standard form. Seven recommended the use of the video in combination with the standard form.

Discussion

This randomized study of two novel consent interventions across six different clinical trials demonstrated a statistically significant difference in participant understanding based on overall scores among those exposed to the video intervention compared to those exposed to the standard consent. This difference was most notable in the responses to the open ended questions where participants scored more than 6 points better after exposure to the video when compared to the standard. This difference may indicate that the audio-visual presentation of the information was more effective at delivering the information about the study than written textual presentation and resulted in participants being better able to articulate their understanding when asked open-ended questions. In addition, participants exposed to the video (but not the fact sheet) reported higher levels of satisfaction when compared to those exposed to the standard consent.

While others have conducted randomized controlled trials of the effect of concise consent form information on subject understanding, ours is the first to include participants from several different, ongoing studies.^23,34,36 Most recently, Grady and colleagues conducted a cluster randomized non-inferiority trial comparing the use of a standard consent form with a more concise version in a single large study assessing timing of antiretroviral therapy for HIV-1 infected individuals and found no difference in understanding but higher satisfaction with the concise version (i.e. in terms of understanding, the concise form was not inferior to the standard form).³⁶ Hall and colleagues conducted a randomized controlled trial of four different modes of consent (standard consent, study information in FAQ format, professionally produced video, study team produced video) for a web-based behavioral survey among men who have sex with men and found that video based interventions improved understanding.²³ Similar to Hall we found that those exposed to a concise video presentation achieved higher understanding scores compared to standard consent and found no difference in understanding when comparing scores of those exposed to the standard consent group to the scores of those exposed to a concise version of the consent information.^23,36 Grady found no difference in level of understanding when comparing participants who received the concise form to those who received the standard form. ³⁶ Participants in our study did not express more satisfaction with the concise form. Our study participants did, however, express more satisfaction with the video presentation (p-value = .0007). Hall and colleagues did not assess subject satisfaction.²³ In summary, our findings confirm that exposure to a concise video does lead to better understanding scores and that exposure to written concise consent information results in similar levels of understanding as those exposed to standard consent forms. , In another similarity to Grady and colleagues and others, we found that understanding of consent information is independently associated with age^{34,36,40,43,44} and level of education.^{34,40,43,44,45,46,47,48,49,50}

Given that a number of studies have shown that exposure to concise versions of the consent form result in the same level of understanding and in some cases increased satisfaction, whether all consent forms could adopt a concise format (written or audio) ought to be explored. That is, concise documents that include key information about the study could be adopted more widely at no cost to understanding and some gain in satisfaction.

It is interesting to note that study coordinators who facilitated the implementation of the study, when interviewed by us, considered the fact sheet to be the more likely of the two interventions to improve understanding. Indeed, they found that providing key summary points to participants to be a clearer way of sharing the most important information. The discrepancy between what research coordinators think and what actual measurement of participant understanding reveals the importance of formal measurement of understanding.

Recent revisions to U.S. federal regulations (the “Common Rule”) now require that consent forms present key information up front to facilitate study comprehension despite limited evidence to support this regulatory shift.⁵¹ Based on our findings, and that of others, participant understanding is improved when exposed to a concise video and comparable when exposed to either the standard consent form or the fact sheet alone. As we conducted our project before the revision of the Common Rule was in force did not address directly the new regulatory requirement that consent forms include key information at the beginning of the consent form.⁵¹ Whether these findings can be extended to the delivery of key information to participants considering enrollment in research ought to be explored further. A future study could explore, among other things, whether including key information delivered in video form facilitates understanding and/or decision satisfaction.

Our findings have a number of limitations. First, it was not possible, given the nature of the interventions, to blind the collaborating study staff who implemented the consent interventions. Similarly, the consent study staff member who conducted CUE-R interviews over the phone was not masked to study arm due to the need to ask questions using language appropriate to which intervention the participant was exposed to, but questions about satisfaction with the intervention appeared late in the assessment and the interviews were recorded and reviewed for overt bias. Another, more practical limitation of our study is that each video cost approximately $1,200 to produce. Creating a video for every study conducted at a high volume institution would require a large investment of resources but could improve research participant understanding and participant satisfaction. , But $1,200 may be a relatively small investment for trials enrolling large numbers of participants and/or more complex studies could easily be written into research budgets.

E. Conclusion

This study, the largest to assess participant understanding across multiple active trials, and demonstrated a statistically significant difference in understanding and satisfaction among those exposed to the video intervention. We believe the data indicate the approach has merit and should be studied further. . The underlying question is whether this level of difference is enough to encourage investigators, IRBs and/or policy makers to advocate for the production of study specific videos to enhance participant understanding or increase satisfaction. Based on these findings, a future trial powered to make a single comparison between the interview based video against the standard consent process would be valuable to those making decisions about consent procedures. Additional research should evaluate whether a video consent would be preferred by certain subgroups of patient and potentially lead to greater diversity of research participants.