I now have comprehensive content. Let me compile the complete exam-oriented notes.
Principles of Epidemiology & Epidemiological Methods
Community Medicine - Park's Textbook (Exam-Oriented Notes)
I. DEFINITION OF EPIDEMIOLOGY
Etymology: Epi (among) + Demos (people) + Logos (study)
Historical Definitions:
- "That branch of medical science which treats epidemics" - Parkin, 1873
- "The science of the mass phenomena of infectious diseases" - Frost, 1927
- "The study of disease, any disease, as a mass phenomenon" - Greenwood, 1934
- "The study of the distribution and determinants of disease frequency in man" - MacMahon, 1960 (most quoted)
Modern/Current Definition (Park's, most comprehensive):
"The study of the occurrence and distribution of health related events, states, and processes in specified populations, including the study of the determinants influencing such processes, and the application of this knowledge to control relevant health problems."
- Study = surveillance, observation, screening, hypothesis testing, experiments, prediction
- Distribution = analysis by Time, Place, Person
- Determinants = geophysical, biological, behavioural, social, cultural, economic, political factors
- Application = to promote, protect, and restore health
Milestones:
- W.H. Frost = first professor of epidemiology in USA (1927)
- Major Greenwood = first professor of epidemiology and medical statistics, University of London
II. USES OF EPIDEMIOLOGY (Morris - 7 Uses) - HIGH YIELD
- Historical study - rise and fall of disease in population; time trends
- Community diagnosis - identify and quantify health problems (mortality, morbidity rates); forms basis for health planning
- Working of health services - evaluating effectiveness and efficiency of health services
- Individual risks and chances - probability that an individual will develop a certain disease; "risk" in epidemiology
- Identification of syndromes - description and classification of clinical syndromes, determination of natural history
- Completion of the clinical picture - full spectrum of disease (clinical + subclinical), natural history
- Search for causes - aetiological research; identification of risk factors
Mnemonic: H-C-W-I-I-C-S or remember "7 uses by Morris"
III. CONCEPT OF CAUSATION
Pre-Pasteur Theories:
- Supernatural theory, Theory of humors, Contagion theory, Miasmatic theory, Spontaneous generation theory
Germ Theory (Koch-Pasteur era):
- One-to-one relationship: Disease agent → Man → Disease
- Limitations: not every exposed person develops disease; cannot explain chronic/multifactorial diseases
Koch's Postulates:
- Organism must be present in every case of the disease
- It must be isolated from the diseased host and grown in pure culture
- Disease must be reproduced when pure culture is introduced into a healthy susceptible host
- Organism must be re-isolated from the experimentally diseased host
Epidemiological Triad (Agent-Host-Environment):
- Three interactive factors determine disease occurrence
- Agent: biological, chemical, physical, nutritional, social causes
- Host: intrinsic factors (age, sex, genetics, immunity)
- Environment: physical, biological, psychosocial
- Disease occurs when there is imbalance between these three
Multifactorial Causation:
- Concept first proposed by Pettenkofer of Munich (1819-1901)
- Revived and developed by epidemiologists
- Most diseases have MULTIPLE causes - genetic, social, economic, environmental, psychological
- Gave rise to concept of "risk factors"
Web of Causation:
- Proposed by MacMahon
- Considers ALL predisposing factors and their complex interrelationships
- Visualizes multiple possible interventions
- Example: Web of causation for myocardial infarction
- Key principle: Removal of just ONE important link may be sufficient to control disease
IV. AGENT, HOST, ENVIRONMENT FACTORS - HIGH YIELD
Agent Factors:
- Biological agents: bacteria, viruses, fungi, protozoa, helminths, rickettsia
- Nutrient agents: excess or deficiency (e.g., protein deficiency, vitamin D excess)
- Physical agents: heat, cold, radiation, noise, trauma
- Chemical agents: poisons, allergens, drugs, industrial chemicals
- Mechanical agents: chronic friction, compression
- Absence/insufficiency/excess of a necessary factor
- Social agents: poverty, stress, occupational hazards
