3 observational study designs
1) cross sectional 2) case control 3) cohort
In cross sectional studies, data are collected at (different times, a single point in time), describes (associations, exposure, causes), and can provide info on (morbidity, prevalence, incidence).
A single point in time, associations, prevalence
What is prevalence?
How common exposure/outcome is in a population
3 strengths of cross sectional studies
excellent for DESCRIPTIVE stats, relatively fast, relatively cheap
One weakness of cross sectional studies
can't determine cause/effect
Case control studies evaluate association as (relative risk, incidence, prevalence, odds ratio).
Cohort studies evaluate associations as (relative risk, incidence, prevalence, odds ratio).
4 advantages of case control studies
1) very good if disease is rare 2) relatively cheap 3) relatively fast 4) can examine many exposures
3 disadvantages of case control studies
1) finding controls can be difficult 2) CONFOUNDING by unexamined exposures possible 3) highly susceptible to bias b/c retrospective in nature
(Cohort, case control, cross sectional) studies are often first approaches to testing a hypothesis.
What kind of study begins with disease-free pt, determines exposure status, records outcomes in the future, and evaluates relative risk?
5 advantages of cohort studies
1) establish time sequence 2)measure incidence 3) eliminate recall bias 4) can evaluate multiple outcomes 5) can adjust for confounders
6 disadvantages of cohort studies
1) expensive 2) time consuming 3) not appropriate for rare outcomes 4) exposure may change over time 5) disease may have a pre-clinical phase 6) attrition of study sample
Randomized, controlled trials are a special kind of ____ studies.
Which study design is the best for determining causality?
Randomized controlled trials
Purpose of having a exclusion criteria
ensure pt safety
Purpose of having inclusion criteria
Advantage of having controlled groups in a study
help draw appropriate conclusions
3 different types of controls in a study
placebo, active, or historical
What is open label blinding?
What does parallel treatment allocation mean?
Groups receive DIFF treatments CONCURRENTLY
What does paired treatment allocation mean?
Subjects are matched for similarities
What does crossover treatment allocation mean?
Subjects receive each treatment
3 Advantages of randomized controlled trials?
1) best to assess causal relationships 2) best at controlling bias 3) can evaluate multiple outcomes
5 disadvantages of randomized controlled trials?
1) expensive 2) time consuming 3) ethical issue 4) loss to follow up 5) less "real world"
T/F: RCTs w/ or w/o meta-analysis is more reliable than RCTs.
T/F: Case reports are more reliable than cohort studies.
False; less reliable
T/F: Cohort studies are less reliable than case control studies.
Unproven therapies are most likely used in what two settings?
Critical care, pediatrics
Goal of clinical practice?
Well being of pt
Goal of clinical research?
Creation of generalizable knowledge where ppl conducting research on LIVE PEOPLE, not animals
T/F: Clinical practice and clinical research are often performed by the same person.
Goal of clinical practice is ____
well being of pt
Goal of clinical research is ____
generate new knowledge
T/F: Much "std care" involves intervention never adequately tested for safety and efficacy.
Unproven theories are most likely to be used in what 2 areas?
Critical care, pediatrics
PICO stands for______
patient, intervention, comparison, outcome
6 steps of critical literature appraisal
1) ID clinical question 2) format clinical question (PICO) 3) ID type of question (therapy, diagnosis, prognosis, etiology) 4) ID and summarize article 5) appraise article on basis of criteria appropriate to question 6) draw conclusion
3 types of publication bias
1) submission 2) methodologic 3) acceptance
10 departures in research process
1) conflict of interest 2) funding 3) study design 4) informed consent 5) randomization 6) data collection/recording 7) stat analysis 8) data ownership 9) research description 10) publication bias