CM7.1-11 | Epidemiology Methods — Graded Quiz

Correct. All standard definitions (Last, Park) include 'specified populations' — epidemiology is inherently population-based.

Last's definition: 'study of the distribution and determinants of health-related states or events in specified populations.' Population specification is non-negotiable.

Epidemiology's defining feature is its focus on specified populations, not individuals or experimental designs.

Correct. Determining the natural history and prognosis of disease in a population is a core use of epidemiology — it requires population-level follow-up data, not individual clinical care.

Park lists uses of epidemiology: (1) community diagnosis, (2) study of causation, (3) planning health services, (4) evaluation of interventions, (5) identification of syndromes, (6) natural history of disease.

Epidemiology's population-level uses include: community diagnosis, determining natural history, identifying risk factors, evaluating interventions, planning health services. Treatment of individuals is clinical medicine.

Correct. The chain runs: Infectious Agent → Reservoir → Portal of exit → Mode of transmission → Portal of entry → Susceptible host.

Every control measure targets one or more links: killing the agent, eliminating the reservoir, blocking exit/entry portals, interrupting transmission, or increasing host resistance (vaccination).

The 6-link chain must flow logically: the agent lives in a reservoir, exits through a portal, travels via a mode of transmission, enters the new host through a portal of entry, and infects a susceptible host.

Correct. Isolation applies to confirmed cases (those with active, communicable infection). Quarantine applies to exposed contacts who may be in the incubation period. Confirmed measles infection would trigger isolation.

Duration of quarantine = incubation period of the disease (measles: up to 21 days). Duration of isolation = communicable period of the disease.

Quarantine = exposed contacts during incubation period. Isolation = confirmed infectious cases during communicable period. The trigger for switching is confirmed disease.

Correct. Prevalence = Incidence × mean duration. Long disease duration accumulates cases over time, so prevalence is much higher than incidence. (High cure rate would eventually lower prevalence but is contradicted by 'long duration'.)

The relationship: P ≈ I × D (P = prevalence, I = incidence, D = mean duration). Short duration or high fatality → prevalence ≈ incidence. Long duration (chronic diseases like diabetes, hypertension) → prevalence far exceeds incidence.

Prevalence = Incidence × mean duration (for steady-state diseases). Long duration means cases accumulate, making prevalence >> incidence.

Correct. MMR = (Number of maternal deaths / Number of live births) × 100,000. India's MMR (SRS 2018-20) was 103 per 100,000 live births.

MMR uses live births (not total pregnancies) because live births are more accurately registered than all pregnancies. It does NOT include stillbirths or abortions in the denominator.

The denominator for MMR is live births (not total pregnancies, not women of reproductive age). Multiplier is 100,000.

Correct. Cross-sectional studies measure prevalence (existing cases at a point in time). The correct measure of association is the prevalence ratio (PR) = prevalence in exposed / prevalence in unexposed = 40%/20% = 2.0.

Study design determines the measure of association: Cohort → RR; Case-control → OR; Cross-sectional → Prevalence Ratio (PR). PR here = 0.40/0.20 = 2.0, meaning obese individuals have 2× the prevalence of hypertension.

RR requires incidence (longitudinal follow-up, which cross-sectional studies lack). The appropriate measure for cross-sectional data is prevalence ratio.

Correct. Berkson's bias (hospital admission bias) occurs when both cases and controls are drawn from hospitalised patients, and the exposure-disease relationship among hospitalised patients differs from the general population. Orthopaedic patients advised healthy diets are an unrepresentative control group.

To avoid Berkson's bias, community-based controls are preferred in hospital-based case-control studies. Recall bias is an information bias (differential reporting of exposure). Neyman bias = prevalence-incidence bias (prevalent cases differ from incident cases in exposure history).

Using hospitalised controls who are unrepresentative of the general population creates Berkson's (hospital admission) bias — a type of selection bias.

Correct. Wilson-Jungner criteria (WHO 1968) require: the condition is an important health problem, natural history is understood, there is a recognisable latent stage, acceptable treatment is available, and the test is simple, safe, and acceptable.

Key Wilson-Jungner criteria: (1) Important public health problem, (2) Known natural history, (3) Recognisable latent stage, (4) Effective treatment, (5) Suitable test — safe, simple, acceptable, (6) Cost-benefit acceptable.

Wilson-Jungner criteria for screening include: important health problem, understood natural history, recognisable latent/early stage, accepted treatment, suitable test, agreed policy on who to treat, acceptable cost-benefit. Rarity and experimental treatment are disqualifiers.

Correct. SAR = (Secondary cases) / (Susceptible contacts of primary cases) × 100. It measures the probability of transmission from primary to secondary cases — an index of communicability.

A high SAR (e.g., measles SAR 75-90%) means the pathogen is highly communicable. A low SAR (e.g., cholera SAR ~15%) means most contacts do not develop disease. SAR guides quarantine decisions for household contacts.

Primary attack rate: cases/exposed population. Secondary attack rate: secondary cases/susceptible contacts of primary cases. SAR is an index of how easily the pathogen spreads person-to-person.

Correct. Inferring individual-level causation from group-level (aggregate) data is the ecological fallacy. Countries with high chocolate consumption may differ from low-consumption countries in many other ways (wealth, healthcare access, life expectancy) — and the individuals who consume chocolate may not be the ones without dementia.

Ecological studies use aggregate data (rates, means per country/district) and are hypothesis-generating only. They cannot establish individual-level causation because of the ecological fallacy and the inability to link individual exposure to individual outcome.

The ecological fallacy: associations observed at the group (country) level cannot be automatically attributed to individuals. The confounding by wealth, healthcare, and other country-level factors makes this inference invalid.

Correct. PICO: Population = diabetic patients 40-60 years (as defined in the question); Intervention = daily aspirin; Comparator = no aspirin or placebo; Outcome = cardiovascular events (not a surrogate like blood pressure).

PICO is the structured format for clinical questions: P (Population/Patient), I (Intervention), C (Comparator), O (Outcome). A well-formed PICO guides the literature search and defines what counts as relevant evidence.

PICO: P = the specific patient/population group, I = the intervention studied, C = the comparator, O = the clinically meaningful outcome. The population should match the study's inclusion criteria.