Postgraduate-level comprehensive NEET PG notes covering epidemiology, biostatistics, demography, environmental health, national health programs, nutrition, MCH, international health, and NCD control — with 80+ tables, 55+ mnemonics, 40+ PYQs, and 35+ quick-revise summaries.
Epidemiology is the study of disease distribution and determinants in populations — it is the foundation of public health, answering who gets sick, when, where, and why.
Key exam topics:
Epidemiologic triad (host-agent-environment)
Bradford Hill criteria for causality
Prevalence vs Incidence relationship
Most common trap:
Prevalence = Incidence × Duration — high prevalence may mean long disease duration, not high incidence. Pancreatic cancer and diabetes have similar incidence but very different prevalence.
Think of epidemiology as detective work for public health — you are figuring out who gets sick, when, where, and why.
Epidemiology is the study of the distribution and determinants of health-related states and events in specific populations, and the application of this study to control health problems.
The core functions include identifying the causes of disease, measuring disease burden in communities, studying natural history and prognosis, evaluating preventive and therapeutic interventions, and providing the evidence base for health policy.
The epidemiologic triad (host, agent, environment) is the fundamental model of disease causation — disease results from the interaction between a susceptible host, a pathogenic agent, and a conducive environment.
Descriptive Epidemiology
Descriptive epidemiology answers the questions of who (person), when (time), and where (place).
Person variables include age, sex, socioeconomic status, occupation, ethnicity, and behavioral factors. Time patterns describe epidemic curves (point source vs. propagated), secular trends (long-term changes over years), and seasonal variation (e.g., influenza peaks in winter). Place variables include geographic distribution, urban vs. rural settings, and clustering of cases.
The classic example is John Snow's investigation of the 1854 Broad Street cholera outbreak — identifying the contaminated water pump as the source by mapping cases.
Analytic Epidemiology
Analytic epidemiology tests hypotheses about causal relationships between exposures and outcomes.
Observational studies include cross-sectional (prevalence surveys), case-control (retrospective — comparing exposure histories between cases and controls), and cohort (prospective or retrospective — following exposed and unexposed groups forward to compare disease incidence).
Hierarchy of Evidence: Systematic reviews / Meta-analyses → RCTs → Cohort studies → Case-control studies → Cross-sectional studies → Case series → Case reports
The Bradford Hill criteria (strength, consistency, specificity, temporality, biological gradient, plausibility, coherence, experiment, analogy) provide a framework for assessing causality — temporality (exposure must precede the outcome) is the only essential criterion.
PGI Chandigarh 2022: Which of the following is NOT a Bradford Hill criterion? Options: (a) Strength, (b) Specificity, (c) Cost-effectiveness, (d) Temporality. Answer: (c) Cost-effectiveness — it is an economic concept, not a causal criterion.
Bradford Hill criteria mnemonic: "Some Strong Studies Show Consistent Specific Trends — Biologically, Experimentally." Corresponds to: Strength, Specificity, Consistency, Temporality, Biological gradient, Experiment.
Measures of Disease Frequency
Incidence (rate of new cases in a population at risk over a specified time period) and Prevalence (total existing cases at a point in time or over a period).
Prevalence = Incidence × Duration of disease.
Incidence rate (person-time denominator) is the most accurate measure for dynamic populations. Attack rate is the cumulative incidence in a closed population over a short epidemic period (used in outbreak investigations). Standardization (direct and indirect) adjusts rates for differences in population age structures. The standardized mortality ratio (SMR) is the ratio of observed to expected deaths using indirect standardization.
Types of Incidence
Cumulative incidence (CI = number of new cases / population at risk at baseline) — assumes no loss to follow-up. Incidence density (ID = number of new cases / person-time at risk) — accounts for varying follow-up times and produces a rate.
Attack rate is a special type of cumulative incidence during an epidemic in a defined population (e.g., food-borne outbreak). Secondary attack rate = number of new cases among contacts / total contacts.
Incidence and prevalence must be distinguished carefully: a high prevalence does not necessarily mean high incidence — it could reflect longer disease duration (e.g., diabetes has high prevalence due to long survival, while pancreatic cancer has similar incidence but low prevalence due to short survival).
Information Bias and Misclassification
Information bias (observation/measurement bias) arises from systematic errors in measuring exposure or outcome.
