1 IRB review and assessment of risks / benefits Bernard Lo, M.D. [email protected] August 9 and 12, 2010.

1

IRB review andassessment of risks / benefits

Bernard Lo, M.D.

[email protected]

August 9 and 12, 2010

2

RCR course

CHR course

Book in bookstore

Epi201 website Updates

Room changes

Starting time? 9 AM?

3

RCR course

Final project

1. If primary data collection, ethics part of your grant

2. If secondary data collection, ethics part of your grant under which data collected

3. Take exam with cases that pose ethical dilemmas

http://ctsi.ucsf.edu

5

Overview of today

Why do we have IRBs?

What are the federal regulations for

human subjects research?

What risks are acceptable in research? Case examples

6

Nazi “experiments”

1. Cause intentional and lethal harm

2. No consent

3. Use vulnerable subjects who were

coerced into participating

7

Tuskegee study

1932 Study started

1936 Journal told that local MDs asked

not to treat subjects

1940 Subjects not treated in military

1947 USPHS Rapid Treatment Centers

8

Tuskegee study

1968 Whistleblower Peter Buxtun

1969 CDC, local chapters of AMA and

NMA reaffirm support

1970 News coverage

9

Tuskegee study

1974 DHEW issues regulations on

funded research

1974 Tuskegee Benefit Program

10

Fundamental ethical tension in research

Primary goal is generalizable

knowledge, benefit to society

Participants experience risks but

benefit to others

11

Ethical violations in Tuskegee

1. Serious harm to participants Deliberately withhold standard treatment

2. Consent not informed Deception during consent process

3. Unfair selection of participants Take advantage of vulnerable population

12

Regulations respond to Tuskegee

1. Beneficence Risks must be acceptable in light of

benefits Risks must be minimized IRB must approve study Psychosocial risks?

13

Regulations respond to Tuskegee

2. Respect for persons Informed and voluntary consent Not capable of consent (children, adults

who lack decision-making capacity) Impracticable to obtain consent

14

Regulations respond to Tuskegee

3. Justice Equitable selection of subjects Protections for vulnerable subjects Access to benefits of research?

Federal regulations

1. Risks / benefits

2. Informed and voluntary consent

3. IRB review and approval

15

Topics to cover

What risks must we consider? Biomedical interventions Secondary analysis of existing data Psychosocial risks

How to decrease risks

16

HIV vaccine trial

Vaccine to induce cellular immunity to

HIV

Subjects at high risk for HIV Injection drug users Multiple sexual partners

• Commercial sex workers

18

Biomedical risks of study

Injection-related adverse effects

19

Psychosocial risks of study

Behavioral disinhibition May increase high-risk behaviors

Stigma and discrimination False +HIV test If confidentiality breached

20

Measures to reduce risk

At every visit, risk reduction

counseling, condoms Reduces power

Monitoring of high-risk behaviors No difference in placebo and vaccine

groups

Cards, letters to explain false + HIV test21

Outcomes of trial

No reduction in HIV incidence Higher incidence in men with Ad5

antibodies, uncircumcised

Research intervention may have

serious unexpected adverse effect

22

Birmingham VA research

1.3 million MDs SSN

Not encrypted, not password protected

Employee not authorized to have data

Risk in secondary data analysis

23

Letter from VA

“We at the VA take information security

and privacy very seriously. We

apologize for any inconvenience or

concern this situation may cause, but

we believe it is important for you to be

fully informed of any potential risk to

you”24

How to protect confidentiality

Train staff

Use coded or de-identified data

whenever possible

25

26

Study 1: How to minimize risks?

Data security Locked paper files Password protection No identified data on laptops, removable

devices Encryption

Certificate of confidentiality

Study on use of marijuana in cancer patients

Determine prevalence of use Compare pain relief and chemotherapy-

related nausea in users and non-users

One study site MD Anderson in Texas

27

Question for audience

What are risks of study?

How to minimize risks?

28

29

If confidentiality breached

Legal risk: illegal activities, prison

sentence

Economic harm: loss of employment

Certificates of confidentiality

May withhold names and identifiers in

case of subpoena or court order

Issued by NIH Need not be NIH-funded study

30

Certificates of confidentiality

May withhold names and identifiers in

case of subpoena or court order Not well tested in courts

Issued by NIH Need not be NIH-funded study

31

How to decrease risk

Monitor for adverse events

Respond to serious adverse event

32

33

New HIV vaccine trial

DNA plasmids express gag, pol, nef

Adenovirus vector booster that

encodes for gag, pol, env

Differ from earlier vaccine More balanced CD4 and CD8 response More immunity to HIV rather than Ad

34

Question for audience

Do you regard benefit / risk balance as

acceptable? Yes No Unsure

35

How to reduce risks of study?

Exclude persons at increased risk Exclude Ad5 +, uncircumcised

Monitor adverse effects carefully Continuous monitoring of incident

infections

36

How to reduce risks of study?

Stop trial if significantly increased

infections in vaccine group Data and Safety Monitoring Committee

• Independent of sponsor and investigators

Statistical stopping rules

37

Questions regarding new trial

How to combine risks and benefits into

overall assessment? Primary efficacy endpoint = viral load in

new infections Primary safety endpoint = new infections

38

Federal regulations

Risks must be reasonable compared to

potential knowledge gained

Risks must be minimized (consistent

with valid research design)

Framework for analyzing study

Underlying condition/prognosis is

serious Foregoing trial has costs

Potential benefit of study is high Likelihood and magnitude of benefit

requires in-depth scientific review

39

Framework for analyzing study

Risk of study intervention is high DSMB decided that any statistically

significant increase in HIV incidence would terminate the study

• 8-2 split in first 10 incident infections

40

Framework for analyzing study

Importance of informed consent Explicitly discuss that may increase

infection Test comprehension in participants

41

Take home message

Risks can be serious, even in

secondary analysis and questionnaire

research

Assessment of benefits and risks made

by IRB independent of investigators

42