semester 2 Flashcards
what is the principle of evidence based medicine?
use of current best evidence in making decisions about the care of an individual patient
what is the difference between descriptive and inferential statistics?
descriptive = describing data
inferential = analysing data + drawing conclusions from the data
what is epidemiology?
study of distribution and determinants of health-related states or events in a specified population
and the application of this study to control health problems
what are the 2 methods used to carry out epidemiological investigations?
Surveillance + descriptive studies (used to study distribution):
* one group studied
* no explicit hypothesis
* study ends in development of possible hypothesis regarding cause + effect relationship
Analytical studies (used to study determinants)
* 2 or more groups studied for comparison
* definite hypothesis
* reject or accept hypothesis at end
what are the requirements of a sample population?
representative
unbiased (on target)
precise (well grouped)
what are the 2 types of validity?
internal validity:
* free from confounding, bias or random error
external validity:
* degree to which results from study can be applied to population
what are the 2 types of error?
Chance error (random):
* due to sampling variation (people selected)
* reduces as sample size increases
Bias error (systematic):
* difference between true value + expected value
* doesnt reduce with increase in sample size
what are the 2 types of bias?
selection bias:
* sample isnt representative of entire population
* groups within a study may not be comparable
* workers usually exhibit lower overall mortality than general population
Information bias:
* recall error - difference in recollection from study participants
* observer or interviewer has preconceived expectations or knowledge that influence result
* measurement errors
* misclassification - participants put in wrong group
what are the different measurements taken in a survey?
- prevalence (absolute risk)
- incidence (absolute risk)
- incidence rate ratio (relative risk)
- risk ratio (relative risk)
- odds ratio
- risk difference
- person years
what is meant by the term prevalance and how is it calculated?
- proportion of people who have disease at any given point in time
- counts people with disease (old and new)
- determines burden of disease
- useful to determine resource allocation
prevelance = num of people with disease / total population
what is meant by the term incidence and how is it calculated?
- num of new cases for disease in a given time frame
- useful when monitoring epidemics
- reported as a rate
incidence rate = num of new cases / sum of patient time at risk
what is meant by the term incidence rate ratio and how is it calculated?
- compares incidence rate in one group to another to see if exposure is associated with condition
IRR = incidence rate in GP 1 / incidence rate in GP 2
what is meant by the term relative risk and how is it calculated?
- determined in cohort study where there is a consistent follow-up for all patients
relative risk = absolute risk ratio of GP A / absolute risk ratio of GP B
what is meant by the term odds ratio and how is it calculated?
- comparison of odds of disease in one group compared to another
odds ratio = ad / bc
what is meant by the term risk difference and how is it calculated?
- difference in risk of an event in GP A compared to GP B
- no difference = 0
risk difference = absolute risk of GP A - absolute risk of GP B
what is meant by the term person years and how is it calculated?
- sum of total time of everybody followed up in a study
e.g. 1000 people followed up for 1 year = 1000 person years
what is the 95% confidence interval and how is it calculated?
- the range within which we can be 95% certain that the true value lies within
- if P > 0.05 = findings are statistically insignificant and failure to reject null hypothesis
- if P < 0.05 = findings are statistically signifant and null hypothesis is rejected
- if 95% CI spans over 1 (e.g. 0.5 - 1.52) then P > 0.05
(OR / e.f) to (OR x e.f)
- what is confounding?
- how can you overcome confounding?
- mixing of effect between exposure, the disease and a third variable (the confounding variable) - age, sex, ethnicity, smoking
- minimise by matching case and controls using confounding variables + standardisation
- what are ecological studies?
- what are issues with ecological studies?
- identify groups of people to study (countries rather than individuals) + gather data on group-level characteristics
- measurement variation, confounding + chance
- what are cross-sectional surveys?
- what are the issues with cross-sectional surveys?
- surveys which measure exposure + outcome simultaneously + is most useful in determining prevalence + used for individuals
- sampling bias, participant bias, chance + confounding
what are case-control studies?
- always retrospective (looking back in time)
- groups of cases and non-cases
- ascertains previous exposure status
- compares levels of exposure in case + controls
Analysis:
* only odds ratio
- what is the difference between conventional case-control study and nested case-control study?
- what is the advantage of nested case-control study?
- conventional = always retrospective data whereas nested = data from a database of concurrent or prospective cohort study
- incidence rates can be calculated, can collect more detailed info for a minority of participants
- what are the advantages of case-control studies?
- what are the disadvantages of case-control studies?
- good for rare diseases, cheap + quick
- selection bias, information bias and confounding
what are cohort studies?
- always prospective look forward in time
- group individuals according to level of exposure
- select outcome free individuals
- ascertain outcomes for everyone
- compare incidence rates for each group
Analysis:
* can use odds ratio or rates ratio
what is the difference between (concurrent or prospective) and (historical or retrospective) cohort studies?
concurrent or prospective:
* find exposed and unexposed at current time
* wait for outcomes to occur
historical or retrospective:
* use data from pre-existing records
* look at past exposure and determine whether exposure has occured
- what are the advantages of cohort studies?
- what are the disadvantages of cohort studies?
- better at studying a range of different outcomes + studying rare exposure
- some people may drop out so cant follow up, information bias, confounding + expensive
what is PICO?
population
intervention
comparison
outcome
what is the standardised mortality ratio and how is it calculated?
ratio between observed num of deaths in study population / num of deaths that would be expected in general population
SMR = observed num of deaths / expected num of deaths