Quant Methods

What is the difference between fixed and dynamic populations?

fixed - membership is defined based on some event

that happens once, permanent

dynamic - membership is defined as being in a state or

condition, transient

What is a subset of the source population called?

a study sample

What is a case report vs case series?

case report - a detailed report on one individual with a new or unusual health outcome

case series - a detailed report on a single group with a new or unusual health outcome

What are the strengths vs limitation of case report and case series?

What is a cross-sectional study?

a study that compares current exposure status to current disease status at a single point in time

What are the strengths and limitations of a cross-sectional study?

What is an ecologic study?

a study that examines associations between exposures and outcomes on a population level

What are the strengths and limitations of an ecologic study?

What is an ordinal variable type?

more than 2 ranked responses

A questionnaire has responses that participants have to choose between low income, median income, or high income. What type of variable is this?

ordinal

What is a categorical variable type?

responses are grouped and unordered

A survey asking about hair color is using what type of variable?

categorical

What is a continuous variable?

assumes any numerical value between a minimum and maximum

An acid burn is an example of what exposure?

acute

Outdoor air pollution is an example of what exposure?

chronic

Genetics and ancestry and what type of exposure?

constituent

Marital status, diet, and smoking are all examples of what exposure?

time-varying

What is a limitation of a cross-sectional study in regards to exposure?

will not catch a time-varying exposure since it is at a single point of time

What are the 3 major exposure routes?

1. inhalation

2. ingestion

3. dermal (skin) contact

Exposure variables can be cateogorized into what?

• continuous

• dichotomous

• ordinal

• categorical

Non-smoker, light smoke, and heavy smoker would be categorized into what?

ordinal

Racial groups would be categorized into what?

categorical

What is the difference between critical window and induction time?

critical window - when is exposure relevant during life course

induction time -  time lag between the exposure and the biological onset of disease

If the same amount of a contaminant is ingested by two people of different body weights, who would have the higher concentration of contaminant?

the smaller person

The first and second trimesters of pregnancy in regards to a Zika virus infection are considered what?

critical window

What is the difference between induction time and latent period?

induction time - period between exposure and disease beginning

latent period - interval between when disease begins and diagnosis

What shortens the latency period and why?

screen-detected cases because one is not waiting until apparent symptoms

What is an exposure assesment?

an applied science for measuring human contact with an exposure of interest

What is the difference between reliability and validity?

reliability = consistent answers every time

validity = is it related to other measures

What is an exposome?

measure of all the exposures of an individual in a lifetime

and how those exposures relate to health

What is prevalence?

proportion of people who have a condition at a specified point in time

What is cumulative incidence (risk)?

proportion of at-risk people who develop a condition during a specified time period

What is incidence rate?

speed at which at-risk people develop a condition during a specified time period

What are the characteristics in common for measuring disease frequency?

# of cases / size of population

measured in source population of interest

What is the equation for prevalence?

What are the units for prevalence?

no units, cancels out

In 2016, there were 21 million adults with Type 2 Diabetes out of a total of 244,186,465 adults in the U.S. What was the prevalence of Type 2 Diabetes in adults in the U.S. in 2016?

21 000,000 / 244,186,465 = 8.6%

In 2018, a survey was conducted of 1,891 children 12-23 months of age in Zambia. Among these children, 1,719 reported receiving the measles vaccine. What was the prevalence of measles vaccination among these 12-23 month-old children in Zambia in 2018?

1719 / 1891 = 90.9%

How do you calculate cumulative incidence (risk)?

What are the units for risk?

no units

What does “at risk” mean?

theoretically capable of developing diasese

If you currently have a condition are you at risk for that certain condition?

no

In a cohort of pregnant women receiving prenatal care at BMC during a 1-year period from 2014-2015, there were 252 deliveries. Of these newborns, 13 were classified as low birthweight (<2,500 grams). What was the 1-year cumulative incidence (risk) of low birthweight in this group?

13 / 252 = 5.2%

Suppose we want to estimate the cumulative incidence (risk) of low birthweight in first pregnancies. Of the 252 deliveries, 67.1% of infants were a first pregnancy. Of these newborns, 11 were classified as low birthweight. What is the 1-year cumulative incidence (risk) of low birthweight for first pregnancies?

252 x 0.671 = 169.092 —> 11 / 169.092 = 6.5%

What are the limitations of cumulative incidence?

• no loss to follow up

• does not consider time of occurrence

How do you calculate incidence rate?

What are the units for incidence rate?

1/time or cases/person-time

What is person-time?

amount of time each study participant is under observation

In an open cohort of children under 5 in the UK, the effect of socio-economic deprivation on injury was examined from 1990-2009. A total of 20,668 fractures, 15,796 burns, and 10,155 poisonings were reported over a total of 2,725,000 person-years. What was the incidence rate of burns among children under 5 in the UK from 1990-2009?

