MHS Flashcards
What are the three types of Systematic Error or Bias in clinical trials? And what is it minimised by?
Selection bias (selection, loss to follow up)
Information bias e.g. recall bias
Reporting bias e.g. publication bias
Minimised by study design, blinding and randomisation
What is a confounding error and what is it minimised by?
Known or unknown third factors associated with relationship under investigation
Minimsed by randomisation
What two types of random error are there?
Type 1 (alpha) or Type 2 (beta)
What is a type 1/alpha error? What is it controlled by?
False +ves
Level of significance
What is a type 2/beta error? What is it controlled by?
False -ves
Large sample sizes
What are the 3 main models of stress?
Stimulus, Response and Transactional
What are the 3 phases of stress?
Phase 1 - Alarm
Phase 2 - Resistance
Phase 3 - Exhaustion
What are the 2 sympathetic pathways (stress lectures)
SAM - sympathoadrenomedullary axis
HPA - hypothalamic pituitary adrenal axis
Where is serotonin made?
Raphe nucleus
Where is noradrenaline made?
locus coerulus
What are the 5 main categories of stressors?
Life events, internal stressors, physical, lifestyle and environmental
What is the choice principle?
That we should seek out events that are pleasurable and avoid those that are harmful or threatening
What is the yerkeses- dodson principle
Performance increases with mental and physiological arousal but only up to a point
Too high performance decreases and there is tension and anxiety
What is eustress?
Positive cognitive responses to stress e.g. athletic competition
What was the study by Seligman and Maier (1967) looking at?
Learned helplessness
What was the crowded train study?
Lundgberg 1976 - those who had been on the train from beginning were less stressed - were in control e.g. of finding and choosing where to seat
What study in 1976 also showed that giving people a sense of personal control increased health?
Langer and Rodin
What are the three types of decisions that have to be made?
Approach-approach
Avoidance - Avoidance
Approach - avoidance
What is the purpose of randomisation in RCT’s
to ensure potential confounders (known and unknown) are balanced between intervention and control group
What are the top 3 observational studies (in order)
Cohort studies
Case control studies
Self controlled case series
What is an experimental study?
Studies in which exposure is determined as part of the study design
What is an observational study?
Studies in which exposure is not determined as part of the study design. So already existing exposures and their consequences are studied
What is a null hypothesis?
The hypothesis being tested imagining there are no links between either factor being investigated
What is an alternative hypothesis?
What is believed to be true if the null hypothesis is false
What is the p value?
The probability of seeing a difference of size by chance
Smaller p value is more likely to be real result
What does it mean if p=0.05
That 5% or 1/20 chance of seeing a difference of the size observed as a result of chance
What p value is said to be a statistically significant difference?
p<0.05
0.5 or above the null hypothesis is rejected and alternative accepted
What is a case control study?
A study where subjects are chosen on the basis of whether or not they have the disease or health related state - retrospective
Identification of exposure in the past
What is a cohort study?
A study where the subjects are chosen on the basis of whether they are exposed or not - prospective
Recruitment of cases as they happen
What does the WHO define health as?
A state of complete physical, mental and social well-being and not merely the absence of disease of infirmity
What is health inequality?
difference in health status of individuals and groups
What is health inequity?
unfair differences in health status that are potentially avoidable
What does the marmot review describe?
proportionate universalism - how disadvantage starts before birth and accumulates throughout life
What are epidemic models?
Sets of equations that describe the rate of change of numbers infected
How do we “deal” with chance associations?
By statistics and defining a probability level that we find acceptable - the alpha level
What is the accepted alpha level?
1/20 i.e p<0.05
How do you deal with confounding variables in a study?
Study design: to restrict the study to one stratum of the confounder or match for potential confounders
Analysis: stratified analysis after having collected data on confounders
What are residual confounding factors?
When all known confounders have been accounted for but there are other unknown confounders that are exaggerating or marking a true relationship
What is the best way to deal with both known and unknown confounders?
randomisation
What is incidence?
New events within a time dimension
What is prevalence?
existing cases within a time
What is point prevalence?
existing cases a specific time
What is period prevalence?
existing cases over a time course
What is cumulative incidence?
new cases over a time period but each with a different length of follow up
What is standardisation?
Set of techniques used to remove as much as possible the effects of differences in confounders when comparing two or more populations
What is indirect standardisation?
When from the reference population to find stratum specific rates and you apply these to the study population the generate the expect value –> then compare expected number of deaths with observed
What is direct standardisation?
When from the study population to derive stratum specific rates and apply these to the reference population and derive the expected rate
What is sensitivity?
The probability that a test on a patient with the condition will give a positive result = True positive
= true positive/total with condition
What is specificity?
The probability that a test on a patient without the condition will give a negative result = true negative
= true negative/ total without the condition
What is the positive predictive value?
The proportion of patients with positive test results that actually have the disease
= true positive/ total positive
What is the negative predictive value?
The proportion of patients with negative test results that don’t have the disease
true negative/total negative
How do you calculate attributable risk?
Difference between absolute risk for control and treatment groups
How do you calculate absolute risk?
no of events/ total no of people
How do you calculate relative risk?
Absolute risk of treatment group / absolute risk of control group
What is a de facto census?
People are enummerated according to where they stay on census night
What is a de juro census?
People are ennumerated according to where they usually live
