Basic of Epidemiology😊

Basic of Epidemiology

Definition

• Study of:
• distribution
• determinants
• diseases
• health-related states
• events

• Plus:
• measures to control or prevent the disease

PSM Scientists

• Father of Public health → Cholera

Microbiology scientists

Louis Pasteur

• Mnemonic: FATHER CAR
• Developed vaccines for:
• Cholera.
• Anthrax.
• Rabies.

• Louis → Liquid (Liquid media)

• Vaccine CAR (coined vaccine, Germ cell theory) → Autoclave

• Pasteurisation → Fermentation

• Father of modern microbiology.

• Founded Pasteur Institute in Paris.

• Developed Pasteurisation (for milk).

• Key contributions:
• Liquid media.
• Fermentation Principle.
• Autoclave.
• Disapproved Theory of abiogenesis.
• Proposed Germ cell Theory.
• Coined the term "vaccine"

Robert Koch

• Father of modern microbiology.

• Significant contributions:
• Mnemonic:
• Koch organisms (koch bacilli - tb, cholera organism)
• are solid (solid culture media) → hang (hanging drop motility) → apply paint (aniline dye)

• Koch postulates.

• Discovery of Koch Bacilli (Tuberculosis).

• Identification of Cholera organism.

• Development of Solid culture media.

• Application of Aniline dye colour.

• Introduction of Hanging drop motility.

Koch Postulates

• Old (4) Postulates:
1. Constant association of causative organisms with disease.
2. Isolation in culture media is possible.
3. Culture growth inoculated in animals should produce the same lesion.
4. Re-isolation from the experimental animals is possible.

• New (1) Postulate:
• Humans should produce antibodies in serum whenever there is an infection (antigen).

• Exceptions from Postulates:
• Mycobacterium leprae.
• Treponema Pallidum.
• N. Gonorrhoea.
• My pallu gone

Paul Ehrlich

• Identified the Ehrlichia organism.

• Father of chemotherapy.

• Developed Toxin-antitoxin standardisation.

• Developed the Acid-fast stain/Ziehl Neelsen stain.

Other Scientists

• Walter Reed
• Worked on yellow fever
• Identified Aedes mosquito as vector
• Walter → Water → aedes → yellow fever

• Joseph Lister:
• Father of Antiseptic surgery.
• Used carbolic acid.

• Anton Von Leeuwenhoek:
• Father of the Light microscope (unilocular)
• First to observe "little animalcules"

• Ernst Ruska:
• Invented the Electron microscope

• Edward Jenner:
• Developed the first vaccine for smallpox.
• Coined the term "vaccination"
• Wardil (Edward) kidakkunna small (smallpox) kidsnu vaccination () cheyynm

• Karry B Mullis:
• Developed PCR.
• KBM PCR 3 letters

• H C Gram:
• Developed Gram staining.

• Kleinberger:
• Discovered L forms (cell wall deficient forms).
• K- L

• Alexander Fleming:
• Discovered Penicillin.

• Barbara McClintock:
• Discovered Transposons (Jumping genes)
• Tick tok - Jumping genes

Nobel Prizes During Covid Era

• HCYV mechanism:
• Michael Houghton
• Harvey J. Alter
• Charles M. Rice

Tools of Measurement

Ratio

• Numerator & denominator are separate entities.

• tio → two → 2 entities

Examples:

• Maternal Mortality Ratio (MMR) =
• (Maternal deaths / Live births) × 1,00,000

• No. of healthy / No. of sick

• Dead / Alive

Non-Ratio

Rate

• Numerator is part of denominator

• ate a part of big part

• Expressed in relation to time

• Uses multiplicative factors (1000, 10,000, etc.)

• Examples:
• Maternal Mortality Rate =
(Maternal deaths / Women in reproductive age) × 1,00,000
• Incidence = No. of new cases / Time
• Birth rate
• Death rate

Proportion

• Numerator is part of denominator

• No account of time

• Expressed in percentage

• Multiplier in proportion is always 100.

