Population Science Flashcards
Population health is affected by these factors (3)
Demographic shape
Economic composition
Behavioural and lifestyle factors
Burden of disease, disability and ageing depends on these (3)
Population size
Population shape
Age-sex specific rates
What is the purpose of statistics?
To generalise to, I.e. Infer about, the population
Sample is used and must be as representative of the population as possible though samples from the same population can vary
What does sample averages mean?
Means of the sample and can vary within a sampling distribution
What is the confidence interval?
- an estimate of the precision of the observed values sample
95% confidence interval is the range that is likely to contain the mean of the population values 95% of the time
When is the 95% confidence interval wider?
- The greater the variation in the population values
2. The smaller the size of the sample used to calculate it
What is selection bias?
When your sample is not representative of the general population (error in generalisability)
Errors in comparability:
- groups being compared are not from the same population
- groups being studied is not representative of the sampling frame from which it was drawn (due to slow response rate, drop outs (attrition rate)
Examples of information bias (4)
- recall error (esp. in case control studies)
- observer/interviewer error (all studies)
- measurement error (all studies)
- mis-classification (all studies)
What are confounding factors?
Factors that can cause or prevent the outcome interest. They are not associated with the factor under investigation.
Direct standardisation = directly standardised rates
Indirect standardisation = standardised mortality ratios
What is a cross sectional survey?
Data are collected on the whole study population at a single point in time to examine the relationship between disease and other variables of interest
Can be analytical or descriptive
What is descriptive epidemiology?
Involves describing the population studied
Done using many cross-sectional surveys
Analysed to give prevalence
= (no of cases in defined population at one point in time) / (no persons in defined population at the same point in time)
What is analytical epidemiology?
Involves comparison with another group (choosing the best comparison for a specific population)
Can be cohort or case-control
What is a prospective study?
Carried out from the present time into the future
- can collect specific exposure data
- involves long follow up, inefficient and loss of follow-up
What is a retrospective study?
Carried out at the present time and look into the past
- limited control over the data collected
- immediate availability
How can cohort studies be analysed?
- risk ratio
- incidence rate ratio
- odds ratio
What can you analyse case-control studies with?
Odds ratio –> because case-control is a retrospective study (know he outcome already)
What is a cohort study?
- can be prospective or retrospective
- exposed vs unexposed
- followed for a number of time
- incidence of exposed vs incidence of unexposed
Good for examining rare exposures, can study many outcomes simultaneously.
What is risk ratio?
Ratio of prevalence proportions of outcome in groups defined by levels of exposure at a particular time (e.g. Per 1,000 persons)
What is prevalence?
- directly proportional to incidence and to average duration of disease, which is affected by levels of mortality and cure
- existing cases
- proportion of the chosen population who have a specific characteristic (disease/risk) in a given time period
What is incidence?
Number of new (or newly diagnosed) cases of the disease within a period of time in a specified population
What is rate ratio?
Ratio of incidence rates of outcome in groups defined by levels of exposure after a particular time period (e.g per 1,000 person-years)
If groups have identical time periods of follow-up then,
Rate ratio = risk ratio
What is odds ratio?
Ratio of odds of outcome in groups defined by levels of exposure at a particular time (no measurement of unit)
Measure of association between an exposure and an outcome
Odds that an outcome due to exposure: odds of the outcome in absence of exposure
What is a case-control study?
- backwards study/ retrospective
- help determine if an exposure is associated with an outcome
- start with cases (group known to have the outcome) and control (group known to be free of the outcome) then look back in time to learn which subjects in each group had the exposure
- compare frequency of exposure in case vs control
Advantages of case-control studies
- quick
- inexpensive
- easy
- can get a bigger sample (you already know the outcome)
Good for: - investigating outbreaks
- studying rare diseases/outcomes
