Epidemiology Exam 1 Flashcards
(96 cards)
1
Q
Epidemiology is a public health basic science which studies the _____ and _____ of health-related ____ or ____ in specific populations to control disease and illness and promote health.
A
Distribution.
Determinants.
states.
events.
2
Q
Distribution of disease involves figuring out the _____ and _____ of disease occurrences?
A
Frequencies.
Patterns.
3
Q
This type of epidemiology, _____ epidemiology, studies the ____ of disease by analyzing these three factors: ____, ____, and ____.
A
Descriptive. Distribution. Who. Where. When.
4
Q
This type of epidemiology, _____ epidemiology, studies the determinants of disease by analyzing ____ vs. ____, aka the ____ and ____.
A
Analytic. associations. causes. why. how.
5
Q
The 6 core functions of epidemiology, which help to promote health in ____.
A
- public health surveillance
- field investigation
- analytic studies
- evaluation
- linkages
- policy development
Populations.
6
Q
This core function portrays ongoing patterns of disease occurrence, so investigations, control and prevention measures can be developed and applied. The ____ registry can be used for data management and interpretation.
A
Public health surveillance.
NNDSS.
7
Q
This core function of epidemiology helps to determine sources/vehicles of disease; to learn more about the history, clinical spectrum, descriptive epidemiology, and risk factors. The CDC has a department, the ____, dedicated to this.
A
Field Investigation.
Epidemic Intelligence Service (EIS).
8
Q
This core function of epi helps to advance the information (hypotheses) generated by descriptive epi techniques. Some characteristics include:
A
Analysis or Analytic Studies.
Characteristics: study design, use of a comparison group, data interpretation, communication of study data/findings.
9
Q
This core function systematically and objectively determines relevance, effectiveness, efficiency and impact of activities.
A
Evaluation.
10
Q
This core function collaborates, communicates, links to other public health and healthcare professionals and the public themselves.
A
Linkages.
11
Q
This core function provides input, testimony, recommendations regarding disease control and prevention strategies, reportable disease regulations, and health-care policy.
A
Policy development.
12
Q
Epidemiologists are experts at describing and comparing groups by ____, ____, and ____.
A
Counting (frequencies).
Dividing (percentages).
Comparing.
13
Q
He was one of the first epidemiologists and he found this contaminated, communal water source as a common point source of disease.
A
John Snow.
Broad Street pump.
14
Q
List the three types of surveillance systems.
A
- Passive
- Active
- Syndromic
15
Q
Define passive surveillance system.
A
Relies on healthcare system to follow regulations to report on new diagnoses. Health system passively waits for reports to come in so they can track disease frequency/occurrences over time and within populations.
16
Q
Define active surveillance system.
A
public health official go into communities to search for new disease/condition cases.
17
Q
Define syndromic surveillance system. This type of surveillance, ____, is also considered syndromic.
A
a system that looks for pre-defined signs/symptoms of patients related to trackable-but-rare diseases/conditions.
Biosurveillance
18
Q
The time between exposure and onset of disease can be referred to as _____, or _____, period.
A
Induction/Incubation period.
19
Q
The time between onset of disease and disease detection (symptoms or diagnosis) can be referred to ____ period.
A
Latency. Sometimes patients are diagnosed presymptomatic.
20
Q
The four stages of the natural history of disease timeline are:
A
- Stage of susceptibility
- Stage of subclinical disease.
- Stage of clinical disease.
- Stage of recovery, disability, or death.
21
Q
This demarcates the stage between susceptibility and subclinical disease.
A
Exposure.
22
Q
This demarcates the stage between subclinical disease and clinical disease.
A
Onset of symptoms. It becomes “clinical” when the symptoms are recognized.
23
Q
____ ____ occur during the stage of subclinical disease.
A
Pathologic changes.
24
Q
The usual time of diagnosis occurs during this stage.
A
Stage of clinical disease.
25
The most critical element that must be defined/delineated BEFORE any of the "who" of descriptive epi can be acquired?
CASE definition: What requirements must be met for the person to be diagnosed.
26
Define Case definition.
A set of uniform criteria used to define a disease/condition for public health surveillance
27
Case definitions are updated every year using the recommendations made by the ____. They recommend that state health departments report cases of selected diseases to ____.
Council of State and Territorial epidemiologists (CSTE).
CDC's National Notifiable Diseases Surveillance System (NNDSS).
28
___ vs. ____ case definitions must be distinguished in order to accurately define how we detect and diagnose cases.
confirmed vs. probable.
