Conducting Research With Special Populations Flashcards
What are special populations?
- vulnerable to coercion or undue influenced
- diminished autonomy - reduced mental capacity (ability to understand / process info) or voluntariness (freedom from the control of others)
- capacity to consent is questionable
What are some examples of special populations?
- children and young people
- specific ethnic, cultural or minority groups (CALD)
- prisoners and other institutionalised individuals
- health and medical populations
- people in dependent relationships
- pregnant women and the human foetus
- people highly dependent on medical care who are unable to give consent
- people with a cognitive impairment, intellectual disability, mental illness
- people involved in illegal activities
- homeless individuals
- people in other countries
Discuss children and young people as a special population:
- Traditional views of children were that they were vulnerable, dependent and immature
- Now we want to focus on their competency and place an emphasis on listening to children’s perspectives (involving the child’s preferences and rights)
- Spans from preconception up to age 25.
Families can include people with
- competing or non-congruent interests
- varied levels of openness and candour
- different decisional capacity and dependence
We need to consider
- rights
- capacity to consent
- possible coercion
- privacy and confidentiality
- compensation
Example: Sol Krugman Hepatitis Experiment
In a school for children with intellectual disability, he deliberately infected children with hepatitis to study the development of a vaccine.
Parental consent were eventually sought (however he didn’t outline risks), but ethics were questionable.
Example: Johnson and Tudor - the Monster Study
Used orphaned children with and without a stutter, and manipulated positive reinforcement vs punishment.
Some children in the punishment condition developed a stutter.
They didn’t undo this distress they caused.
What is informed consent?
- From young children on, they have a developing ability to understand information and should be involved in the consent program as much as possible.
- No set age - but typically children under 16 do not give legal consent –> From age 18, must give informed consent.
- But as soon as children are able to process information, we should be providing age appropriate information so they can give verbal assent.
- need to also manage multiple relationships (parent-researcher and child-researcher)
What are some other forms of consent?
Opt out consent: e.g. school agreeing to a class being part of research and the parents have to opt out.
Standing consent: when parents give consent for children’s to be in studies throughout the year.
Rights and respect
Balance
- risk
- harm
- benefit
Need to consider if the research is likely to be physically or emotionally dangerous for a child.
And respect the developmental level of child when engaging them.
Privacy and confidentiality
- discuss confidentiality and its limits with the parties upfront
- explain (in age appropriate language) what confidentiality means and the exceptions
- ensure the research team have plans for how breaching confidentiality will be handled
- common reasons for breaching confidentiality include mandatory reporting and unsuspected risk of the child
- reporting is required when a researcher suspects on reasonable grounds that a child is at risk of significant harm
Compensation
consider whether payment is appropriate and how much
for instance, small gifts may be more appropriate for children
How are people recruited?
- schools (may need approval from the school system)
- childcare
- sporting and extracurricular
- hospitals and clinics
- parents / guardians vs children
- government / health records
- online and social media
- newspapers and magazines
- newsletters
What are some practical considerations of assessing chidren?
- access
- who is recruiting –> power and perceived coercion - if you’re looking for a kid with a specific condition, is it appropriate for the healthcare provider to be advertising that
- consent process
- participation time (during the day or straight after school - attention span), location and duration: i.e. testing in schools and hospital settings requires separate ethics approval
- age appropriate assessment: measures (normed for that age), language, child vs other report, mother vs father vs teacher report
Clinical populations
Mental health research can help to answer many questions including:
- risk factors and causes of mental health problems
- prevention of mental health problems
- attitudes, opinions and preferences
- distinguishing typical vs atypical emotional, cognitive and behavioural functioning
- understanding the impact of mental health problems (personal, societal, economic)
- changes over time
- treatment and service delivery
Have to consider that mental health conditions are heterogenous and affect people in different ways (brief - chronic, mild - severe)
Can affect thoughts, feelings, experiences, and relationships of individuals and loved ones.
Dan Markingson - treatment of clinical populations gone wrong
- he was involuntarily admitted to inpatient care
- dr olsen his treating psychiatrist said that if he would be part of his study, he wouldn’t have to go into admission
- they kept him in the study despite worsening symptoms and he ended up killing himself
Dan didn’t have the capacity to consent - he was experiencing psychosis, was coerced and there was a conflict of interest being that his lead investigator was his lead psychiatrist.
What are some ethical considerations in place?
- the right to be included nad have opinion heard
- beneficence and non-maleficence (risk vs benefits)
- respect (regard participants with dignity and appreciation of their autonomy / privacy / capacities - use of language - protecting rights)
- consent: consider whether capacity may be impaired (nature of the condition, medication / treatment, distress / discomfort, complexity f the project, fluctuations in condition, having the right to withdraw consent at any time
Other ethical considerations
Privacy and confidentiality
- personal implications of breaching confidentiality (stigma, employment)
- publishing of personal information
Conflict of interest
- clinicians involved in research
- client welfare vs desire to complete the research
Compensation
- should not be of a value that would unduly influence participation
What are some practical considerations in selecting the correct sample to answer the research question?
Is a clinical sample necessary to answer the question?
- can it be answered using nonclinical samples
- can it be answered using clinical analogue samples (samples that have high levels of a symptom but no diagnosis)
- experimental manipulation ( can you do a mood induction instead?)
Do you need a control sample? Who is the best comparison?
These may be needed to know how
- clinical samples differ to non-clinical samples
- individuals with one disorder differ from those with other types of disorders
- if an intervention results in different outcomes
How do you recruit clinical samples?
Feasibility of access:
- some conditions are low prevalence (vulnerable populations may not be eager to disclose their mental health status)
- consider the gate-keepers to accessing potential participants
Recruitment sources may include:
- clinics / hospitals
- health practitioners
- support groups
- online
- community advertising
- disorder related associations and foundations
Important to plan how to manage distress and risk
- immediate distress or risk
- referral options
What are some methods for studying clinical populations?
- direct vs indirect
–> direct = ask individual people
–> indirect = ask people who treat people with that disorder, or friends and family of those individuals - identifiable vs anonymous: sometimes you do need to identify people for compensation
- methods: interviews, focus groups, questionnaires, observation, document reviews
- manipulation vs no manipulation
- self-report vs other-report vs interview vs observed
- clinical trials
Clinical trial comparison conditions
- placebo control and waitlist control
–> yields some effects (30% in pharmacological trials)
–> allows comparison to active ingredient
–> concerns about withholding effective treatment
–> may bias recruitment with more severe cases not enrolled (reduced generalisability)
–> smaller sample sizes (cumulative risk is reduced) - active control
–> is the experimental treatment better than other treatments
–> controls for non-specific elements - washout studies
–> gradually reduce existing medication to study when drug / intervention free
–. differentiate symptoms related to a medication and/or withdrawal from disorder - must minimize risks and ensure rapid response to symptoms if necessary
