General Epidemiology

📚 04 PSM & Preventive Pediatrics • 🎯 39 High-Yield MCQs • Nelson 22nd Edition Standard

Research Methodology & Study Design Hierarchy

  • Hierarchy of Clinical Research & Gold Standard: ((EPID001)): At the pinnacle of the evidence pyramid lies the systematic meta-analysis, which combines outcomes from multiple studies to yield the most reliable and statistically precise conclusions.
  • Reporting Guidelines: ((EPID007)): The MOOSE guidelines are specifically designated for standardizing the reporting and execution of meta-analyses for observational studies (such as cohort and case-control studies). In contrast, CONSORT applies to Randomized Controlled Trials (RCTs).

Epidemiological Study Designs & Associated Biases

  • Cohort vs. Case-Control Studies: ((((EPID003)) · ((EPID028)) · ((EPID031)) · ((EPID032)))):
  • Cohort Studies: proceed from exposure to disease (prospective/concurrent or retrospective/historical), are superior for rare exposures, but are more expensive, require long follow-up periods, and are not ideal for rare diseases.
  • Case-Control Studies: proceed from disease to exposure, are cost-effective, yield rapid results, and are the design of choice when the disease under investigation is rare.
  • Recall Bias in Retrospective Designs: ((EPID002)): Recall bias is a systematic error characteristic of retrospective studies (like case-control designs) where subjects with the outcome (cases) differentially and more vividly recall past exposures compared to controls. This distortion is eliminated by utilizing a prospective design.
  • Randomized Controlled Trials (RCTs) & Attrition: ((((EPID004)) · ((EPID005)) · ((EPID035)))): - In RCTs, subjects are selected based on inclusion/exclusion criteria before undergoing randomization (which allocates subjects into intervention vs. control arms to ensure baseline comparability). - Dropouts must not be excluded from the final analysis; doing so introduces attrition bias and violates the Intention-to-Treat (ITT) principle. All subjects must be analyzed in the groups to which they were originally randomized.
  • Internal Validity: ((EPID008)): Internal validity defines the degree to which a study's results are free from bias and confounders, dictating how accurately the outcome can be attributed to the experimental intervention. External validity, conversely, addresses generalizability to other populations.

Causality & Epidemiological Terminology

  • Criteria for Causation: ((EPID009)): Temporality (proving that the exposure precedes the onset of the disease) is the single most difficult criterion to establish in chronic illnesses with long latent periods, yet it is the only mandatory/essential Bradford Hill criterion for causation.
  • Infectious Disease Intervals & Proxies: ((((EPID010)) · ((EPID011)) · ((EPID012)) · ((EPID013)) · ((EPID014)))):
  • Incubation Period: : Time from exposure to symptom onset. Useful for calculating quarantine duration and distinguishing primary from secondary cases. The period of communicability guides isolation and cannot be used as a proxy for the incubation period.
  • Generation Time: : Time interval between receipt of infection by a host and the moment of maximum infectivity of that host.
  • Paradoxical Carrier: : An asymptomatic carrier who acquires infection from another carrier.
  • Droplet Nuclei Transmission: ((EPID012)): Small particles (<5 µm) capable of long-range airborne transmission (seen in Measles, Diphtheria, Pertussis), not involved in fecal-oral pathogens like Typhoid.
  • Immunity Concepts: ((((EPID015)) · ((EPID016)))):
  • Herd Immunity: : Community-level protection when a critical proportion of a population is immune; it is significantly more effective for infections featuring a high sub-clinical/asymptomatic phase and does not protect individuals against non-communicable or environmental pathogens (e.g., tetanus).
  • Premunition: : A state of active individual immunity that persists only as long as the sub-clinical/low-grade infecting organism remains present in the host; it does not confer herd protection.

Levels of Prevention

  • Primary Prevention Target & Measures: ((((EPID017)) · ((EPID018)) · ((EPID020)))): Primary prevention aims to avert the initial occurrence of disease among apparently healthy individuals or those with risk factors via health promotion and specific protection. - Examples include prospective screening (e.g., screening blood donors or healthy at-risk groups) and wearing helmets/seatbelts during transit. - Asymptomatic patients (who already have subclinical disease pathology) are targets for secondary prevention (early detection and prompt treatment).

Biostatistics & Epidemiological Measures

  • Outbreak Metrics & Rates: ((((EPID019)) · ((EPID022)) · ((EPID023)) · ((EPID024)) · ((EPID025)) · ((EPID027)))):
  • Median Incubation Period: : The exact time required for 50% of exposed cases to manifest clinical symptoms.
  • Neonatal Mortality Rate (NMR): : Cannot be calculated unless neonatal deaths (0–28 days of life) are explicitly separated from total infant deaths.
  • Case Fatality Rate (CFR): : $\frac{\text{Deaths from disease}}{\text{Total cases diagnosed}} \times 100$ (measures virulence/severity).
  • Attack Rate: : A specialized incidence proportion applied to closed populations during short outbreaks: $\frac{\text{Ill persons exposed}}{\text{Total persons exposed}} \times 100$.
  • Secular Trends: ((EPID026)): A gradual, long-term progressive change in disease frequency over years or decades, driven by permanent shifts in lifestyle, urbanization, or environment.
  • Point-Source Epidemics: ((EPID027)): Characterized by an epidemic curve with a sharp rise and rapid fall, clustering of cases within a single incubation period, and an absence of person-to-person transmission.
  • Measures of Association & Impact: ((((EPID029)) · ((EPID030)) · ((EPID033)) · ((EPID034)))):
  • Odds Ratio (OR): : Used in case-control studies. Calculated via cross-product ($AD/BC$) of a 2x2 contingency table. An $\text{OR} > 1$ denotes a positive association/risk factor.
  • Relative Risk (RR): : Ratio of incidence in exposed ($\text{I}_e$) to unexposed ($\text{I}_u$) groups: $\text{RR} = \frac{\text{I}_e}{\text{I}_u}$.
  • Attributable Risk (AR / AR%): : The proportion of disease among the exposed that can be directly attributed to the exposure: $\text{AR}\% = \frac{\text{I}_e - \text{I}_u}{\text{I}_e} \times 100$.
  • Population Attributable Risk Percent (PAR%): : Measures the reduction in disease incidence expected in the entire population if the risk factor were completely removed.

Epidemiology & Biostatistics

  • Confounding Factor: ((EPID036)): A hidden variable that influences both exposure and outcome, distorting the true association. E.g., concluding beta-carotene protects against colon cancer when the true protective agent is high-fibre intake inherent to a vegetarian diet.
  • Kappa Statistic ($\kappa$): ((EPID037)): Measures inter-observer or inter-instrument reliability/agreement beyond chance for categorical data. Range: -1 to +1 ($\kappa = 1$ perfect, $\kappa = 0$ chance agreement, $\kappa < 0$ less than chance). Alternate options like Sensitivity, Specificity, and P-value evaluate validity or statistical significance, not observer concordance.
  • Serial Interval: ((EPID038)): The time gap between the onset of symptoms in a primary case and a secondary case (e.g., measles onset Jan 10 to Jan 17), used to estimate transmission dynamics and outbreak spread.
  • Secular Trend: ((EPID039)): Long-term changes in disease occurrence over years/decades driven by environmental, socioeconomic, or healthcare shifts, best visualized using a Line diagram (continuous data on a chronological scale). Alternate graphical tools: Bar graphs (discrete/categorical data), Box and whisker plots (distribution and median/quartiles), Stem-leaf plots (individual numerical data distribution).