Ethics/Stats Flashcards
4 principles of consent
VICS
voluntary
informed
capable
specific
4 factors negligence
Duty
Breach
Harm
Causality
CanMEds (7)
MCCASPL
Medical expert
Communicator
Collaborator
Advocate
Scholar
Professional
Leader
4 principles of ethics
Autonomy
Non-maleficence
Beneficience
Justice
4 things to discuss during consent
nature of procedure
common and severe risks
alternative treatments
consequences of refusing
capacity evaluation (4)
nova scotia criteria
understanding
appreciation
reasoning
expression of choice
2 to proceed without consent
1) imminent treat to life or health
necessary to proceed immediately
and patient / sustitute is not not available
2) psych imminent risk to themselves / others
3) mandatory reporting for public health
4) vital statistics
exceptions to confidentiality (3)
- statutory reporting (child abuse, fitness to drive or fly, communicable disease, coroner)
- imminent danger (clear risk to identifiable person or group, risk of serious bodily harm and danger is imminent)
- court order
What contributes to a type 1 error? (2)
Bias
Confounding variables
What contributes to a type 2 error ? (1)
Low power
What are ways to decrease confounding variables ? (4)
Multivariate analysis
Stratification
Randomization
Matching
Study shows a relative risk of 2.1. How do you interpret this ?
i. Exposed group is 2.1 times more likely to have breast cancer than the unexposed group
200 person study. 50 have the illness. After 10 years, 10 patients develop illness.
- What is the cumulative incidence?
- What is the prevalence on 10 years?
- 2 ways to associate risk factors with the disease?
- Cumulative incidence = # new cases/ # at risk over x years
10/(200-50) = 1/15 per ten years - a. Prevalence = number of sick people / people
b. 60/200 = 30% - Relative risk & odds ratio
3 advantages of a meta-analysis
- Increased power
- Can do subgroup analysis that initial studies were underpowered to do
- Can resolve uncertainty when studies disagree
What does I squared mean? I=0
No inconstency in the studies
What are two factors that increase heterogeneity in meta-analysis?
- Clinical diversity
- Methodological diversity
5 levels of evidence with 1 example of each
- High quality RCT
- Lesser quality RCT, prospective cohort
- Retrospective study
- Case-series
- Case report
Phases of studies with 3 characteristics of a phase 1 treatment ?
a. Phase 1 = safety
i. Small cohort
ii. Low doses
iii. New medication
b. Phase 2 = efficacy
c. Phase 3 = large scale (RCT)
d. Phase 4 = post-rollout monitoring
Define type 2 (beta) error
Risk of accepting a null hypothesis that is false
Define type 1 (alpha) error
Rejecting a null hypothesis that is true
6 comments on RCT and surgery
- Often RCT will happen after the adoption of a new surgical technique
- Difficulty to recruit (rare, urgent cases)
- Surgeon is not blinded (patient and outcomes recorder are)
- Learning curve for new surgery
- Some centers specialized for one surgery, some surgeons do both but have preference, bias is introducted
- Ethical quetsion of placebo surgery
d. En chirurgie plastique, MEILLEUR type d’étude pour déterminer laquelle parmi 2 intervention est la plus bénéfique économiquement
i. Prospective Cost-benefit analysis
- Définir “intention to treat analysis”
Data analysed per assigned treatment group without accounting for therapeutic changes over the course of the study
Two exams in frostbite
Technetium 99 triple bone scan
Angiography
What does it mean if you have a power of 0.9
90% chance you will reject a null hypothesis that is false
2 acceptable and unacceptable ways to randomize
> 200 participants, toss coin
block randomization
computer random number generator
preference of patient, choice of physician