15,796 / 2,725,000 = 0.00579 x 1000 = 5.8 burns/1000 person years

When would you look at absolute measure vs relative measure?

absolute - quantifies excess risk of disease in exposed group compared to unexposed group; measure of public health impact

relative - quantifies the relative risk of disease in exposed group compared to unexposed group; measure of strength or magnitude of association

How do you calculate prevalence difference (PD)?

P_E – P_U

How do you calculate risk difference (RD)?

R_E – R_U

How do you calculate incidence rate difference (ID)?

IR_E – IR_U

If there is NO ASSOCIATION between the exposure and disease, then the RD_____?

RD = 0

If the exposure is associated with INCREASED RISK of disease then the RD ____?

RD > 0

If the exposure is associated with DECREASED RISK of disease then the RD ______ ?

RD < 0

What is the prevalence ratio?

P_E / P_U

What is risk ratio?

R_E / R_U

What is incidence ratio?

IR_E / IR_U

If there is NO ASSOCIATION between the exposure and disease, then the RR ___?

RR = 1

If the exposure is associated with INCREASED RISK of disease, then the RR ____ ?

RR > 1

If the exposure is associated with DECREASED RISK of disease, then the RR ____ ?

RR < 1

What are experimental studies?

a study in which investigators assign participants to exposure groups and track them over time for incidence of health outcomes

For randomization, what should be more or less than same?

the baseline characteristics

How do we preserve randomization?

• blinding

• placebo

• compliance

What are the strengths of experiemental studies?

• Clear temporal sequence between E and D

• Minimize bias and confounding

• RCT's are optimal to detect statistically small to moderate, but clinically meaningful, effects

What are the limitations of experiemental studies?

• Must have equipoise

• Acceptability to subjects

• Can be time-consuming and costly

• Differential loss to follow-up can bias results

• Lack of compliance can bias results

What are cohort studies?

study in which investigators observe participants classified in exposure groups and track them over time for incidence of health outcomes

What are prospective cohort studies?

on the date investigators start the study (begin data collection), outcomes have not yet occurred

What are retrospective cohort studies?

on the date investigators start the study, outcomes have already occurred

What are the advantages of prospective studies?

•Better exposure and confounder data

•Less prone to bias

What are the disadvantages of prospective studies?

•More expensive and time-consuming

•Not efficient for diseases with long latency

What are the advantages of retrospective studies?

•Efficient for rare exposures

•Cheaper, faster

•Efficient for disease with long latency

What is a disadvantage of retrospective studies?

•Exposure data may be inadequate / unreliable

What defines an internal comparison?

have to come from same population

What are case-control studies?

study in which investigators enroll participants with disease (cases) and a sample of the source population (controls) and compare their exposure histories

What are the 2 rules of control selection?

1. Controls must come from the same source population as the cases

2. Controls must be selected independently from exposure

What is the “would criterion”?

if the controls had experienced the outcome, would they have been identified as cases in your study?

What study is it appropriate to use prevalence ratio or difference?

cross-sectional

What study is it appropriate to use risk ratio or rate?

cohort, experimental

How do you calculate odds ratio?

outcome of interest / outcome unexposed

What study is it appropriate to use odds ratio?

case control

What are the strengths of case-control studies?

Fewer ethical concerns than RCT

Less expensive and time-consuming

Efficient for rare outcomes and diseases with long latency

Can evaluate effects of multiple exposures on single outcome

What are the limitations of case-control studies?

Potentially ambiguous temporal sequence between E and D

Study single outcome

Inefficient for rare exposures

More opportunity for systematic bias (selection and recall)

CANNOT directly calculate risks and rates

How do you calculate probability?

What are the characteristics of probability?

no units

ranges from 0-1

What is the difference between screening vs diagnostic test?

diagnostic - confirms disease or condition

screening - positive or negative but does not give condition

What does a wider curve tell us for a probability distribution?

more spread/variability

What are percentiles?

Z score that holds x percent of score below it

How do you find percentile for a normal distribution?

Larger samples produce small or larger standard errors?

smaller (more precision)

What do you use for dichotomous variables/outcomes?

use z, not t

What is hypothesis testing?

a framework of specifying competing hypotheses (null and research) and evaluating the extent to which the sample data support the research hypothesis

What is the null hypothesis?

H₀: null hypothesis, or hypothesis of no difference, no association, or no change

assume no difference, no association, or no change

What is the research hypothesis?

H₁: research hypothesis or alternative hypothesis H_A, or hypothesis of difference, association, or change

What does reject the null hypothesis mean?

strong evidence, statistically significant result

What does fail to reject the null hypothesis mean?

weak evidence, not statistically significant result

What is a 1-tailed test?

specifies direction of association in research hypothesis

What is a 2-tailed test?

does not specify direction of association in research hypothesis (primarily observational studies)