CH114 Study Guide

epidemiology definition

the study of the distribution and determinants of disease frequency in human populations and the application of this study to control health problems

An idea central to epidemiology is that

good and bad health events do not happen to people randomly

distribution of disease mechanisms

person (age, sex, SES)

place (urban, US, near an incinerator)

time (june, annual, decade)

determinants definition

factors that cause some people to get a disease

determinants are also called

exposures or risk factors

frequency

counting the number of cases of a disease in a population over (or at) a specific time period

populations

a group of people with a common characteristic (or many common characteristics) such as residence, age, sex, group membership, etc.

why is it important to apply studies to control health problems?

there is no point to collecting and analyzing data if it cannot be used to help control health problems

sequence of epidemiologic investigations

suspect that exposure influences disease occurrence → form specific hypotheses about exposure-disease association → conduct epidemiologic studies → judge whether association is valid and causal → evaluate preventions + treatment →

we operate under what type of hypothesis when doing research

null hypothesis

null hypothesis

the idea that there is no association between the variables being studied

t/f: epidemiology is concerned with individuals

false; it is involved with large populations

4 branches of public health

policy, scientific, clinical, activist

public health triangle involves which branches of public health

policy, scientific, and clinical

Epidemiologic units of study

person, place, and time

primary level of prevention in public health

actual prevention of a specific outcome

secondary level of prevention in public health

identification of an outcome at an early stage in its natural history, leading to improved outcomes

tertiary level of prevention

the limiting of mortality, morbitidy, and/or disability of the specific outcome after it has occured

example of secondary level of prevention

detecting cancer at stage I

example of teriary level of prevention

EMS care after an accident before arriving at the hospital

basic causal theory

A framework for understanding the relationships between causes, conditions, and health outcomes in disease prevention.

objective causality

a theory that can be objectively tested by people other than the person proposing it

subjective causal theory

a theory that cannot be proven or disproven by assessing data

3 types of cause

necessary, contributing, and sufficient

necessary cause

a cause that must be present for something to develop in a causal manner

example of necessary cause

exposure to HIV is necessary to develop AIDS

contributing cause

a cause that increases/decreases the risk of development of an outcome, but is NOT necessary

contributing cause example

flooding after heavy rains in areas with poor sewage systames can cause outbreak fo diarrheal disease

• flooding is not a necessary cause for diarrhea, but it increases the risk significantly.

sufficient cause

all contributing and necessary causes needed to bring about a certain outcome in a given individual are present

sufficient cause example

exposure to influenza is a necessary, but NOT sufficient cause

• with a weakened immune system, poor nutrition, no vaccination, etc., influenza will likely develop

rothman’s causal pies

a visual representation of causal factors, showing sufficient and necessary causes as a combination of elements within pieces of a pie.

Hill’s 6 Causal Criteria

• strength of association

• consistency of the association

• temporal sequence of the association

• dose-response association

• analogy

• biologic plausibility

strength of association

the stronger the association, the more likely it is to be causal

consistency of association

if association appears very strong in one study but weak in another study, it is less likely to establish causality

temporal sequence of association

if X is the cause of Y, X must occur before Y in time.

dose-response association

causality is more likely if an increase in exposure leads to a corresponding increase in effect.

analogy

when a similar agent causes disease, the agent at hand can be more suspect

smallpox vaccine

• Edward Jenner, while still a medical apprentice, noticed
  that milkmaids who had contracted a disease called
  cowpox, which caused blistering on cow’s udders, did
  not catch smallpox.

• The milkmaids sometimes reporting having smallpox as
  a child, but then they tended to not get cowpox.

• Jenner reasoned that cowpox was caused by something
  similar to smallpox and thus infection by one seemed to
  make the person immune to the other.

• Unlike smallpox, which caused severe skin eruptions,
  dangerous fevers, and substantial mortality in humans,
  cowpox led to few symptoms in these women.

• On May 14, 1796, Jenner took fluid from a cowpox
  blister and scratched it into the skin of James Phipps, an
  eight-year-old boy. Blisters rose up on the scratched spot
  (cowpox), but James soon recovered.

• On July 1, Jenner inoculated the boy again, this time pus
  matter from a smallpox patient, and no disease
  developed. The vaccine was a success

biologic plausibility of the association

the relationship between an exposure and an effect is supported by scientific evidence, making the observed association credible.

4 components for measuring disease frequency

population, cases of disease, size of population, time

population can be called

reference population, base population, or source population

source population

the commonality between the population individuals is a specific location

fixed population

a population whose membership is defined on the bases of some event; permanent membership

dynamic population

a population whose membership is defined by being in a state or condition; membership can be transient

#cases of frequency is always the numerator/denominator of all frequencies

numerator

disease definition

any health outcome (positive OR negative)

2 types of incidence

cumulative (risk) or person-time incidence

if incidence is said without specificity, what are they referencing

generally referring to cumulative incidence

subjects used to calculate point prevalence

individuals who have the disease at a specific point in time

subjects used to calculate incidence

subjects that do not have the disease at the baseline and are followed over time to see if they develop the disease.

