Epidemiology Essentials: Concepts and Methods
Posted by Anonymous and classified in Medicine & Health
Written on in 
with a size of 7.49 KB
I. Foundations
Epidemiology: The study of the distribution (who, where, when) and determinants (why, how) of health and disease in populations.
- Goal: Reduce morbidity and mortality through prevention.
- Scientific Method: Observation → Hypothesis → Test → Conclusion → Action.
Key Definitions:
- Distribution: Frequency and pattern of health events by person, place, and time.
- Determinants: Causes, risk factors, or exposures that affect health outcomes.
- Population: The group under study.
- Prevention: Primary (prevent onset), Secondary (early detection), Tertiary (reduce disability).
II. Measures of Disease Frequency
| Measure | Formula | Notes |
|---|---|---|
| Count | Total number of cases | Simplest measure |
| Proportion | X ÷ (X + Y) | Numerator is part of denominator |
| Percentage | (X ÷ (X + Y)) × 100 | Proportion expressed as percent |
| Ratio | X ÷ Y | Numerator/denominator independent |
| Rate | (Cases ÷ Population) × Time unit | Includes time element |
Prevalence – “Burden of Disease”
Formula: Prevalence = (Existing cases ÷ Population) × Multiplier
- Point prevalence: At a specific time.
- Period prevalence: Over a specific time frame.
- Example: 75 ÷ 50,000 = 0.0015 → 15 per 10,000.
Incidence – “New Cases or Risk”
Formula: Incidence = (New cases ÷ Population at risk) × Multiplier
- Example: 44,232 ÷ 290,809,777 × 100,000 = 15.2 per 100,000.
- Relationship: Prevalence = Incidence × Duration of Disease.
Statistical Measures:
- Central Tendency: Mean, Median, Mode
- Variation: Range, Variance, Standard Deviation
III. Mortality Measures
| Measure | Formula | Notes |
|---|---|---|
| Crude Mortality Rate | (All deaths ÷ Midyear population) × 1,000 | Unadjusted rate |
| Case Fatality Ratio | (Deaths from a disease ÷ Cases of that disease) × 100 | Indicates lethality |
| Proportionate Mortality Ratio | (Deaths from cause ÷ All deaths) × 100 | % of deaths due to specific cause |
| Cause-Specific Death Rate | (Deaths from cause ÷ Population) × 100,000 | Compares causes |
| Age- or Sex-Specific Rate | (Deaths in subgroup ÷ Group population) × 100,000 | By subgroup |
| Age-Adjusted Rate | Removes effect of different age structures | For fair comparison |
IV. Study Design
| Type | Unit | Time | Best For | Measure | Strengths/Limits |
|---|---|---|---|---|---|
| Case Report/Series | Individual(s) | None | New/unusual events | None | Simple, generates hypothesis |
| Ecologic Study | Population | Varies | Group-level trends | Correlation | Fast, cheap, ecologic fallacy |
| Cross-Sectional | Individual | Single point | Prevalence data | Prevalence | No temporality |
| Case-Control | Individual | Retrospective | Rare disease | Odds Ratio | Recall bias; no incidence |
| Cohort Study | Individual | Follow-up | Rare exposures | RR, AR | Costly, loss to follow-up |
| RCT | Assigned | Prospective | Testing intervention | RR, AR | Best for causality; expensive |
V. 2×2 Table Structure
| Disease Present | Disease Absent | Total | |
|---|---|---|---|
| Exposed | A | B | A + B |
| Not Exposed | C | D | C + D |
| Total | A + C | B + D | A + B + C + D |
Formulas:
- Odds Ratio (OR): (A × D) ÷ (B × C)
- Relative Risk (RR): [A ÷ (A + B)] ÷ [C ÷ (C + D)]
- Attributable Risk (AR): [A ÷ (A + B)] – [C ÷ (C + D)]
VI. Causation & Association
Association ≠ Causation
- Necessary Cause: Must be present for disease to occur.
- Sufficient Cause: Combination of factors that can produce disease.
- Component Cause: Part of a causal mechanism.
Bradford Hill’s Criteria
- Temporality: Exposure before outcome (Required).
- Strength of Association: High RR/OR.
- Consistency: Reproducible in different studies.
- Biologic Gradient: Dose-response relationship.
- Plausibility: Makes biological sense.
- Coherence: Does not contradict known facts.
- Experiment: Based on intervention evidence.
- Analogy: Similar known associations.
- Specificity: 1 exposure → 1 disease.
VII. Validity, Bias, and Confounding
- Internal Validity: Study accuracy (free from bias/confounding).
- External Validity: Ability to generalize results.
- Random Error: Sampling variability; minimize by increasing sample size.
- Systematic Error (Bias): Reproducible inaccuracy.
- Selection Bias: Errors in recruitment.
- Information Bias: Misclassification or measurement error.
- Confounding: Third variable distorts the relationship.
VIII. Screening & Test Validity
- Sensitivity: a ÷ (a + c)
- Specificity: d ÷ (b + d)
- PPV: a ÷ (a + b)
- NPV: d ÷ (c + d)
- Trade-off: Raising sensitivity lowers specificity.
IX. Prevention in Public Health
- Primary: Before onset (e.g., Vaccination).
- Secondary: Early stage (e.g., Screening).
- Tertiary: After diagnosis (e.g., Rehabilitation).
X. Randomized Controlled Trial (RCT)
- Randomization: Balances known and unknown confounders.
- Masking: Single (participants), Double (participants/investigators), Triple (analysts).
- Efficacy: Ideal conditions.
- Effectiveness: Real-world conditions.
XI. Health Data Sources
- NHANES: National Health and Nutrition Examination Survey.
- BRFSS: Behavioral Risk Factor Surveillance System.
- YRBS: Youth Risk Behavior Survey.
- NNDSS: National Notifiable Diseases Surveillance System.