Chapter 12 Estimating Risk

Chapter 12  Estimating Risk:  Is there an Association? 

Define, explain and recognize the following in examples: 

Key Definitions 

Risk 

• The chance that something (like a disease) will happen. 

• Example: If 3 out of 10 people get sick, the risk is 3/10 or 30%. 

Absolute Risk (AR) 

• Also called incidence. 

• Just tells you how common the disease is in a group. 

• Example: If 20 out of 100 exposed people get sick → AR = 20%. 

Relative Risk (RR) 

• Tells you how much more likely the exposed group is to get sick compared to the unexposed group. 

• Formula: Risk in exposed ÷ Risk in unexposed 

• Example: If exposed group has 20% risk and unexposed has 10% → RR = 2 (means 2x the risk). 

Excess Risk 

• The extra risk caused by the exposure. 

• Formula: Risk in exposed - Risk in unexposed 

• Example: If exposed = 20%, unexposed = 10%, then excess risk = 10%. 

Odds 

• A way of saying how likely something is compared to it not happening. 

• Example: If 2 people are sick and 8 are not → odds = 2/8 = 0.25. 

Odds Ratio (OR) 

• Compares the odds in exposed vs. unexposed. 

• Example: Odds of disease in smokers vs. nonsmokers. 

Attack Rate 

• A special kind of risk used during outbreaks. 

• Formula: # of new cases ÷ total people at risk 

• Example: 10 people get food poisoning out of 50 who ate at a buffet → Attack rate = 20%. 

Study Designs 

1. Randomized Controlled Trial (RCT) 

• Design: Randomly assign people to treatment or control group. 

• Use it when: You want to test a new drug or treatment. 

• Advantages: Best for proving cause-and-effect (causation), no bias. 

• Disadvantages: Expensive, sometimes not ethical. 

• Calculates: AR, RR, OR. 

2. Cohort Study 

• Design: Group people by exposure, follow over time to see who gets disease. 

• Use it when: You want to study how an exposure affects risk. 

• Advantages: Can find incidence (new cases), can calculate RR. 

• Disadvantages: Takes time, not good for rare diseases. 

• Calculates: AR, RR, OR. 

3. Case-Control Study 

• Design: Start with people who already have the disease (cases), and compare them to people without it (controls). Look back to see exposure. 

• Use it when: The disease is rare or takes a long time to develop. 

• Advantages: Quick, cheap, good for rare diseases. 

• Disadvantages: Can’t get incidence or RR. Subject to recall bias. 

• Calculates: Only OR. 

Association vs. Causation 

• Association: Two things happen together (like smoking and lung cancer). 

• Causation: One thing actually causes the other. 

• Association is always determined first. Causation takes stronger evidence (like from RCTs). 

Q&A Summary 

• Another name for absolute risk: Incidence. 

• Excess risk = Risk in exposed - Risk in unexposed. 

• Relative Risk = Risk in exposed ÷ Risk in unexposed. 

• Interpret RR: 

• RR = 1 → No difference 

• RR > 1 → Exposure increases risk 

• RR < 1 → Exposure lowers risk 

• Why can’t you calculate RR in case-control? 
  You don’t know how many total people were at risk — you start with people who already have the disease. 

• Can you calculate OR in both cohort and case-control studies? 
  Yes! 

• RR vs. OR? 
  RR uses risk, OR uses odds. RR is easier to understand. OR can exaggerate risk when disease is common. 

• When is OR a good estimate of RR? 
  When the disease is rare. 

• Why is absolute risk important? 
  It shows how likely someone is to get the disease — very helpful for doctors and patients. 

• What does RR measure? 
  How much the exposure increases or decreases the risk. 

• What does OR measure? 
  How much more likely the exposure is in cases vs. controls