Recall bias (differential recall between cases and controls) is a major concern in case-control studies. Observer bias occurs when investigators differentially assess outcomes. Misclassification can be non-differential (random — biases toward null) or differential (non-random — biases in either direction).
Non-differential misclassification of a binary exposure biases the association toward the null, reducing the ability to detect a true effect.
Interviewer bias (differential probing by interviewer aware of hypothesis) and detection bias (systematic differences in how groups are monitored) also distort results.
Aspect
Descriptive Epi
Analytic Epi
Purpose
Describe disease distribution
Test hypotheses about causality
Question
Who? Where? When?
Why? How?
Designs
Cross-sectional, ecological
Case-control, cohort, RCTs
Comparison group
Not required
Essential
Output
Rates, time trends
RR, OR, AR, PAR%
Measure
Formula
Best For
Cumulative Incidence
New cases/Population at risk
Closed populations, short follow-up
Incidence Density
New cases/Person-time
Dynamic populations, variable follow-up
Point Prevalence
Cases at a point/Population
Cross-sectional surveys
Period Prevalence
Cases during period/Mid-period pop
Disease burden over year, planning
Attack Rate
New cases during epidemic/Pop at risk
Outbreak investigation
Common exam confusion: Point vs Period prevalence. Chronic diseases (DM, HTN, asthma) have similar point and period prevalence; acute diseases (influenza, GE) have very different values. NEET PG frequently tests this.
Bias Type
Definition
Example
Recall bias
Differential accuracy in remembering exposures
Mothers of children with birth defects recall more drug use
Observer bias
Investigator differentially assesses outcome
Radiologist reads CXR as abnormal if told patient is smoker
Interviewer bias
Differential probing by interviewer
Asking cases 10 follow-up questions, controls only 2
Detection bias
Outcome more likely detected in one group
Smokers get more chest X-rays - lung cancer more likely detected
Non-differential misclassification
Misclassification independent of outcome/exposure
Faulty glucometer reads high for all participants
Differential misclassification
Misclassification depends on outcome/exposure status
Cases self-report weight more inaccurately
1. Epidemiology = distribution + determinants + control of health problems
2. Epidemiologic triad: Host + Agent + Environment
3. Prevalence = Incidence x Duration; high prevalence can mean long duration, not high incidence
4. Attack rate = cumulative incidence in closed pop during epidemic
5. Secondary attack rate = new cases among contacts / total contacts
6. SMR = Observed deaths / Expected deaths (indirect standardization)
7. Temporality is the ONLY essential Bradford Hill criterion
8. Non-differential misclassification biases toward null
9. Point prevalence (single moment) vs Period prevalence (year-long window)
10. Recall bias = #1 threat in case-control studies
The epidemiologic triad (host-agent-environment) and transmission dynamics of infectious diseases.
Study Designs and Bias in Epidemiology
Cross-sectional Studies
Cross-sectional studies measure exposure and disease simultaneously in a population at a single point in time.
They provide prevalence estimates and are useful for generating hypotheses, assessing disease burden, and planning health services. However, they cannot establish temporality (the chicken-or-egg problem) and are susceptible to survival bias.
The Neyman bias (prevalence-incidence bias) occurs when prevalent cases differ systematically from incident cases — e.g., studying risk factors for myocardial infarction using prevalent survivors misses those who died rapidly.
Case-control Studies
Case-control studies select individuals with the disease (cases) and without the disease (controls), then retrospectively compare exposure histories.
They are efficient for rare diseases and diseases with long latency periods. Selection of appropriate controls is critical — controls should be representative of the population that gave rise to the cases (the study base). Matching (individual or frequency matching) can control for confounding but prevents the study of matched variables as risk factors.
Sources of Bias in Case-control Studies
Selection bias
— control selection is the most vulnerable point (Berkson's bias/hospital admission bias)
Recall bias
— cases remember exposures more accurately than controls
Interviewer bias
The odds ratio (ad/bc) approximates the relative risk when the disease is rare (<10% prevalence) — the rare disease assumption.
Cohort Studies
Cohort studies identify a group of individuals, measure their exposure status, and follow them forward in time to compare disease incidence between exposed and unexposed groups.
