The Electronic Health Record in Research: Ethical Implications

1
The Electronic Health Record in Research
Ethical Implications

• Nancy M. P. King, JD
• Department of Social Sciences Health Policy
• Program in Bioethics, Health, Society
• Wake Forest University School of Medicine

2
Why Are There Issues?

• Clinically derived data
• Increasing volume
• Increasing availability
• HIPAA
• Designed to promote access to EHRs
• The best of intentions
• Ability to manipulate merge data sets
• Desire to address clinical public health needs

3
What Are The Issues?

• Informed consent--or not?
• Confidentiality ongoing data linkage
• Reporting results incidental findings
• Dissemination, Publication, Translation
• Is genetic information different?
• Group rights interests
• The Citizenship Model

4
Ethics Science in EHR Research

• Design
• Value of and reason for the research question
• Methodology and validity of design
• Data-sharing plans
• Conduct of Research
• Explaining research purpose and data use
• Community advisory board?
• Consent for use of clinical genetic information
• Privacy confidentiality
• Dissemination of Results
• What is being measured?
• How do the results help address the problem
  identified?
• Investigator responsibility group interests

5
Informed Consent

• Is consent necessary? Is it desirable?
• Not Human Subjects Research
• Exempt Research
• Consent Waiver
• Informed consent
• scope limits
• Informing without Consent

6
Informed Consent Exceptions

• Not Human Subjects Research
• secondary uses of coded private information
  when identifying information is not provided to
  researcher
• but linking new records risks reidentification
• Exempt Research
• already existing data, recorded by researcher
  without identifiers
• but linking new records risks reidentification
• Consent Waiver
• minimal risk
• consent makes research impracticable
• subjects rights and interests unaffected

7
Informing Limiting Consent

• Informed consent to research use of clinically
  derived data
• is prospective consent feasible?
• can scope of use be anticipated?
• how broad can meaningful consent be?
• is limiting scope of EHR use feasible?
• Informing without consent
• patients deserve to know they may be research
  subjects
• public deserves to know value of EHR research
• basis for citizenship model of research
  participation

8
Protecting Confidentiality

• Privacy vs. Confidentiality
• Expectations treatment vs. research
• ongoing data linkage
• how to safeguard patient-subjects sensitive
  information?
• how to enable continuing data collection?
• monitoring the balance

9
(No Transcript)
10
Recontact in EHR Research

• Why (Re)contact?
• to expand or alter subjects participation -- and
  seek (new or amended) consent
• to gather additional data in direct encounter
• to expand subject pool by reaching relatives of
  subjects
• to return individual results or incidental
  findings
• to disseminate aggregate results -- and inform
  subjects how they have assisted in research

11
Critical Issue Returning Results

• Recontact if medically significant information is
  found?
• what is significant?
• nature, magnitude, likelihood, timeliness
• confidentiality tradeoffs
• how long is contact information kept?
• whose duty is this?
• who has the knowledge to determine medical
  significance?
• contact information and knowledge may be widely
  separated
• information and oversight infrastructure could be
  extensive (and expensive)

12
(No Transcript)
13
Critical Issue Interpretation Dissemination

• Group interests and group harms
• IRBs usually do not consider, but their
  significance is acknowledged (and growing)
• Dissemination of research results
• How should results be described?
• race/ethnicity categories are inadequate or
  pernicious
• personalized medicine is a long way off
• Duties of investigators, sponsors, institutions?
• communicating with public about meaning (and
  limits) of information -- especially genetic
  information

14
Is Genetic Information Different?

• Genetic information is poorly understood by most
  patients and many physicians
• Genetic information is high-profile -- though
  individual and family history information can be
  just as revealing
• Genetic predispositions, risks, and links to
  environment can be misused
• Misuse is infrequent misunderstanding and fear
  are real
• Group characterizations can stigmatize and
  discriminate

15
Should Groups Have a Research Voice?

• Group rights interests
• Individuals choices about scope of research
  using information from them may reflect concerns
  about groups even when individuals are not
  identifiable
• individuals with family histories of stigmatized
  conditions
• African-Americans, Native Americans, Ashkenazi
  Jews, other First Peoples ethnic groups
• Reporting and dissemination can stigmatize and
  foster discrimination

16
Information and Identity

• Havasupai Tribe v. Arizona State University
• Private information and ownership
• just stuff (U.S. perspective)
• I saw my mother (Maori health researcher
  studying anonymized medical records)
• Understanding genotype phenotype in social
  context?
• nature nurture
• nature culture

17
A Citizenship Model for EHR Research

• protection model
• analogy to patients
• focus on rights limiting power
• limited consent
• utility model
• analogy to consumers
• focus on data security public health
• blanket consent
• citizenship model
• agency, awareness, democratic engagement
• focus on increasing science literacy, database
  transparency
• broad consent with opt-out mechanism
• --Vilhjalmur Arnason, University of Iceland

18
Conclusions Recommendations

• Enormous utility of research using EHR data is
  not sufficient justification
• Informing involving public can increase science
  literacy
• Awareness of data sensitivity, for both
  individuals groups, can increase both
  researcher understanding public trust
• EHR has capacity to tag tailor research uses to
  benefit both public health public interest