Host Factors (Intrinsic):
- Age, Sex, Genetic constitution
- Immunity (specific and non-specific)
- Nutritional status, Physiological state
- Personality, habits, lifestyle
Environmental Factors (Extrinsic):
a. Physical environment
- Geography, climate, soil, water, air quality
- Urbanization, radiation, noise, housing
b. Biological environment
- Living things: microbes, insects, rodents, animals, plants
- These act as agents, reservoirs, intermediate hosts, vectors
- Constant adjustment and readjustment (ecological balance)
c. Psychosocial environment
- Cultural values, customs, habits, beliefs, attitudes, religion, education, lifestyle
- Community life, social and political organization
- Psychosocial factors can cause psychosomatic disorders: duodenal ulcer, bronchial asthma, hypertension, CHD, mental disorders
V. RISK FACTORS
Definition: "An attribute or exposure that is significantly associated with the development of a disease"
Characteristics of a Risk Factor:
- Associated with increased probability of disease
- Precedes the disease
- Not sufficient in itself to cause disease (only increases probability)
Types:
- Modifiable: smoking, diet, hypertension, obesity, sedentary lifestyle
- Non-modifiable: age, sex, genetic constitution, family history
Risk Factor vs Cause:
- Risk factor = increases probability; need not be necessary or sufficient
- Cause = produces the disease (necessary, sufficient, or both)
VI. CRITERIA FOR CAUSATION (Hill's Criteria) - VERY HIGH YIELD
Proposed by Sir Austin Bradford Hill (1965):
- Strength of association - stronger the association (high relative risk), more likely causal; e.g., smoking-lung cancer: RR = 9-10
- Consistency - association observed repeatedly in different populations, places, times, by different researchers
- Specificity - one cause - one effect; a factor leads to a specific disease
- Temporality - exposure (cause) must PRECEDE the outcome (effect); MOST essential criterion
- Biological gradient (Dose-response) - as dose increases, disease frequency increases; e.g., more cigarettes smoked = higher lung cancer risk
- Plausibility - biologically plausible mechanism
- Coherence - association does not conflict with known natural history and biology of the disease
- Experiment - disease reduced when exposure removed (intervention evidence)
- Analogy - similar factors causing similar diseases (analogous situations)
Mnemonic: SCSTT BPCEA (Strength, Consistency, Specificity, Temporality, dose-response, Biological plausibility, Coherence, Experiment, Analogy)
Most essential criterion = Temporality (cause must precede effect)
VII. BASIC MEASUREMENTS IN EPIDEMIOLOGY - VERY HIGH YIELD
Key Terms:
- Numerator: number of times event has occurred
- Denominator: population at risk or total events
- Population at risk: those capable of having or acquiring the disease
- Mid-year population: population on 1st July (used as denominator for annual rates)
- Person-years: combination of persons × time (used in cohort studies)
Rate vs Ratio vs Proportion:
| Term | Definition | Example |
|---|
| Rate | Number of events per population per time | Incidence rate = new cases/population at risk/time |
| Ratio | Comparison of two numbers NOT part of each other | Male:Female ratio = 1.06:1 |
| Proportion | Numerator IS part of denominator; expressed as % | Attack rate = cases/exposed × 100 |
VIII. MEASURES OF DISEASE FREQUENCY
A. Incidence
Incidence rate = (Number of NEW cases in a period / Population at risk at beginning of period) × constant
- Measures rate of occurrence of new events
- Dynamic measure (time-dependent)
- Best for acute diseases
- Used to study aetiology
Attack Rate (Incidence proportion):
= (Number of new cases during outbreak / Population at risk at start) × 100
- Used in epidemic investigations
- Food-specific attack rate: determines which food caused outbreak
Secondary Attack Rate (SAR):
= (Number of cases among contacts / Total number of contacts) × 100
- Measures transmission in households or closed populations
- Indicator of communicability of the disease
B. Prevalence
Point Prevalence = (Number of cases at a point in time / Total population at that time) × 100
Period Prevalence = (All cases during a period / Average population during that period) × 100
- Measures existing cases (old + new)
