case unit 1 - cholera Flashcards
which averages should you use for quantitative data
median
mean
why is the mode not useful for quantitative data
there might be more than one mode
each value in the study might only appear once
mode could be a high or low number (far from middle)
what methods do you use to quantify variation?
sd
iqr
range
benefits of case report study design
can identify new trends/diseases
helps detect new drug side effects
identifies rare manifestations of a disease
case report study
reports a new case of newly identified symptoms or outcomes
observation of symptoms, diagnosis and treatment of the individual case
presents hypothesis to be confirmed by another study type
types of observational study
case report
case control
cohort
limitations of case report study
may not be able to generalise the individual cases
report not based on systematic studies
other possible explanations for outcomes
report only focuses on rare event, may be misleading
case control study
‘retrospective’ - disease has already occurred
compares patients with disease to those without
evaluates relationships between risk factors and disease
estimates odds
benefits of case control study
good for studying rare diseases
takes little time as disease has already occurred
multiple risk factors can be studied at the same time
risk factor-disease associations can be established
limitations of case control
problems with data quality and data relies on memory
recall bias as people remember worse things
hard to find suitable control group
cohort study
evaluations of causative risk factors determined after following cohort populations during their disease
measurements taken of outcomes
benefits of cohort study
subjects are matched which limits influence of confounding variables
standardisation of criteria/outcomes
easier and cheaper than rct
limitations of cohort study
difficult to identify cohorts due to confounding variables
no randomisation
blinding of subjects is difficult
takes time for outcome of interest to occur
randomised control study
subjects randomly assigned to control or intervention group
only variable studied is the expected difference between the groups
benefits of rct
no population bias due to randomisation
easier to mask/blind subjects
statistical analysis using known methods easy
clearly identified populations
limitations of rct
expensive
time-consuming
volunteering populations not representative of entire population
causation of disease not revealed
practice guidelines
produced by panel of experts
guidelines on prevention/treatment/diagnosis/prognosis of disease
informs health care professionals
systematic review
combines and summarises all information from previous studies on one health topic or question