Prospective cohort studies follow participants into the future (stronger design, less bias, but time-consuming and expensive). Retrospective cohort studies use existing records (more efficient but subject to data quality limitations).
Cohort studies are the gold standard for establishing incidence, natural history, and multiple outcomes from a single exposure.
Loss to follow-up is a major potential bias (differential loss by exposure and outcome status can severely distort results).
Measures of Association:
1. Relative Risk (RR = Ie / Iu)
2. Attributable Risk (AR = Ie - Iu)
3. Attributable Risk Percent (AR% = [Ie - Iu] / Ie × 100)
4. Population Attributable Risk (PAR% = [It - Iu] / It × 100)
5. Number Needed to Treat (NNT = 1 / ARR)
6. Number Needed to Harm (NNH = 1 / ARI)
Randomized Controlled Trials (RCTs)
RCTs are experimental studies that provide the strongest evidence for causal inference.
Randomization (simple, blocked, stratified) minimizes confounding by distributing both known and unknown confounders equally between groups. Blinding (single, double, triple) reduces performance and detection bias.
Phases of Clinical Trials
Phase I
— safety, dose-finding (20-80 healthy volunteers)
Phase II
— efficacy, dose-ranging (100-300 patients)
Phase III
— comparative efficacy (500-3000 patients, pivotal for approval)
Phase IV
— post-marketing surveillance for rare adverse events
Confounding and Effect Modification
Confounding is a distortion of the exposure-outcome association by a third variable that is independently associated with both the exposure and the outcome and is not on the causal pathway.
Criteria for a confounder: (1) associated with the exposure, (2) independent risk factor for the outcome, (3) not an intermediate step in the causal pathway.
Methods to control confounding: randomization, restriction, matching, stratification (Mantel-Haenszel), multivariable regression analysis.
Effect modification (interaction) occurs when the magnitude of the association differs across levels of a third variable — e.g., the effect of smoking on lung cancer risk differs by asbestos exposure (synergistic interaction).
Systematic Reviews and Meta-analysis
Systematic review uses explicit, reproducible methods to identify, appraise, and synthesize all available evidence on a specific question.
Meta-analysis combines results statistically to produce a pooled estimate.
Forest plots display point estimates and confidence intervals for each study plus the pooled estimate. The diamond at the bottom represents the summary estimate. Heterogeneity (I² statistic: 0-40% might not be important, 50-90% substantial) guides whether random-effects or fixed-effects models are appropriate.
Publication bias is assessed using funnel plots (asymmetric funnel suggests missing studies) and Egger's test.
The GRADE system rates quality of evidence as high, moderate, low, or very low based on study design, risk of bias, inconsistency, indirectness, imprecision, and publication bias.
Feature
Cross-sectional
Case-control
Cohort
RCT
Temporality
Cannot establish
Can infer
Can establish
Establishes
Measure
Prevalence ratio
OR
RR
RR, ARR, NNT
Cost
Low
Moderate
High
Very high
Rare diseases
Not suitable
IDEAL
Needs large n
May be unethical
Multiple outcomes
Yes
No (one disease)
Yes
Yes (limited)
Framingham Heart Study (1948-present): prospective cohort, 3 generations, identified major CVD risk factors. Nurses Health Study (1976): 121,700 women, cancer/CVD risk factors.
NEET PG trap: OR always overestimates RR when disease prevalence > 10%. In hypertension study (prev ~25%), OR=2.5 but actual RR ~1.8.
1. Cross-sectional: prevalence, no temporality, Neyman bias (prevalent vs incident cases)
2. Case-control: OR = ad/bc, ideal for rare diseases, recall bias = major threat
3. Cohort: RR = [a/(a+b)]/[c/(c+d)], gold standard for incidence
4. RCT: randomization controls known + unknown confounders
5. Phase I = safety; Phase II = efficacy; Phase III = comparative; Phase IV = post-marketing
6. Confounding = nuisance to eliminate; Interaction = finding to report
7. I-squared < 40% low heterogeneity; 50-90% substantial (use random-effects)
8. GRADE system: high (RCTs) to very low (observational)
9. OR > RR when disease prevalence > 10%
10. Framingham = prototype prospective cohort
Classification of epidemiologic study designs with common sources of bias and confounding.