- Static measure
- Best for chronic diseases
- Used for health planning (resource allocation)
Relationship: P = I × D
Prevalence = Incidence × Mean duration of disease
- High prevalence relative to incidence → long duration (e.g., TB, diabetes)
- Low prevalence relative to incidence → short duration (acute, rapidly fatal or rapidly recovering)
- Effective treatment that prolongs life without cure → increases prevalence
IX. MORTALITY MEASURES - HIGH YIELD
| Measure | Formula | Notes |
|---|
| Crude Death Rate (CDR) | Deaths/Population × 1000 | Affected by age-sex composition |
| Age-specific death rate | Deaths in age group/Population in age group × 1000 | More specific |
| Infant Mortality Rate (IMR) | Deaths <1 yr/Live births × 1000 | Most sensitive indicator of community health |
| Neonatal Mortality Rate | Deaths <28 days/Live births × 1000 | |
| Post-neonatal Mortality Rate | Deaths 28 days-1 yr/Live births × 1000 | |
| Child Mortality Rate (1-4 yr) | Deaths 1-4 yr/Mid-year population 1-4 yr × 1000 | |
| Under-5 Mortality Rate | Deaths <5 yr/Live births × 1000 | MDG indicator |
| Maternal Mortality Rate (MMR) | Maternal deaths/Live births × 100,000 | |
| Case Fatality Rate (CFR) | Deaths from disease/Cases of disease × 100 | Measures severity of disease |
| Proportionate Mortality Rate (PMR) | Deaths from a cause/Total deaths × 100 | NOT a true rate; shows relative importance |
Standardization of Rates:
- Direct standardization: apply age-specific rates of study population to a standard population
- Indirect standardization: apply standard rates to study population → calculate SMR (Standardized Mortality Ratio)
- SMR = (Observed deaths / Expected deaths) × 100
- SMR >100 = excess mortality; SMR <100 = deficit mortality
X. DESCRIPTIVE EPIDEMIOLOGY - HIGH YIELD
"Describes the distribution of disease in terms of Person, Place, Time"
Person (Who?):
- Age - most important host factor; strongly related to disease
- Bimodality: two peaks in age-incidence curve (e.g., Hodgkin's disease, leukaemia, breast cancer)
- Sex - sex-specific rates; some diseases more common in women (diabetes, hypothyroidism, obesity); others in men (lung cancer, CHD)
- Ethnicity/race - genetic and socio-cultural factors
- Occupation - occupational exposures, lifestyle
- Socioeconomic status - income, education, poverty
- Marital status, religion, diet, habits, migration
Place (Where?):
- International differences, national differences, urban-rural differences
- Spot maps, area maps
- Point clusters, area clusters
Time (When?):
- Secular (long-term) trends - changes over years/decades (e.g., decline of TB, rise of CHD)
- Cyclic/Periodic fluctuations - regular patterns:
- Seasonal variation: peak in certain seasons (e.g., cholera in summer, respiratory infections in winter)
- Annual cycles
- Short-term fluctuations (Epidemics):
- Common source: single contaminated source; explosive onset; all cases within one incubation period
- Propagated: person-to-person spread; gradual rise and fall; multiple incubation periods
- Irregular/Random variation
XI. EPIDEMIC INVESTIGATIONS - HIGH YIELD
Types of Epidemics:
A. Common Source Epidemics
- Point source (single exposure):
- Epidemic curve: rises and falls rapidly, no secondary waves
- All cases within ONE incubation period
- Explosive onset
- Example: food poisoning at a party
- Continuous/repeated exposure:
- Exposure ongoing from same source
- Epidemic extended or irregular; exceeds one incubation period
- Example: contaminated water supply, Legionnaire's disease (Philadelphia, 1976)
B. Propagated (Progressive Source) Epidemics
- Person-to-person transmission
- Gradual rise, tails off over longer period
- Multiple incubation periods visible on epidemic curve
- Continues until susceptibles depleted or exposure stops
- Speed depends on herd immunity, contact opportunities
- Examples: hepatitis A, polio, influenza
XII. ANALYTICAL EPIDEMIOLOGY - VERY HIGH YIELD
Types of Study Designs:
EPIDEMIOLOGICAL STUDIES
├── Observational
│ ├── Descriptive (person, place, time)
│ ├── Analytical
│ │ ├── Case-Control (Retrospective)
│ │ └── Cohort (Prospective)
│ └── Cross-sectional (Prevalence survey)
└── Experimental (Interventional)
├── Clinical trials
├── Field trials
└── Community trials
XIII. CASE-CONTROL STUDY - HIGH YIELD
