How to Improve Research Participants’ Understanding

James Flory, BA

Bioethics FellowDepartment of Clinical Bioethics

Warren G. Magnuson Clinical CenterNational Institutes of Health

Background

• Understanding is a goal of the informed consent process

• Empirical research can provide information about how well research participants understand

• And how research participants can understand better

Disclaimer

The views expressed in this presentation do not represent the views of the NIH, the United

States DHHS, or any other government agency or official.

Central question

What can empirical research tell us about how to improve research participants understanding?

Disclosure

Typical disclosure requirements

• Purpose of research

• Risks

• Benefits

• Alternatives

• Voluntariness and right to withdraw

• Confidentiality of records

• Compensation for injuries

• Who to contact with questions

How is disclosure usually done?

• A written version of the required information is almost always used: the consent form

• Usually there is also discussion

• In spite of emphasis on consent form, it is only one element in an interpersonal process of communication

Understanding

What is understanding?

• Understanding is processing of disclosed information

• Understanding is hard to measure

• Understanding is not the same as memory

What should research participants understand?

• Purpose of the research

• Research methods (randomization, placebo)

• Risks

• Benefits

• Requirements—e.g. number of clinic visits

• Right to withdraw

• Alternatives

Empirical question

• How well do research participants understand these key areas?

Purpose

Study Participants Understanding

Lynoe

1991

Swedish women in gynec. study

98% research study

Joffe

2001

US oncology studies

30% drugs unproven

Pitisuttithum

1997

Thai HIV vaccine study

80% vaccine might not work

Ciriscione

2003

US rheumatoid arthritis drug study

100% research experiment

Randomization

Study Participants Understanding

Hietanen

2000

Finnish women breast cancer

23% randomization

Harrison

1995

US HIV vaccine 21% everyone get vaccine

Howard

1981

US men cardiac

Beta blocker

42% randomization

Pace

2003

Thai HIV

IL-2 treatment

42% randomization

Risks

Study Participants Understanding

Bergler

1980

U.S. Hypertension

28% not recall 2 risks 2 hours after consent

Miller

1994

Analgesia 52% not recall even one risk 60 days later

Dougherty

2000

US oncology 100% recall at least 1 side effect

Leach

1999

Gambian mothers vaccine

56% recall 1 or more risks

Right to Withdraw

Study and Year

Country Subjects Knew they could

Withdraw

Pitisuttithum

1997

Thailand 33 HIV vaccine

88%

Karim

1998

South Africa

56 perinatal HIV prevent

93%(freedom to quit)

Lynoe

2000

Bang-

ladesh

105 iron for pregnancy

48%

Conclusions on understanding

• Understanding varies a lot from setting to setting

• Understanding of randomization is particularly poor

• Can understanding be improved?

Improving understanding

Improving understanding

• Comprehensive review identifies 42 trials of ways to improve understanding

• Four main approaches:– Video or computer-based disclosure– Shorter, more readable forms– More time spent on discussion– Using short tests to confirm understanding and

guide discussion

Video or Computer-Based Presentation

• Only 3 out of 12 trials showed an improvement in understanding due to video or computerized presentation (P<.05)

• Although videos and computers may improve understanding sometimes, they usually have not worked

Recommendations: Multimedia

• Multimedia can be expensive, and high expense does not guarantee success

• Investigators who invest in multimedia should consider research on its effectiveness

Shorter, More Readable Consent Forms

• 6 out of 15 trials showed an improvement in understanding (P<.05)

• But 5 of the 6 trials that showed improvement were simulated consent processes

• Improved consent forms may be no better than multimedia

How were consent forms improved?

• Lower reading levels

• Better formatting

• Graphics

• Shorter length

Recommendations: Consent Forms

• Better consent forms are not a magic bullet

• The one realistic trial that showed effectiveness used a form with less boilerplate

• Removing unnecessary material may be effective

Extended Discussion

• 3 out of 5 trials showed an improvement in understanding (P<.001)

• The other 2 trials showed trends toward improvement (P=.054, P=.08)

Recommendations: Extended Discussion

• Limited, but encouraging, evidence

• One effective implementation was a 30 minute phone conversation with a trained research nurse

• Stronger case can be made for extended discussion than for other interventions

Using a Test to Confirm Understanding

• 5 out of 5 studies showed an improvement in understanding (P<.05)

• But these studies may have mistaken rote memorization for understanding

Recommendations: Test to Confirm Understanding

• Research that avoids flawed methodology needed

• This approach often proposed but not yet convincingly evaluated

Effect of Education

• In most studies, even effective interventions did not make as large a difference as education

• In one study, the gap between the most and least educated participants was 51 points out of 100

Conclusions on Improving Understanding

• Improving understanding is difficult

• The most effective kind of intervention appears to be extended discussion

• Less educated research participants may need help the most

• Further research important

Practical Recommendations

• Recognize limitations of all approaches

• One-on-one interactions may be better than forms or multimedia

• Less educated research participants may need help most