Epidemiology and Public Health Flashcards

Flashcards to review public health and epidemiology concepts.

Epidemiology

A method of reasoning about disease that deals with biological inferences derived observations of disease phenomena in population groups.

Public Health

The community effort to protect, maintain, and improve the health of a population by organized means, including preventative programs, hygiene education, and other interventions.

Primary prevention

Interfering before health effects occur, through measures such as vaccinations, altering risky behaviors, and banning substances known to be associated with a disease or health condition.

Secondary prevention

Screening to identify diseases in the earliest stages, before the onset of signs and symptoms, through measures such as mammography and regular blood pressure testing.

Tertiary prevention

Managing disease post diagnosis to slow or stop disease progression through measures such as chemotherapy, rehabilitation, and screening for complications

Prevalence Count

Number of cases divided by total population.

Cohort

a group of people usually followed through time

closed cohort: start following everyone at the same time

open cohort: may add people over time, may lead to a larger or smaller follow up over time

Prevalence

the description of the population with existing disease (or exposure) at a single point in time (or over a period)

Fraction

= ratio, x/y

Proportion

x/(x+y)

Odds

x/y

Rate

x/N * 1/t

Prevalence proportion

n cases / total population

Prevalence odds

prevalence proportion / 1-prevalence proportion

n cases / population - n cases

Incidence count

number of new cases during a fixed period

Incidence proportion (or risk/cumulative incidence)

new cases during a fixed period divided / population at risk during a fixed period

Incidence odds

incidence proportion / 1 - incidence proportion

Incidence rate (density)

new cases during a fixed period / person-time at risk during a fixed period

Person-time

amount of time observed for all people under study while at risk of experiencing an incident outcome

Measures of association

use of the fundamental measures of occurrence to obtain derivative measures that aid in quantifying relationships between exposure and disease

Risk difference (RD)

risk exposed / risk non-exposed

Incidence rate difference (IRD)

IR exposed - IR non-exposed

Relative effect

RD/risk non-exposed = (risk exposed / risk non-exposed) - 1 => risk exposed/risk non-exposed = risk ratio (RR)

Risk ratio (RR)

risk exposed / risk non-exposed

Incidence rate ratio

IR exposed / IR non-exposed

Odds ratio

Odds exposed / odds non-exposed

Causality

causality refers to the relationship between causes and effect

Causal interference

aims to determine whether an exposure leads to a particular outcome

the 9 viewpoints for causality

• strength of association

• consistency of association

• specificity of association

• temporality

• biological gradient

• plausibility

• coherence

• experiment

• analogy

Confounding

A variable that is causally associated with the outcome and associated with the exposure, but it is not an intermediate variable in the causal pathway between expose and outcome

Mediator

a step in the pathway between exposure and outcome, it explains how or why an exposure leads to an outcome

Selection bias

occurs when the selection of participants distorts the measure of association between exposure and outcome

Information bias

results from either imperfect definition of study variables or flawed data collections

• exposure identification bias: recall bias and interview bias

• outcome identification bias: observer bias and respondent bias

Validity

the ability of a test to distinguish between who has a disease and who does not

reliability (repeatability)

the extent to which the results obtained by a test that are replicated if the test is repeated

sensitivity

how good the test is in correctly identifying those who had the disease

= TP/(TP+FN)

specificity

how good the test is in correctly identifying those who did not have the disease

= TN/(FP+TN)

positive predictive value (PPV)

the probability that someone has the disease given a positive test result

=TP/(TP+FP)

negative predictive values (NPV)

the probability that someone does not have the disease given a negative test result

=TN/(FN+TN)

Randomized clinical trial

randomized, double-blind and (placebo) controlled, it has the most scientific value and are generalizable

advantage: randomization tends to produce comparable groups, reduce bias and produces valid statistical tests

disadvantage: generalizability issues due to volunteer effects and recruitment challenges and administrative complex