HaDPop Flashcards
To cover all Health and Disease in Populations content you are required to know for ESA 1 (may not be everything) I take no responsibility for any of the flashcards featured here...mistakes/shit happens.
Define fecundity
The physical ability to reproduce
Define fertility
The realisation of the fecundity (physical ability to reproduce) as actual births
What is the crude birth rate defined as?
The number of live births per thousand people in a population
What is the general fertility rate defined as?
The number of live births per 1000 of the populations women who are ages 15-44
What do we mean when we say the total period fertility rate?
All of the age specific fertility rates together
What is the crude death rate defined as?
The amounts of deaths per 1000 people in a population
What is the age-specific death rate defined as?
The amount of deaths per 1000 people in an age group per year
What does SMR stand for?
Standardised mortality ratio
What is the point of using an SMR?
It adjusts for age and sex distribution which removes confounding
How does an SMR work? i.e. think of how it is calculated
It works by comparing a population being studied to a reference population, any values of 100 show no difference
Observed pop./Expected pop. (x100)
Define Census
The simultaneous recording of demographic data at a particular time pertaining to all individuals who live in a particular territory
How is an incidence rate worked out? (NOT HOW IS AN INCIDENCE RATE RATIO WORKED OUT)
Incidence rates are worked out by how many new events of something happen in a population over time. Therefore:
Incidence rate = New Events ÷ (Population x Time)
How is an incidence rate ratio worked out?
One incidence rate divided by another
Define incidence rate
The number of new cases of a disease in a population over a period of time
How would you work out the two confidence limits?
Error factor x IR
IR ÷ Error Factor
Can you ever accept the null hypothesis?
NO!
You can only ever reject or not reject the null hypothesis.
Define Prevalence
The number of existing cases of a disease in a population
What is a confounding factor?
A confounding factor is any factor that links the exposure with the outcome but is not on the causative pathway
Allocation bias is a form of which type of bias?
Selection bias
Recall bias ia form of which type of bias?
Information bias
What type of plot can be used to identify publication bias? What does it do?
Funnel plot which shows whether there are unpublished results omitted
What is the healthy worker effect? What type of bias is it?
The healthy worker effect is a form of selection bias that occurs when populations who are in work are being to compared to just the general population
What is the main difference between internal and external cohort studies?
External cohort studies use a reference population whereas an internal cohort study is done with different sub groups
What is a cohort study?
Where you look at people who have been exposed and people that have been unexposed then you wait to see who has developed the disease
Why is a cohort study useful?
- It can be used for looking into a lot of different outcomes of an exposure
- It is useful for rare exposures but not rare diseases
Why is a cohort study not that useful?
- Expensive
- Time consuming
What is a retrospective cohort study?
A cohort study where disease free individuals are recruited and their exposure status is recorded from historical documentation and they are followed up
What is a prospective cohort study?
A cohort study where disease free individuals are recruited, their exposure status determined after they have been recruited and they are then followed up
What is a case control study?
A study looking into people who are ill and a group of healthy people working backwards to determine an exposure
Why is a case control study useful?
- Cheap
- Quick
- Useful for rare diseases but not rare exposures
Why is a case control study not that useful?
- Prone to recall bias
- Prone to selection bias
How is an odds ratio worked out?
(Exposed cases ÷ Unexposed cases) ÷ (Exposed controls ÷ unexposed controls)
How can you reduce the error factor of an odds ratio?
Increasing controls
What are the three main steps to a randomised control trial?
- Identify patients
- Randomly allocate participants
- Follow up participants identically
Why is double blinding often used?
- Removes the placebo effect
- Removes selection bias
What is the intention to treat model?
The idea that even if people don’t keep using the treatment correctly/drop out they should still be included
List the Bradford Hill Criteria and briefly describe what it means?
9 points (Eurgh…I want to die)
- Strength of association is how much difference is observed i.e. a high IRR
- Specificity of association is how specific the disease/exposure is i.e. cancer is a very broad group
- Temporal sequence is asking whether the disease precedes the exposure i.e. a nested case control study
- Dose response is how much you give compare to how much difference it makes
- Reversibility can you reverse the effect you are having
- Biological plausibility does it have a mechanism that makes sense
- Coherence of theory confirms current thinking about things
- Analogy similar disease have similar outcomes
- Consistency means that the results can be replicated again by other people
How would you minimise losses and increase compliance?
- Minimise losses by honesty and make clear, accessible follow up
- Compliance by making it easy to use and allowing patients to ask questions
What 3 ways could you ensure a randomised control trial was ethical?
- Clinical equipose (no idea which is better)
- Scientifically robust (for the good of the science)
- Use volunteers
What 3 features should a study have to be included in a systematic review?
- Transparent
- Replicable
- Explicit
How is the data from a meta-analysis displayed? What is shown and how?
Shown via forest plots:
- Odds ratios as squares where weighting given is size of square
- Pooled estimates are diamonds with 95% confidence limits
What does the fixed effects model assume?
It assumes that studies are homogenous and that any data variation comes from within the study
What does the random effects model assume?
Assumes that the studies are heterogenous and that variation comes from between the studies as well
What will a well balanced study show in a funnel plot?
A well balanced study will show a funnel shape
What are the advantages of using a systematic review?
- Reduce bias and exclude shit studies
- Give an overall figure for multiple studies
- Used in evidence based guidelines