• Examples:
• Prevalence = Total no. of existing cases
• 2° attack rate

Incidence vs. Prevalence

Prevalence = Incidence × Duration

Impact of Treatment/Prevention on Incidence & Prevalence

Epidemiological Methods

Epidemiological Study Design

• ODA EaRN Evidence

A. Observational (Mnemonic: ODA)

1. Descriptive

• Buzzword
• Distribution of a disease in terms of time, place and person
• Formulation of hypothesis

• Describe a disease → No comparison group

• Collection, Formulation, presentation

• Types
1. Case study:
• Single case description.
2. Case series:
• Multiple cases description

2. Analytical

• Testing of hypothesis

• Buzzword → Determinant

• Analyse → Comparison group is always present

B. Experimental

• Confirmation of hypothesis

Experimental (Intervention) Studies

C. Evidence-based medicine

• Systematic review

• Meta analysis

Hypothesis

• An assumption that is yet to be verified.

Descriptive Studies :

Important

• Buzzword → Distribution of a disease in terms of time, place and person

• No comparison group

Time distribution

• Short term trend
• Outbreaks
• Epidemics

• Periodic trend
• Cyclic trends:
• Dengue: 1-3 years
• Measles: 2-3 years
• Rubella: 6-9 years
• Influenza: 7-10 years

• Long term trend
• AKA secular trends.
• Means progressive change in disease occurrence
• Epidemiological transition
• Shift from an era of communicable diseases to non-communicable diseases.
• ↑↑ cases of non-communicable diseases

Person distribution

• Age, gender, marital status, social class.
• Age distribution:
• Chicken pox: 1-10 years
• Measles: 2 yr – 3 years
• Typhoid: 5-19 years
• Diphtheria: <5-7 years
• Rubella: 6 - 8 yrs
• Developing countries: 3-10 years
• Developed countries: >15 years

Place distribution

• Stomach Ca: Japan

• Breast Ca: India

• Cervical Ca: India

• Yellow Fever: Africa

Analytical Epidemiology

• Buzzword → Determinant, Testing of hypothesis

1. Case Control

• To study rare diseases.

• Cases → disease present

• Control → disease absent

 

Steps of a Case-Control Study

• Take SMEAr
1. Selection of cases and control
• For 1 case:
• maximum 4 controls can be taken
• if the cost of collecting cases & control equal
• 1:1 is also sufficient
2. Matching
3. Measurement of Exposure
4. Analysis and interpretation

A landmark Case control study

• Doll and Hill study
• Relation between Lung cancer and smoking

Many children from a particular community coming to a hospital were detected to have acute lymphoblastic leukemia (ALL). It was assumed that it is due to the presence of cytotoxic waste in the water of that community. If a case-control study has to be done to find whether the chemical and ALL are associated, what will be taken as the control?
A. Children from the area exposed, but unaffected with the disease
B. Children from the area not exposed and affected with the disease
C. Children coming to your OPD, who do not have the disease
D. All children with ALL irrespective of exposure status

ANS

Odds Ratio (OR):

• Cross product ratio.
• (a x d) / (b x c)

ㅤ	Disease +	Disease -
Risk factor +	a	b
Risk factor -	c	d

• Mnemonic: Against (X → cross) Odds

• So, if odds ratio is 1.5
• This can be explained as “people with myocardial infarction are 1.5 times more likely to be exposed to smoking than non smokers”
• Because it is a case control study, which starts with outcome.
• DO NOT INTERPRET AS 👉 “people with smoking are 1.5 times more likely to develop mi.”

2. Cohort (Longitudinal) study

• one group includes workers directly involved in handling the dye,

• while the other group consists of office clerks who are not directly exposed to the dye.

• The duration of occupation was recorded from available records.

Relative Risk (RR):

• AKA Risk ratio.

