Statistics Flashcards
What is a cross sectional study?
OBSERVATIONAL - analyses data from
- a population
- a representative subset
- a specific point in time
- recall biased group
e. g. a survey
What is a cohort study?
OBSERVATIONAL
- in which participants are recruited for a study
- followed up over time
- exposures and diseases are measured prospectively (likely, or expected to happen)
What is a randomised control trial/clinical trial (RCT)?
INTERVENTIONAL
- participants allocated by chance to different interventions and followed up with outcomes assessed
- Key features - inclusion/exclusion criteria, control/comparison, blinding, randomisation
- Disadvantages - more difficult to control than observational, costly, ethical issues, not suitable for all types of research questions
What is a case series report?
Description of medical history of one or more patients
Outcome of interest
Lacks a control group
What is a case control study?
People with a disease are matched to those without it
Earlier exposure to different factors are compared
Lacks control group
Great for aetiology
What is a Systematic review/Meta analysis?
All the evidence for RCTs looking at effectiveness of a particular treatment are synthesised
Require multiple RCTs in order to exist
find literature > assess quality > combine it
What is the difference between prevalence and incidence?
Prevalence - related to a specific point in time
Incidence - is the rate something in a period of time
How do you calculate risk?
Risk = number of events of interest/total number of observations
e.g. 24 ppl ski down slope, 6 fall
risk of falling = 6/24 = 0.25 or 25%
How do you calculate odds?
Odds = number of events of interest/number without the event
e.g. 24 ppl ski down slope, 6 fall
odds of falling = 6/(24-6) = 0.33
How do you calculate absolute risk difference?
Difference between groups
ARD = RISK 1 (%) - RISK 2 (%)
e.g. Risk of G1 = 63%, Risk of G2 = 18%
ARD = 63-18 = 45%
If no benefit then ARD = 0
What is the number needed to treat?
Is the number of patients youd need to treat to prevent one from developing disease/condition/outcome
Calculated as
NNT - 1/ARD
e.g. if ARD was 45%
1/0.45 = 2.22 NNT = 2.22
therefore youd need to treat 3 people
How do you calculate risk ratio?
RR = RISK1/RISK2
e.g. 63% and 18%
63/18 = 3.42
so person is 3.42 x more likely to get pain relief than someone in placebo group
Cannot get 0 value for this so if answer is 1 then there is no difference in risk ratio between 2 criteria
How do you calculate odds ratio?
Ratio of odds of pain relief in both groups (usually intervention/control studies)
OR = ODDS1/ODDS2
e.g. O1 –> 40/23 = 1.74
O2 –> 5/22 = 0.23
1.74/0.23 = 7.56
REMEMBER ODDS NOT IN %
What are confidence intervals?
Taken a sample of participants/patients
CI’s quantify level of uncertainty of risk ratios etc
For a ratio where value of no difference = 1
- if CI contains/overlaps/straddles 1 - insufficient evidence theres a difference between intervention and placebo
- if CI doesnt contain/overlap/straddle 1 - suffieicent evidence to say theres a difference, i.e. value > 1
