How to search lit Flashcards
PICO method
Population/problem
Intervention
Comparison
Outcome
Questions in critical appraisal
what is the clinical question? why was the question asked? what did they do? what was the answer? what did they say about the answer? what do I do with this info?
Diagnostic test
evaluates a test for diagnosing a disease
differential diagnosis
weighing the probability that one disease rather than another accounts for a pts illness
tries to sort out what proportion of the pts with a single sign or symptom has various diseases
economic analysis
provides accurate values to assess the cost of disease and the cost-benefit of interventions
harm/etiology
addresses how to identify causes for disease including those caused by the healthcare system
prognostic
how to estimate the pts likely clinical course over time and anticipate likely complications of disease
practice guideline
a systematically developed statement on medical practice for specific medical conditions
Qualitative study
no stats
ie. beliefs, attributes, meanings, symbols
systematic review
focused on a single question and gather all research in relation to the question
meta analysis
the results of studies are similar enough that they are combined and analysed as one
therapy
how to select treatments to offer pts that do more good than harm and are worth the efforts and costs
qualitative research
study things in their natural setting
holistic view
inductive reasoning: bottom up approach, make obs and recognise patterns
deductive reasoning: top down approach, start with a theory and develop a hypothesis based on that theory
Methods of qualitative research
case studies behavioural studies ethnopgraphy action research personal
quantitative research
uses numbers and data
strength of research lies in reliability of study
uses deductive reasoning
types: observational - patterns of injury and illness in population
experimental - compare to baseline or placebo
Observational research: case series
cases selected on a single topic for defined period
comparison includes opinion of reviewer
observational research: cross-sectional
collects data from a sample population at one point in time
Observational researchL case control
series of cases and a control or comparison group
Observational research: cohort studies
participants are defined on basis of exposure or non exposure to a suspected risk factor
exposure measured before outcome occurs
Experiement research: controlled clinical trials (CCT)
experimenter manipulates the situation
comparison of groups
experiemental research: randomised controlled trials (RCT
subjects randomly allocated to groups
study and control groups
power calculations
sample size calcs are about making an educated guess to help ensure the study has a sufficiently high change of finding and effect and is not overpowered and wasteful
type 1 error
probability of finding a statistically significant effect when the truth is that there is no effect
Type II error
probability of not finding a statistically significant effect when one really does exist
Mechanics of hypothesis testing
null hypothesis alternative hypothesis decision rule p value apply decision rule
null hypothesis (Ho)
Ho = no effect or association
assume its true (Straw man)
alternative hypothesis (Ha)
finding of some effect
one or 2 sided
decision rule
virtually always: reject Ho if p less than .05
this cutpoint = alpha level
calculate p value
assume Ho is true
do study/gather data
calculate probability that we’d see something as extreme as what we observed IF Ho were true
applying decision rule
if P less than cutpoint reject Ho ie. there is an effect
if P greater than cutpoint do not reject Ho i.e there is not an effect
to calculate sample size
effect size, SD(d), cutpoint for decision rule, power we want for study, what statistical test we will use
