Please enable JavaScript.

Coggle requires JavaScript to display documents.

Biostat - Coggle Diagram

- - - - Randomization
  - - - Double blinding
  - - - Case control studies
  - - - Migraine headache
  - - - New screening test is introduced
    - - Poor prognosis diseases
      - Lung cancer
      - Pancreatic cancer
  - - - Smoking confounds relation between maternal alcohol intake and low birth weight
  - - - Keep subjects unaware of being studied
  - - - Few false negatives
      - Good for screening
  - - - Few false positives
      - Good for confirmation
  - - - Decrease sensitivity
      - Increase specificity
    - - Increase sensitivity
      - Decrease specificity
  - - - Positive result does not strongly indicate disease
  - - - Low NPV
    - - High NPV
    - - NPV 96% means 96% chance patient is disease free if test is negative
  - - - High sensitivity
      - High NPV
      - Decrease false negatives
      - Good for ruling out disease
    - - High specificity
      - High PPV
      - Decrease false positives
      - Good for ruling in disease
  - - - Increase NPV
      - Decrease false negatives
    - - Increase PPV
      - Decrease false positives
- - - - Risk of developing a disease over a time period
      - Example: 1 in 10 = 10%
    - - Compares risk between two groups
      - Example:
        
        Absolute risk in non smokers = 4 in 100
        
        Relative risk increase in smokers = 50%
        
        Absolute increase = 50% of 4 = 2
        
        Absolute risk in smokers = 6 in 100
    - - Risk = 2 in 20
      - Treatment reduces relative risk by 50%
      - 50% of 2 = 1
      - New absolute risk = 1 in 20
  - - - Drug group = 2%
      - Placebo = 4%
      - ARR = 4% − 2% = 2%
    - - Absolute risk = 6 in 100
      - Relative risk = 50%
      - Drug reduces risk by one third
      - Absolute risk becomes 4%
      - Correct answer: a
  - - - NNT = 1 / ARR
    - - Risk = 4 in 20 = 20%
      - Reduced to 2 in 20 = 10%
      - ARR = 10%
      - NNT = 10
    - - Condition in 12 out of 100
      - Relative risk = 25%
      - Drug decreases condition by 1/3
      - NNT = 25
      - Correct answer: d
    - - Condition in 8 out of 100
      - Drug decreases condition by 25%
      - NNT = 50
      - Correct answer: b
  - - - p = 0.06 → not significant
      - p = 0.04 → significant
      - p = 0.01 → most accurate and superior
  - - - Neither investigator nor patient knows allocation
    - - Most important for validity