Statistics

• What is selection bias?

  when there are systematic differences between comparison groups
• Type 2 error

  The failure to reject the null hypothesis when it is false.
• Phase 2 trial

  Testing of drug on large group of patients to assess efficacy and safety
• How does randomisation of subjects make a randomised controlled trial more rigorous?

  reduces the risk of selection bias
• List the components of the ‘evidence pyramid’ going from strongest to weakest form of evidence.

  • Meta-analysis
  • Systematic Review
  • Randomized Controlled Trial
  • Cohort Studies
  • Case-control studies
  • Cross-sectional studies
  • Case reports
  • Expert Opinion
• Risk ratio

   probability of an event occurring in an exposed group compared to the probability occurring in a non-exposed group
• How to calculate the Number Needed to Treat?

  Number Needed to Treat is inversely proportional to the Absolute risk reduction (%)

  E.g. if ARR is 4% , then NNT is 1/4 x 100 = 25
• NNT = 25. What does this mean?

  An average of 25 people will have to receive the intervention for 1 of them to benefit compared to someone not receiving the intervention
• Pre-clinical trial

  Animal/cell testing to gather information about efficacy and toxicity
• Phase 0 trial

  Sometimes not done

  small group of volunteers, used to assess pharmacodynamics and pharmacokinetics
• Phase 1 trial

  Testing of drug on healthy volunteers to find appropriate dosing
• Phase 3 trial

  Testing of drug on a large group of patients to confirm effectiveness, safety and comparing to existing interventions
• How to calculate sensitivity?

  $\frac{true\ positive}{true\ postive\ +\ false\ negative}$
• How to calculate specificity?

  $\frac{true\ negative}{true\ negative\ +\ false\ positive}$
• How to calculate positive predictive value?

  $\frac{true\ positive}{true\ positive\ +\ false\ positive}$
• How to calculate negative predictive value?

  $\frac{true\ negative}{true\ negative\ +\ false\ negative}$
• How to calculate false positive rate?

  1 - specificity
• How to calculate false negative rate?

  1 - sensitivity
• Rate ratio

  Also known as incidence ratio

  Measure of association which compares the incidence rate in an exposed group to the incidence rate in a non-exposed group
• Confidence interval

  Range within which the true treatment effect is likely to lie
• Selection bias

  When there are systematic differences between comparison groups
• Which studies commonly use the odds ratio as a measure of association?

  Case-control studies

  May also be used in cross-sectional studies
• What is the main disadvantage of cross-sectional studies?

  Do not show causality or account for confounding factors which may influence data

  So, associations may be difficult to interpret
• Attributable risk

  Measure of association

  Amount of proportion of disease incidence (disease risk) that can be attributed to the exposure
• Cross-sectional study

  Collection of data from a population in a specific time point
• Type 1 error

  Incorrect rejection of the null hypothesis when it is true
• Calculating relative risk

  $\frac{risk\ of\ poor\ outcomes\ in\ intervention\ group}{risk\ of\ poor\ outcomes\ in\ the\ control\ group}$
• Measurement bias

  Error due to data collection during measurement

  Can occur in both qualitative & quantitative studies
• Odds ratio calculation

  $\frac{AD}{BC}$

  
• Interventional studies

  Participants receive 1 or more intervention or treatment

  Types

  • randomised control trial
  • pre-post study
• Observational studies

  Participants are observed without an intervention

  Types

  • cohort
  • case-control
• What is the Bradford-Hill Criteria?

  9 viewpoints to help determine if observed epidemiological associations are casual

  • strength -> strong associations
  • consistency -> repeated observation of an association in different populations under different circumstances
  • specificity -> cause leads to a single effect (not multiple)
  • temporality -> cause precede the effect in time
  • biologic gradient -> unidirectional dose responsive curve
  • plausibility -> biologic plausibility of hypothesis
  • coherence -> interpretation does not conflict with known natural history & biology
  • experimental evidence
  • analogy
• Odds ratio

  Probability that an event will occur given an exposure, compared to the probability that that event will occur absent the exposure

  Used in case-control studies, may be used in cross-sectional studies
• Interpretation of odd ratio

  OR = 1

  • exposure will not affect outcome

  OR < 1

  • exposure is associated with lower probability of outcome

  OR > 1

  • exposure is associated with higher probability of outcome
• Absolute risk reduction (AAR)

  Absolute difference in risk between control group & experimental group

  ARR = Risk(control group) - Risk(experimental group)
• Relative risk reduction (RRR)

  Proportional reduction in risk bestowed by the intervention compared to the control situation

  RRR = 1 - (Risk(experimental group) / Risk(control group)
• Selection bias

  Selection of individuals, groups or data for analysis in such a way that proper randomization is not achieved, thereby ensuring that the sample obtained is not representative of the population intended to be analysed

  Types

  • sampling
  • time interval
  • susceptibility
  • attrition