• Incidence in exposed (IE)
  Incidence in non-exposed (INE)
• = a/(a+b)
  c/(c+d)

Interpreting OR & RR:

• ⇒ Confidence interval should not include 1

Example 1

 

✅ Significant Risk Factors
(OR > 1 + Confidence Interval does not include 1):

• Early menarche: OR = 1.3 (1.2–1.9)

• BRCA Mutation: OR = 1.9 (1.1–2.8)

• Obesity: OR = 1.25 (1.1–1.7)

❌ Not significant
(Confidence Interval includes 1 or OR < 1):

• High socioeconomic status: OR = 1.4 (0.5–3.5) → Not significant (CI includes 1)

• Age < 50 yrs: OR = 0.04 (0.01–0.32) → Protective factor

• Smoking: OR = 0.5 (0.02–0.9) → Protective factor

• Alcohol: OR = 1.0 (0.8–1.8) → Not significant

• No family history of Ca. breast: OR = 0.3 (0.01–0.9) → Protective factor

Mixed cohort study

• Bidirectional / Ambispective Cohort Study
• Has both retrospective and prospective components
 

• Example (Study starting in 2023)
• Retrospective:
• Collect past data
• Divide into exposed and non-exposed groups
• Check incidence of skin rash at baseline
• Prospective:
• Follow-up from 2023 → Continue tracking

ANS

• 1 and 2

• 1 → 0.5, 2
  2 → 0.5, 2
  3 → 0, 2

 

Nested Case-Control Study

• Definition
• Smaller case-control study within a larger cohort.
• Start as a cohort study → Develop into case control study
• For rare/expensive tests,
• only analyze cases and a subset of controls

• Example
• Population: 100 children born in Delhi hospitals
• Initial data/specimen collection from mothers:
• Blood samples
• Urine samples
• Stem cells
• Prenatal exposure details
• Study type: Prospective cohort (followed for 50 years)
• Observation
• 10 children developed disease → Cases
• 90 children did not → Eligible pool for controls

• Advantages
• Prospective
• Temporality maintained
• Recall bias eliminated
• Cost-effective
• Limited testing only
• Suitable for rare investigations
• Useful for biological precursor studies
• Efficient:
• No need to test entire cohort

Attributable Risk:

• measures the absolute difference OR disparity in the incidence of a disease between exposed and non-exposed

• IE - INE
  IE

• % of disease d/t risk factor under study among exposed group.

• risk difference

• Community physician/epidemiologist

Population Attributable Risk:

• I(total) - INE
  I(total)

• % decline in disease if risk factor was lowered in general population.

• Provides an estimate of amount by which the disease could be reduced in any population if suspected factor was eliminated or modified

• Used to design health programs.

• Policymakers (smoking cessation program)

IMP: Attributable risk of an exposure refers to which of the following options?

 

1. The female population is at greater risk of developing blindness.

2. People dwelling in rural areas are at increased risk of developing blindness

3. The risk of developing blindness increases with age

4. Illiteracy leads to an increased risk of developing blindness

Note:

• If p value < 0.05 → accept the study

• If odds ratio > 1 → strong association

ANS

 

 

Number Needed to Treat (NNT)

• NNT = Number of patients to treat to prevent one additional bad outcome
(e.g. pain, death, MI, stroke, haemorrhage)

✅ Ischemic stroke prevention example

• ARC = 22 / 100 = 0.22

• ART = 17 / 100 = 0.17

• ARR = 0.22 - 0.17 = 0.05

• NNT = 1 / 0.05 = 20

Experimental Epidemiology

Types:

Randomization:

• 1st step
• Selection (Eg 200 people)
• Not random, so bias prone

• 2nd step
• Allocation into groups.
• Randomisation is done at this stage
• Removes selection bias & confounding factors
• Known and equal chance.

Evidence-Based Medicine

• Research on previous study design.
• Systematic reviews
• Preliminary studies to assess factors, associations, and other variables.
• Meta-analysis (Better)
• Best level of causal association.
• Systematic study + statistical approach.
• To find the final effect of an intervention, or risk factor.
• Approach through forest plot.