Screening, Outbreak Investigation and Infectious Disease Epidemiology
Screening is the presumptive identification of unrecognized disease in apparently healthy individuals using tests that can be applied rapidly.
The Wilson-Jungner criteria (1968) guide the decision to screen: the condition should be an important health problem, its natural history should be understood, there should be a detectable preclinical phase, a suitable and acceptable test should exist, and there should be an effective treatment.
Lead time bias makes screening appear to improve survival even if it does not change the natural history (the extra survival time is just the lead time).
Test Validity
Measure
Formula
Interpretation
Sensitivity
a/(a+c)
Proportion of diseased correctly identified (true positive rate)
Specificity
d/(b+d)
Proportion of non-diseased correctly identified (true negative rate)
PPV
a/(a+b)
Probability a positive test truly indicates disease (falls with decreasing prevalence)
NPV
d/(c+d)
Probability a negative test truly indicates no disease
LR+
sens/(1-spec)
How much a positive test increases probability of disease
LR-
(1-sens)/spec
How much a negative test decreases probability of disease
The Receiver Operating Characteristic (ROC) curve plots sensitivity vs. 1-specificity across different cutoff values — the area under the curve (AUC) is a summary measure of test accuracy (0.5 = no discrimination; >0.9 = excellent).
Additional Screening Biases
Length bias: screening tends to detect slower-growing, less aggressive lesions (longer preclinical phase), while aggressive, fast-growing lesions are more likely to surface clinically between screening rounds.
Selection bias: people who participate in screening tend to be healthier, more health-conscious, and have better outcomes regardless of screening (volunteer bias).
Outbreak Investigation
Outbreak investigation follows a systematic protocol:
Prepare for field work
Establish the existence of an outbreak (compare to expected baseline)
Verify the diagnosis (clinical, lab, epidemiologic confirmation)
Define and identify cases (case definition)
Describe data in terms of person, time, place (epidemic curve)
Develop hypotheses (based on descriptive data and interviews)
Evaluate hypotheses (through analytic studies)
Refine hypotheses and conduct additional studies
Implement control and prevention measures
Communicate findings through a formal outbreak report
Example: An outbreak of gastroenteritis among attendees of a wedding reception. Epidemic curve shows a point-source pattern. Attack rates calculated by food item: chicken (80% among exposed vs 15% among unexposed, RR = 5.3). Salmonella confirmed by stool culture. Control measures: food handler hygiene, proper cooking temperatures, public notification.
Infectious Disease Epidemiology
The basic reproduction number (R0) is the average number of secondary cases generated by a single infectious case in a completely susceptible population. If R0 > 1, the infection can spread; if R0 < 1, it will die out.
The effective reproduction number (Rt) reflects the actual transmission potential accounting for population immunity and interventions. Serial interval is the time between successive cases in a chain of transmission (e.g., COVID-19 serial interval ~4-5 days). Generation time is the interval between infection of primary and secondary cases.
Forest Plots and Publication Bias
Forest plots are graphical displays of individual study results and the pooled meta-analysis estimate.
Each row shows a study's effect size (square) and 95% CI (horizontal line). The pooled estimate is shown as a diamond at the bottom.
Heterogeneity quantified by I² (percentage of total variation due to between-study differences). Funnel plot asymmetry suggests publication bias — small negative studies are missing. The GRADE approach classifies evidence quality: randomized trials start as high quality, observational studies as low quality, then are downgraded or upgraded based on specific criteria.
Publication bias: studies with positive results are more likely to be published, in English, and cited. Funnel plot inspection and Egger's regression test are used to detect it.
Screening Bias
Definition
Effect
Lead time bias
Early detection shifts diagnosis date backward
Overestimates survival benefit
Length bias
Slow-growing lesions overrepresented
Overestimates benefit of screening
Selection bias
Health-conscious attend screening more
Overestimates benefit
Overdiagnosis
Detecting disease that would never cause symptoms
Artificial incidence increase
Screening is SECONDARY prevention - reduces MORTALITY, NOT incidence. Only primary prevention (immunization, lifestyle) reduces new cases.