| Feature | Description |
|---|
| Direction | Backwards (retrospective) - from effect to cause |
| Starting point | People with disease (CASES) and without (CONTROLS) |
| Exposure | Past exposure to risk factor is measured |
| Measure | Odds Ratio (OR) |
| Speed | Faster, cheaper, less subjects needed |
| Best for | Rare diseases, long latency diseases |
| Bias risk | Recall bias, selection bias (major limitations) |
Odds Ratio (OR):
= (a × d) / (b × c) from 2×2 table
| Disease (+) | Disease (-) |
|---|
| Exposed (+) | a | b |
| Exposed (-) | c | d |
- OR >1 = positive association (risk factor)
- OR <1 = protective factor
- OR = 1 = no association
Selection of Cases:
- Incident cases (newly diagnosed) preferred over prevalent cases
- Diagnostic criteria must be specified before study
- Sources: hospitals or general population
Selection of Controls:
- Must be free from the study disease
- Must be similar to cases except for absence of disease
- Sources: hospital controls, neighbourhood controls, relatives, general population
XIV. COHORT STUDY - HIGH YIELD
| Feature | Description |
|---|
| Direction | Forward (prospective) - from cause to effect |
| Starting point | People exposed and unexposed to risk factor |
| Outcome | Disease development is followed over time |
| Measure | Relative Risk (RR) = Attributable Risk, Incidence |
| Speed | Slower, costly, large numbers needed |
| Best for | Common diseases, rare exposures |
| Advantage | Can calculate incidence rates; temporal relationship clear |
| Bias risk | Loss to follow-up, Neyman bias (prevalence-incidence bias) |
Types:
- Prospective cohort - groups identified in present; followed into future
- Retrospective cohort (historical) - groups identified from records in the past; outcome assessed in present
- Ambidirectional - combination
Relative Risk (RR):
= Incidence in exposed / Incidence in unexposed
= [a/(a+b)] / [c/(c+d)]
- RR >1 = positive association
- RR = 1 = no association
- RR <1 = protective
Attributable Risk (AR):
= Incidence in exposed - Incidence in unexposed
Measures the absolute excess risk due to the factor
Comparison groups in cohort studies:
- Internal comparisons - different exposure levels within same cohort
- External comparisons - exposed vs. separate unexposed cohort
- Comparison with general population - exposed vs. general population rates
XV. CROSS-SECTIONAL STUDY (Prevalence Survey)
| Feature | Description |
|---|
| Timing | One point in time (snapshot) |
| Measures | Both exposure AND outcome at same time |
| Measure | Prevalence rate |
| Best for | Planning health services, prevalence of chronic diseases |
| Limitation | Cannot establish temporality (chicken-egg problem) |
| Also called | Prevalence study |
XVI. EXPERIMENTAL EPIDEMIOLOGY (Interventional Studies) - HIGH YIELD
Key difference from observational:
- Investigator directly controls the conditions; deliberately applies/withdraws intervention
- Always prospective
Types:
- Clinical trials - evaluate therapeutic agents (drugs, procedures) in patients
- Preventive trials - evaluate preventive measures (vaccines, chemoprophylaxis) in healthy persons
- Risk factor trials - intervene to modify risk factors (e.g., reduce cholesterol, stop smoking)
- Community trials - intervention applied to entire communities
Randomized Controlled Trial (RCT) - Gold Standard:
- Participants randomly assigned to treatment and control groups
- Randomization ensures comparability
- Blinding: Single-blind (subject unaware), Double-blind (both subject and observer unaware), Triple-blind (subject, observer, analyst unaware)
- Measures: Efficacy (under ideal conditions) vs. Effectiveness (under real-world conditions)
Phases of Clinical Trials:
- Phase I: Safety testing; small group; dose finding
- Phase II: Efficacy testing; small group; short duration
- Phase III: Large-scale RCT; compared with standard treatment or placebo; most important
- Phase IV: Post-marketing surveillance; long-term safety and efficacy
XVII. MEASURES OF ASSOCIATION - SUMMARY
| Measure | Study Design | Formula | Interpretation |
|---|
| Relative Risk (RR) | Cohort | IR exposed / IR unexposed | Risk ratio; how many times more likely |
| Odds Ratio (OR) | Case-Control | ad/bc | Approximates RR when disease is rare |
