4 Observational Study Designs

• Case-control: looks backward from disease
• Cohort: looks forward from exposure
• Cross-sectional: a single-moment snapshot
• Ecological: compares whole groups

Case Tiers

• Confirmed: positive lab
• Probable: full symptoms, no lab yet
• Possible: some symptoms, no lab
• Not a case: other agent or ruled out

CDC 13 Steps (write these)

1. Prepare for field work
2. Establish outbreak exists
3. Verify diagnosis
4. Construct case definition
5. Find cases & record info
6. Descriptive epi (time/place/person)
7. Develop hypotheses
8. Evaluate hypotheses
9. Reconsider & refine hypotheses
10. Compare w/ lab & environmental
11. Implement control & prevention
12. Initiate/maintain surveillance
13. Communicate findings

Grouped as our 10: 4+5, 9+10, 11+12.

Test Accuracy

• Sensitivity = TP ÷ all who have disease
• Specificity = TN ÷ all without disease
• VE = 1 − RR State/Nats

Case Definition Triad

• Person: clinical/demographic criteria
• Place: geographic boundary
• Time: onset date range

Chain of Infection (6 links)

1. Infectious agent
2. Reservoir
3. Portal of exit
4. Mode of transmission
5. Portal of entry
6. Susceptible host

Key Formulas

• Attack rate = cases ÷ population at risk
• Relative risk = AR exposed ÷ AR unexposed
• Odds ratio = (a×d) ÷ (b×c)

Causation Frameworks

• Koch's Postulates (4): single infectious agent
• Hill's Criteria (9): modern, flexible
• Evans's Postulates: bridges the two
• Causal Pie Model: multiple component causes