Statistics Flashcards
Name the study: Snap shot in time, comparing two things in a population at a given time.
Cross sectional study ie a clinical audit
What is the disadvantage of a cross sectional study?
Cant prove causality
What studies can be used to answer a ‘prediction’ question?
cohort (follow-up over time) and case control
Name the study: Identifying patients with uncommon outcomes and looking back in time to link exposure etc
Case control studies
Which study is higher in quality of evidence? case control or cohort?
Cohort
What is the 1st and 2nd best study design for evidence?
Meta analysis
Systemic review
Which study is best to answer an intervention question?
RCT- if ethical
What is a weakness of systematic reviews?
Its only as good as the studies included in it
Name 2 strengths of an RCT
randomisation distributes known/unknown confounders, blinding reduces bias and can measure effect of internvention
Name 2 weakness of RCT
ethics, cost-time, failure of allocation concealment
Name 2 strengths of a cohort study
proves temporal causation, cheaper and easier than RCT, can show PROGNOSIS
Name 2 weaknesses of a cohort study
unknown confounders, rare conditions, lost to follow-up
Name 2 strengths of a case control study
quick, cheap, feasible for rare disorders
Name 2 weaknesses of a case control study
selecting controls, recall and selection bias
A case-control study is a type of observational study commonly used to look at factors associated with diseases or outcomes. The case-control study starts with a group of cases, which are the individuals who have the outcome of interest. The researcher then tries to construct a second group of individuals called the controls, who are similar to the case individuals but do not have the outcome of interest.
Name 2 advantages of a cross sectional study
simple, cheap, can quantify condition/risk factor
Name 2 weaknesses of a cross sectional study
No cause/effect, confounding bias, only to generate hypothesis.
‘audits or surveys on the street’
What is the solution for confounding variables in a study?
randominsation
What is Kappa (Cohen’s) statistics?
agreement between different people assessing results of the study.
1= 100% agreement and 0= no agreement
Clinical studies need high kappa stats
What is external validity
Ability to generalise data
What is the key reason for failure in RCTs?
Poor allocation concealment
Who is the most important person/group to blind in a RCT?
The assessor (intervention should ideally also be blinded)
What is the intention to treat analysis and why is it done?
Analyse participants in groups here they were randomised. This preserves randomisation and is more realistic
What is convenience sampling?
The worst form of sample take 50 people at this shop right now.
Volunteer and selection bias.
What is cognitive bias?
Relate to bias in clinical decision making, not study design. This cognitive bias affects the way we test and evaluate hypotheses every day.