| Attributable Risk (AR) | Cohort | IR exposed - IR unexposed | Absolute excess risk |
| Population Attributable Risk (PAR) | Cohort | IR total - IR unexposed | Excess risk in whole population |
| Population AR% | Any | PAR/IR total × 100 | % of disease attributable to exposure |
XVIII. BIAS IN EPIDEMIOLOGY - HIGH YIELD
Types:
- Selection bias - systematic difference in selection of study and comparison groups
- Berkson's bias (hospital admission bias): hospitalized patients not representative
- Neyman bias (prevalence-incidence bias): prevalent cases ≠ incident cases (survivors only)
- Information (Observation) bias - systematic error in measuring exposure or outcome
- Recall bias: cases remember exposure better than controls (common in case-control studies)
- Observer bias: systematic error by observer
- Hawthorne effect: subjects change behaviour when they know they are being observed
- Confounding bias - third variable is associated with both exposure and outcome; distorts true relationship
XIX. SURVEILLANCE IN EPIDEMIOLOGY
Monitoring vs Surveillance:
| Monitoring | Surveillance |
|---|
| Performance and analysis of routine measurements aimed at detecting changes | Continuous collection, analysis, interpretation, and feedback of systematically collected data |
| More episodic/intermittent | Continuous and ongoing |
| Less sophisticated analysis | Requires professional analysis and judgement |
| Example: Air pollution monitoring | Example: Disease notification system |
Surveillance Definition (Park's):
"Systematic and continuous collection, analysis, and interpretation of data, closely integrated with the timely and coherent dissemination of the results to those who have the right to know so that action can be taken."
Sentinel Surveillance:
- Identifies missing cases not captured by routine notification
- Uses a network of selected health care providers (sentinels) to report specific diseases
- Supplements notified cases; particularly used for influenza, HIV, STIs
XX. CONCEPTS OF CONTROL, ELIMINATION AND ERADICATION
| Concept | Definition | Example |
|---|
| Disease control | Reduction of disease incidence, prevalence, morbidity or mortality to a locally acceptable level | Malaria control |
| Disease elimination | Reduction of incidence to zero in a defined geographical area | Neonatal tetanus elimination in India |
| Disease eradication | Permanent reduction to zero worldwide; intervention measures no longer needed | Smallpox (1980, WHO certified) |
| Extinction | Pathogen no longer exists in nature or labs | Not yet achieved for any pathogen |
XXI. HIGH YIELD EXAM SUMMARY TABLE
| Topic | Key Point |
|---|
| Best definition of epidemiology | MacMahon, 1960: distribution and determinants of disease frequency |
| Uses of epidemiology | Morris 7 uses |
| Most important criterion for causation | Temporality |
| Hill's criteria proposed by | Sir Austin Bradford Hill, 1965 |
| Web of causation - example | Myocardial infarction |
| Case-control measure | Odds Ratio |
| Cohort study measure | Relative Risk + Attributable Risk |
| Gold standard study | RCT |
| P = I × D | Prevalence = Incidence × Duration |
| Study for rare disease | Case-control |
| Study for rare exposure | Cohort |
| Study with recall bias | Case-control |
| Study with loss to follow-up | Cohort |
| IMR most sensitive indicator | Health status of community |
| Attack rate used in | Epidemic investigation |
| Secondary attack rate measures | Communicability |
| Standardized Mortality Ratio (SMR) | Observed/Expected × 100 |
| Sentinel surveillance | Identifies missing cases; used for influenza, HIV |
| Surveillance = | Continuous data collection + analysis + feedback + action |
| Disease eradicated | Smallpox (1980) |
| Bimodality example | Hodgkin's disease, leukaemia, female breast cancer |
| Point-source epidemic | All cases within one incubation period |
| Propagated epidemic | Gradual rise; multiple incubation periods; person-to-person |
| Phase III trial | Large-scale RCT; most important phase |
| Double-blind | Both subject AND observer blinded |
| Confounder | Variable associated with both exposure and outcome |
| Berkson's bias | Hospital admission bias in case-control studies |
Source: Park's Textbook of Preventive and Social Medicine, Chapters 3-4 - Principles of Epidemiology & Epidemiological Methods