Epidemic Curve
Shape
Interpretation
Example
Point source
Single sharp peak
Common source, single exposure
Wedding food poisoning
Continuous common
Plateau then decline
Ongoing exposure
Contaminated water supply
Propagated
Multiple peaks, gradual rise
Person-to-person spread
Measles outbreak
NEET PG 2020: A single sharp peak in epidemic curve with all cases within one incubation period suggests: Point source outbreak.
Epidemic curves: Point source = one steep mountain; Propagated = rolling hills; Continuous = flat mesa.
Disease
R0
Herd Immunity
Incubation
Measles
12-18
92-95%
10-14 days
COVID-19
2-3
50-67%
2-14 days
Influenza
1.3-1.8
23-44%
1-4 days
Pertussis
15-17
93-94%
7-14 days
Polio
5-7
80-86%
7-14 days
Diphtheria
5-6
80-83%
2-5 days
Mumps
10-12
90-92%
16-18 days
Rubella
5-7
80-86%
14-21 days
1. Screening = SECONDARY prevention; reduces mortality, not incidence
2. Sensitivity = a/(a+c); Specificity = d/(b+d) - both INTRINSIC properties
3. PPV increases with prevalence; NPV decreases as prevalence increases
4. LR+ >10 = good rule-in; LR- <0.1 = good rule-out
5. Lead time bias: survival appears longer but death isn't delayed
6. Length bias: slow-growing lesions detected preferentially
7. R0 >1 = epidemic; Herd immunity threshold = 1-1/R0
8. Point source: single peak within one incubation period
9. Forest plot diamond = pooled estimate; funnel plot asymmetry = publication bias
10. Overdiagnosis: detecting disease that would never cause symptoms/lifetime death
Screening test validity parameters (sensitivity, specificity, PPV, NPV) and the systematic outbreak investigation protocol.
Confounding, Effect Modification and Measures of Association
Measures of Association — Key Formulas
Measures of association quantify the statistical relationship between an exposure and an outcome in epidemiologic studies.
2×2 Contingency Table
Disease +
Disease −
Total
Exposed +
a
b
a+b
Exposed −
c
d
c+d
Total
a+c
b+d
N
Relative Risk (RR) = [a/(a+b)] / [c/(c+d)] — used in cohort studies; RR > 1 = positive association (exposure increases risk); RR < 1 = negative association (protective); RR = 1 = no association.
Odds Ratio (OR) = ad/bc — used in case-control studies; approximates RR when disease is rare (prevalence < 10%); OR > RR when disease is common.
Attributable Risk (AR) = Ie − Iu — absolute excess risk in the exposed; useful for clinical decision-making (how much risk is added by the exposure?).
Attributable Risk Percent (AR%) = (Ie − Iu)/Ie × 100 — proportion of disease in the exposed that is attributable to the exposure.
Population Attributable Risk (PAR%) = (It − Iu)/It × 100 — proportion of disease in the entire population attributable to the exposure; depends on both the strength of association AND the prevalence of exposure in the population; highest-impact measure for public health interventions.
Exam trap: PAR% is NOT the same as AR%. PAR% takes exposure prevalence in the population into account — a weak association with a very common exposure can have a higher PAR% than a strong association with a rare exposure.
Confounding
Confounding occurs when a third variable (the confounder) is associated with both the exposure and the outcome, and is not in the causal pathway — thereby distorting the true relationship between exposure and outcome.
Criteria for a confounder: (1) associated with the exposure in the study population, (2) independently associated with the outcome, (3) NOT an intermediate variable in the causal pathway.
Methods to Control Confounding
Stage
Method
Notes
Design
Randomization
Only randomization controls UNKNOWN confounders
Design
Restriction
Limits eligibility to one level of the confounder
Design
Matching
Controls confounders but prevents studying matched variable as risk factor
Analysis
Stratification (Mantel-Haenszel)
Computes stratum-specific and summary adjusted estimates
Analysis
Multivariable regression
Adjusts for multiple confounders simultaneously
Analysis
Propensity score
Balances covariates in observational studies
Effect Modification (Interaction)
Effect modification (interaction) occurs when the magnitude of the association between exposure and outcome differs across levels of a third variable (the modifier) — this is a biological phenomenon to REPORT, not remove.
Key distinction: Confounding is a bias to be eliminated; effect modification is a finding to be described. A Mantel-Haenszel test for heterogeneity
Confounder criteria - "ACE": Associated with exposure, Cause (risk factor) for outcome, Excluded from causal pathway.