2 ways to monitor subjects when calculating P or I

• follow them for specific amount of time

• follow for different periods (usually done if subjects start the study at different times)

when following individuals for a specific amount of time, how do you caluclate CI

CI/R = # individuals who develop the condition / # individuals in the population at the start of the study period

critical assumption regarding CI

All people in the population have been followed for the entire specified time period. If anyone has not, they cannot be used in a risk calculation.

what does CI / risk tell us

the probability of an individual of getting a condition over a certain period of time

if you follow up on subjects over specific periods, how do you calculate incidence rate (IR) / person-time incidence

IR = # new cases of a condition during a specified time period / total person-time of observation during the given time period

fundamental to calculating incidence rate is

person-time

person-years

a measure combining the number of people and the amount of time each person is observed, used to express the total amount of time at risk for a particular condition.

2 types of prevalence

point prevalence and period prevalence

point prevalence

prevalence at a specific point in time

period prevalence

how many people had this condition at any time over a certain period of time

question you would ask for point prevalence

do you currently have x?

question you would ask for period prevalence 

have you had x during this time period?

difference between period prevalence and cumulative incidence

incidence is only new cases, period prevalence is not necessarily new cases

calculating point prevalence

P = # individuals with a condition at a specific time / # individuals in the population at that specific time

when using “prevalence” generally, which one are they usually referring to

point prevalence

CI numerator is 2

Period prevalence numerator for 2012 is 5

Point prevalence numerators:

• May = 5

• July = 4

• September = 3

• November = 2

measures of disease frequency

prevalence, cumulative incidence, incidence rates

how do we compare measures of disease frequency?

by exposure status

why do we compare measures of disease frequency?

if disease occurs more often in one group than in another group, there is an association between the exposure and the disease

t/f: association = causation

NO

exposed group is also called

index group

unexposed/less exposed group is called

comparison/reference group

measure of association definition

comparison of measures of disease frequency

2 types of measures of association

absolute measure + relative measure

absolute measure

calculating the difference between two measures of disease frequency

relative measure

calculate the ratio of 2 measures of disease frequency

You are running late for an appointment, and you get stopped by a
police officer because you are driving 60 MPH in a 30 MPH zone.

How would the police officer describe the offense using absolute terms?

You were traveling 30 mph over the speed limit.

You are running late for an appointment, and you get stopped by a
police officer because you are driving 60 MPH in a 30 MPH zone.

How would the police officer describe the offense using relative terms?

You were traveling twice/double the speed limit.

t/f: all absolute measures of association have no units

false; P and CI have no units, but incidence rate is in person-time units

general formula for absolute measures of associations

RD = R(exposed) - R(unexposed)

what does it mean if RD = 0

There is no difference in risk between the exposed and unexposed groups.

what does it mean if RD > 0

The exposure is associated with an increased risk of disease.

what does it mean if RD < 0

The exposure is associated with a decreased risk of risease

absolute measure of association (Rexposed - Runexposed) is also called

attributable risk (AR)

The Nurses’ Health Study is a longitudinal cohort study of factors
that influence women’s health. NHS researchers investigated the
hypothesis that hypertension increases the risk of myocardial
infarction (MI). Among 13,422 women with hypertension, 117
had an MI over 10 years of follow-up. Among 106,541 women
without hypertension, 125 had an MI during the same follow-up
period.

Based on only this info, which type of absolute difference measure can we calculate?

risk difference

3 things to consider when interpreting risk difference

• Focus on the excess disease in the exposed group

• Specify comparison group

• Specify time period

3 main relative measures of association

CI ratio (or risk ratio), IR ratio (or rate ratio), and odds ratio

relative measures of association tell us

how strongly an exposure is associated with an outcome; likelihood of developing the outcome

what does it mean if relative measure = 1

there is no association between exposure and disease

what does it mean if relative measure > 1

the exposure is associated with a higher risk of the disease.

what does it mean if the relative measure < 1 BUT > 0

the exposure is associated with a lower risk of the disease.

risk ratio (RR) estimates

how strong the association is between a given exposure and outcome

RR formula

[a / (a + b)]/[c / (c + d)]

Calculate + interpret the RR

2.10

Women with a first-degree relative with a history of breast cancer are 2.1 times as likely to contract the disease over four years as women with a negative family history of breast cancer.

***you have to include the time frame !!

person-time incidence ratio (IR) formula

IR = [(a / P-Y1)] / [c / P-Y0]

***numerator is exposed and denominator is unexposed

Calculate and Interpret the IR

2.24

Men 50-55 years of age with diabetes of at least 5 years duration are 2.24 times as likely to develop clinical CHD than men without diabetes.

Note
that since the IR is a rate, it’s considered the “instantaneous” rate of
change and no formal time period has to be given. However, for ease of
understanding the context if the IR, the time period is often given (here,
over 5 years).

odds ratio

the odds of having the outcome in the exposed divided by the odds of having the outcome in the unexposed

when do you use odds ratio

for studies not involving incidence or risk (usually cross-sectional and cross-control studies involve incidence and risk)

odds ratio formula

OR = ad / bc

calculate and interpret the OR

4.37

The interpretation of the Reye’s syndrome
study OR is that children exposed to aspirin
during a recent viral illness were over 4 times
as likely to have (not contract or get or
develop) Reye’s syndrome as unexposed
children.