Clinical Trials:

Licensing Authority:

• USA: FDA

• India: DCGI (Drug Controller General of India)

Phase	Participants	Key Objectives	Study Design
Preclinical	In animals	• Pre-clinical study CPCSEA approval • Clinic (Preclini) for Animals in Sea (CPCSEA)	ㅤ
Phase 0	10–15 healthy volunteers	• Use Radiolabeled substances • [Pharmacokinetics/ Pharmacodynamics] • "Micro-dosing study." • For Expensive/ toxic drugs • [Conducted on humans] • micrO, radiO, tOxic drugs = 0	• Exploratory Maximum drug amount: 100 mcg or (1/100)th of Human Equivalent Dose, whichever is lower.
Phase I	Healthy volunteers	Determine • Maximum Tolerated Dose (MTD) • Toxicity • NOT Efficacy • Safety • Tolerability • I → T → Toxicity, Tolerability, safeTy, mTd	ㅤ
Phase II	Patients 100–300 patients	• First test of efficacy • Safety, Dose range refinement	• Single blind studies • Max failure
Phase III	Patients (up to 5000)	• Confirm efficacy = Efficacy trial • Compare safety and efficacy ↳ with old standard or placebo drug	• Multicentric • Double blind studies
Phase IV	Post-market patients	• Identify rare and long-term adverse effects	• Post-marketing studies • Blackbox warnings
Phase V	ㅤ	• Pharmaco-epidemiology	ㅤ

• Preclinical study
• CPSEA approval
• Clinic (Preclini) for Animals in Sea (CPCSEA)

• Phase 0
• IND application filing: 
• Submit new experimental drug application to CDSCO
• for a human clinical trial.
• 0 = IND (0 discovered by India)

• Phase 2
• Central Drugs Standard Control Organisation (CDSCO)

• After Phase 3
• NDA filing (3 letter = Phase 3)
• leading to the drug entering the market.

• Blackbox warnings
• Anti depressants = Suicidal effect
• No Suicide risk → Clozapine & Lithium
• Monteleukast = Depression & Suicide

Guidelines for Epidemiological Studies:

Evidence based

1. Systematic Review

• A literature review that:
• Collects data
• Critically analyses multiple research studies
• Includes quantitative synthesis of all studies

• Steps of Systematic Review
• QSER
1. Identifying Questions
• Set eligibility criteria
2. Searching & Selecting studies
3. Extract/Abstract Data
4. Analysis / Result interpretation

2. Meta analysis

• One step ahead of systematic review

• Performs quantitative analysis

• Combines previous research systematically

• Arrives at overall conclusion

• Steps
• QSER
1. Identifying Questions
• Set eligibility criteria
2. Searching & Selecting studies
3. Extract/Abstract Data
4. Analysis / Result interpretation
5. Outcome: Forest Plot

Funnel Plot

• Drawn for:
• Systematic review
• Meta-analysis

• Used to:
• Check publication bias
• Assess study quality

Forest Plot (Blobbogram)

• Only for Meta-analysis

• Visual display of effect sizes across studies

• Horizontal line → Confidence interval

Interpretation

• If Interval includes 1 → Not significant

• If Interval excludes 1 → Significant association

• OR > 1 → Risk factor (positive association)

• OR = 1 → No association

• OR < 1 → Protective factor (inverse association)

• Diamond → Represents overall/pooled estimate
• Example: OR = 2.2 (1.9 - 2.7) → Significant, Risk factor

Levels of Evidence/Hierarchy of Epidemiology:

• Best study: Confirmation of hypothesis
• Evidence based medicine 
• Meta analysis → Best
• Systemic reviews
• Experimental studies

• Test hypothesis
• Cohort studies 
• Case-Control studies

• Unknown disease: To formulate hypothesis
• Case-Series/Reports, Editorials, Opinions 
• Animal studies

Bias, Confounding & Casual Associations

Errors:

• Random errors: Sampling errors.

• Systematic errors: Bias
• Even if we do everything systematically → still error occurs → Systematic error

Bias:

Ecological fallacy:

• Results of an ecological study are not applicable at individual level.