NEET PG 2023: Only randomization controls BOTH known and unknown confounders.
1. RR = cohort; OR = case-control (approx RR if disease rare)
2. OR > RR when disease prevalence > 10%
3. AR = absolute excess; AR% = proportion in exposed; PAR% = population burden
4. Confounder: associated BOTH exposure and outcome, NOT in causal pathway
5. Randomization = only method for unknown confounders
6. Effect modification = report stratum-specific ORs; confounding = eliminate it
7. PAR% depends on RR AND exposure prevalence
8. NNT = 1/ARR; NNH = 1/ARI
9. Mantel-Haenszel OR = pooled adjusted OR across strata
10. Breslow-Day test p<0.05 = significant interaction present
(Breslow-Day test) detects whether stratum-specific estimates differ significantly — if yes, interaction is present and a pooled estimate is inappropriate.
PYQ — An OR of 2.5 (95% CI: 1.8–3.4) for smoking and lung cancer means: Answer: Smokers have 2.5 times the odds of developing lung cancer compared to non-smokers; the CI excludes 1.0, so the result is statistically significant at the 0.05 level.
1. RR = cohort studies; OR = case-control studies (approximates RR if disease rare)
2. OR > RR when disease prevalence is high (> 10%)
3. AR = absolute excess risk; AR% = proportion in exposed; PAR% = population burden
4. Confounder: associated with BOTH exposure and outcome, NOT in causal pathway
5. Randomization controls unknown confounders; regression controls measured ones
6. Effect modification = report it (stratum-specific ORs); confounding = eliminate it
Framework for understanding measures of association (RR, OR, AR, PAR%) and distinguishing confounding from effect modification.
Immunization, chemoprophylaxis, health promotion, condom use
Secondary
Early detection of preclinical/early clinical disease
Pap smear, mammography, FOBT, BP/cholesterol screening
Tertiary
Reduce impact of established disease, prevent complications
ACE inhibitors in heart failure, cardiac rehab, stroke rehab
Quaternary
Protect from unnecessary/excessive medical interventions
Avoiding overdiagnosis, de-prescribing in elderly
Primordial prevention is the most upstream and cost-effective form of prevention but requires political will and intersectoral action.
The Ottawa Charter for Health Promotion (1986) identified five action areas: build healthy public policy, create supportive environments, strengthen community action, develop personal skills, and reorient health services.
Immunization is one of the most cost-effective public health interventions
— the Expanded Programme on Immunization (EPI) targets tuberculosis (BCG), diphtheria, pertussis, tetanus (DPT), polio (OPV/IPV), measles, hepatitis B, Hib, rotavirus, and pneumococcal disease.
Herd immunity is the indirect protection of unvaccinated individuals when a high proportion of the population is immune.
Vaccine efficacy = (attack rate in unvaccinated - attack rate in vaccinated) / attack rate in unvaccinated × 100. Vaccine effectiveness measures real-world performance.
Secondary prevention: Screening programs are the main tool.
Examples include Pap smear for cervical cancer (reduced incidence and mortality by 70-80%), mammography for breast cancer (reduces mortality by 20-30% in women aged 50-69), and FOBT/colonoscopy for colorectal cancer.
Primary Prevention Strategies in Detail
Health promotion: education, lifestyle modification (diet, physical activity, smoking cessation), and environmental modifications (clean air, water, sanitation). Specific protection: immunization, chemoprophylaxis (isoniazid for TB contacts, hydroxychloroquine for malaria), use of protective equipment (seatbelts, helmets, condoms), and occupational safety measures.
The concept of the "prevention paradox" (Rose) states that a preventive measure that brings large benefits to the community may offer little benefit to each participating individual.
Level
Timing
Disease Stage
Target
Responsibility
Primordial
Before risk factors
No disease, no RFs
Whole population
Government, policy
Primary
Before biological onset
RFs present, no disease
At-risk populations
Health system
Secondary
Pre-clinical
Disease present, asymptomatic
Screening target groups
Screening programs
Tertiary
After diagnosis
Established disease
Patients
Clinicians, rehab
Quaternary
During medical care
Any stage
At-risk of over-treatment
Clinicians
AIIMS Nov 2019: Ottawa Charter addresses which level? Primordial prevention (upstream determinants).
