Maths for Clinicians 7

avatar
Created by
Shalini

Cards (14)

  • What is a cross-sectional study?
    A snapshot of frequency of outcomes & exposures in a particular population at a specific time-point
    (Observational study so no intervention)
  • How are cross-sectional studies carried out?
    • Define target population
    • Use representative sample to take part
    • Obtain info on outcome of interest & exposure at particular time
  • What are the strengths of cross-sectional studies?
    • Data good quality as it's collected for the purpose of the research (not routinely)
    • Quick & cheap (no follow up)
    • Can study multiple exposures & outcomes
  • What are the weaknesses of cross-sectional studies?
    • Can't draw causal conclusions
    • Prone to measurement bias (e.g, recall bias- may not remember/ accurately report facts in questionnaires/ social desirability bias)
    • Confounding variables could cause the association
  • How is response rate calculated?
    Number of people in study sample (data collected from) / number of people in selected sample
  • What is risk ratio/relative risk?
    Summarises the association between exposure & outcome when outcome is binary
    RR = Risk in exposed group/Risk in unexposed group
    Risk in each group= Number with outcome / Total number in group
  • What does a risk ratio of 1 mean?
    Null value (no association between exposure groups)
  • What does a RR < 1 mean?
    Exposure reduces risk of outcome
    1 - RR = %
  • What does a RR > 1 mean?
    Exposure increases risk of outcome
    RR - 1 = %
  • What do smaller P values indicate?
    Stronger evidence against null hypothesis
  • What is representativeness?
    How representative is the sample of the target population?
  • What is generalisability?
    How generalisable are the results to other populations?
  • What do confidence intervals estimate?
    Population risk ratios
  • EXAMPLE:
    RR = 1.4
    95% CI = (1.3, 1.5)
    Risk of outcome in exposed compared to unexposed group is 40% higher. 95% confident that in general population risk of outcome in exposed compared to unexposed group is between 30 & 50% higher.