Rx of bias:

• Blinding
• BLINDING → Remove Bias, not confounder
• Randomisation → both B and C (RBC)
• Blinding → Only B (BB)
• Matching → Only C (MC)

• Types
1. Single blinding
• Subject is blinded.
2. Double blinding
• M/C done
• Subject + doctor are blinded.
• Machine used.
3. Triple blinding
• Statistician/analyser + doctor + subjects are blinded.
• Best blinding.
• Researcher → not blind

Confounding

Effect Modifier:

• A variable that changes the strength or direction (positive or negative)
• of the observed effect of a risk factor on disease status

Confounder:

• 3rd variable, unequal distribution.

Criteria for a confounder:

• (All 3 required)
1. Associated with risk factor.
2. Associated with disease.
3. Can lead to disease directly or indirectly.

• Example: Nicotine in smoke

Rx of confounding:

• Randomisation is universal treatment for both selection bias and confounding factor

• BLINDING → Remove Bias not confounder
• Randomisation → both B and C (RBC)
• Blinding → Only B (BB)
• Matching → Only C (MC)

Matching

• Matching means Cases and control are similar to almost all the factors that can affect the outcome of the study, except the disease under study

• Matching eliminates all the known confounders in the study.

Mnemonic:

• Confound → Kalyana alochana
• If know the person → match the person
• If no one is there
• find Standard () Random (m/c) person
• Stratify () → list according to preference → easy (engineers/doctors first) → simplest
• Regress () → cross out bad people → difficult but best thing to do

Study of Choice

Hill's criteria of Causal Association:

• Temporality:
• Most essential criteria.
• RF should precede disease

• Biological plausibility.
• Biological explanation

• Coherence of association.
• Every time, RF should lead to disease

• Validity of association.
• Specificity, sensitivity, OR, RR

• Dose response relationship.
• ↑ Rf α ↑ Disease

• Specificity of association:
• Same RF → Same Disease every time
• most difficult to prove.
• Least essential criteria

Validity:

• Internal:
• Means study gives similar results on repeating
• Measure of robustness of study
• Based on bias → If bias + , Internal validity low

• External:
• It shows generalisability of study
• Study results can be applied/extrapolated to larger population.
• Based on sample size >> bias.

RE α 1 / (Precision α Reproducibility/ Reliability)

• Precisely () random () alkkare kuthiyal reproduce () cheyyam

SE α 1 / (Accuracy α Validity)

• Vaccurat (Validity → Accuracy) systematic (systematic error)

• Validity:
• Hit the board
• Results are within a desired range.

• Accuracy:
• Bull’s eye
• Nearness/closeness to the actual true value.

• Reliability:
• Repeatability
• Out of the board, but same result when repeated
• Dependable, reproducible, repeatable results.

• Note:
• Serial interval:
• Proxy indicator for incubation period.
• CFR:
• Denote virulence of a disease.

Ecological study

• Type: Observational analytical study

• Also called:
• Correlational study
• Geographical study
• Population study
• Aggregate study

Design

• Exposure and outcome assessed simultaneously

• Uses secondary data (already collected data) from private or third parties

• Commonly used in nutritional services

• Represented using scatter diagrams to show correlation

Unit of Study

• Population, not individuals

Limitation

• Ecological fallacy:
• Error of applying population-level results to individuals
• Population trends may not reflect individual-level associations
• Example
• Countries with high daily meat consumption show more colon cancer cases
• But people with colon cancer may not necessarily be meat-eaters

Question based study

Drug	Interpretation
A	Inferior
B	Possibly inferior / Inconclusive
C	Equivalent (crosses 0)
D	Equivalent (clear)
E	Equivalent but borderline superior
F	Likely superior
G	Clearly superior

Understanding the graph:

• X-axis: Effectiveness difference between test drug and comparator.
• 0 = no difference.
• +δ = margin for superiority.
• –δ = margin for inferiority.

• Zone of Equivalency:
• If confidence interval lies inside this zone