Ottawa Charter: "Policy, Places, People, Personal skills, Providers."
1. Primordial = upstream (poverty, education, tobacco ad ban)
2. Primary = prevent onset (immunization, chemoprophylaxis, condoms)
3. Secondary = early detection (Pap, mammography, BP screening)
4. Tertiary = reduce complications (rehab, ACEi in HF)
5. Quaternary = protect from over-medicalization (deprescribing)
6. Herd immunity = 1-1/R0; measles threshold 92-95%
7. Vaccine efficacy = (AR unvacc - AR vacc)/AR unvacc x 100
8. Rose prevention paradox: population benefit, minimal individual benefit
9. Only randomization controls unknown confounders
10. Active vs passive immunization: active = own antibodies; passive = pre-formed
The five levels of prevention showing the spectrum from primordial (upstream determinants) to quaternary (avoiding overmedicalization).
Epidemiological Surveillance and Notification
Surveillance is the systematic, ongoing collection, analysis, interpretation, and dissemination of health data for the purpose of planning, implementing, and evaluating public health practice.
Symptom/syndrome data preceding confirmed diagnosis
Early warning, real-time
Low specificity, false alarms
Event-based surveillance
Unstructured reports, media, social media
Rapid detection, complements indicator-based
Verification needed, rumor risk
IDSP (Integrated Disease Surveillance Programme) — India
IDSP (2004) is India's disease surveillance system under NHM. Structure: Central Surveillance Unit (NCDC, Delhi) → State Surveillance Unit (SSU) → District Surveillance Unit (DSU) → Community level. Reporting: S-form (syndromic — by health workers, weekly), P-form (presumptive — by MO, weekly), L-form (laboratory confirmed — by labs, weekly). 22 reportable diseases/conditions under IDSP.
Outbreak threshold: any case of AFP, suspected measles, diphtheria, cholera, plague, dengue, JE, leptospirosis. IDF (Immediate Disease Notification): within 24 hours for VPDs (vaccine-preventable diseases) and epidemic-prone diseases. IHIP (Integrated Health Information Platform) — digital platform replacing paper-based IDSP; real-time, GIS-enabled, individual case-based reporting.
Health Information Systems in India
HMIS (Health Management Information System): web-based monitoring of NHM programs; facility-level data entry; monthly reporting from all public health facilities. HIMS portal generates performance scorecards. NFHS: periodic household survey (DHS model). SRS (Sample Registration System): continuous dual-record system for vital rates. CRS (Civil Registration System): registration of all births and deaths under RBD Act, 1969. Target: 100% registration of births and deaths.
System
Frequency
Type
Data
IDSP
Weekly (S/P forms)
Surveillance
Disease outbreaks, 22 conditions
HMIS
Monthly
Facility-based
Program indicators, service delivery
SRS
Continuous / Annual
Dual-record survey
Birth rate, death rate, IMR, MMR
CRS
Continuous
Registration
Births, deaths, cause of death
NFHS
~5 years
Household survey
Health, nutrition, population indicators
IHIP
Real-time
Digital surveillance
All IDSP diseases + additional
NEET PG 2021: The weekly reporting format for syndromic surveillance under IDSP is called: S-form (Syndromic). P-form for presumptive, L-form for laboratory-confirmed. Immediate reporting (within 24 hours) is required for suspected outbreaks.
1. Passive surveillance: routine provider reporting; cheap but underreporting
2. Active surveillance: health authority seeks cases; accurate but expensive
3. Sentinel: selected sites; quality data for trends
4. IDSP: 22 reportable diseases; S (syndromic), P (presumptive), L (lab) forms
5. IHIP: real-time digital platform replacing paper IDSP
6. HMIS: monthly facility-level NHM program reporting
7. SRS: continuous dual-record for vital rates (CBR, CDR, IMR, MMR)
8. CRS: civil registration under RBD Act 1969; target 100% birth/death registration
9. Outbreak: cases exceeding expected baseline (2 SD above mean for endemic diseases)
10. IDF: immediate disease notification within 24 hours for VPDs
Types of surveillance including passive, active, sentinel, and India's IDSP/IHP digital surveillance platforms.