29
List and define the five categories used to describe the geographic extent of disease frequencies.
1. Epidemic: disease occurrence clearly in excess of normal. community/period are clearly defined.
2. Outbreak (or "cluster"): epidemic limited to a localized increase in the occurrence of disease.
If "outbreak" is used first, on campus for example, then "epidemic" could be used if it spread to a larger area, KC metro, for example.
3. Endemic: normal, constant presence of disease for that specific area, but higher than surrounding areas.
4. Emergency of International Concern: Step before pandemic. alerts the world to the need for high vigilance.
5. Pandemic: An epidemic spread world-wide. (2009 H1N1 pandemic).
30
Epi Curves visually depict both ____ & ____ of disease occurrence and ____ of disease occurrence.
Magnitude & timing.
Pattern (shape)
31
The pattern (shape) of disease occurrence in an epi curve shows either a ___ source or a ____ source. Define both.
Common / Point: not spread person-to-person, spread from a single point source.
Propagated: person-to-person.
32
The magnitude and timing of disease occurrence can help to identify what?
sentinel/index cases, peaks, outliers, duration of outbreak.
33
Epi curves can help to form hypotheses about:
routes of transmission, probable exposure period, incubation period (could help identify/eliminate causes)
34
List the three types of common/point source epi curves
1. Continuous with no sentinel case.
2. Continuous with sentinel case.
3. Intermittent
35
What's the major difference between a propagated graph with an index case and a common/point source graph with an index case?
The propagated with index case graph has a sawtooth pattern, with the gap representing the average incubation period.
36
What are the three numerical representations of measures of disease frequency? Define each.
1. Proportions: simple percentage, division of 2 related numbers (part/whole).
2. Ratios: division of 2 unrelated numbers.
3. Rates: proportion (%) with time incorporated into the denominator.
37
3 key factors comparing measures of disease frequency between groups?
1. # affected/impacted (frequency/count)
2. Size of the source population, or those at risk.
3. length of time population is followed.
38
# Define incidence vs. prevalence.
Both terms are ____.
Incidence: new cases of disease.
Prevalence: existing cases + new cases.
Proportions.
39
Incidence (aka ____ or ____ ____) is calculated by:
Useful for ____ populations when the time frames for the numerator and the denominator are the same.
Incidence is also called ____ when summed over multiple time periods.
Risk or Attack Rate.
(# new cases of illness / # people at risk of illness (or # in pop.))
*The people at risk, in the denominator, does not include those who have been vaccinated, have the disease, or are immune to the disease.
Cumulative Incidence.
40
What if the at risk population for determining incidence cannot be determined?
What do you do if the individuals aren't followed for the same amount of time?
You use the entire population of the area.
You use the start or ending population of the year (or mid-year).
41
Incidence Rate equation:
For dynamic populations, how is the population estimated?
Incidence rate is also called ____ when summed over multiple time periods.
(# of new cases of disease / person-time at risk for the disease (or in pop.))
Use population at start, middle, or end of the year. Or use the average population over the entire year.
Incidence Density.
42
Prevalence is calculated by:
Prevalence at a given time point is?
Prevalence over a period of time is?
* (# of existing cases of disease / # of persons in population)
* Time frames for numerator/denominator must be the same.
Point Prevalence
Period Prevalence
43
Crude Morbidity Rate
of persons with disease / # of persons in population
44
Crude Mortality Rate
of deaths (all causes) / # of persons in population.
45
Cause-Specific Morbidity Rate
of persons with cause-specific disease / # of persons in population.
46
Cause-Specific Mortality Rate
of cause-specific deaths / # of persons in population
47
Case-Fatality Rate
of cause-specific deaths / # of cases of disease.
48
Cause-Specific Survival Rate
of cause-specific cases alive / # of cases of disease.
49
Proportional Mortality Rate (PMR)
of cause-specific deaths / total # of deaths in population.
50
Live Birth-Rate
of live births / 1,000 population.
51
Fertility Rate.
of live births / 1,000 women of childbearing age (15-44)
52
Neonatal Mortality Rate
of deaths <28 days of age / 1,000 live births.
53
Postnatal Mortality Rate
of deaths in those >28 (or =28 days) days but < 1 year of age / 1,000 live births.
54
Infant Mortality Rate
of deaths in those <1 year of age / 1,000 live births
55
Maternal Mortality Ratio
of female deaths related to pregnancy / 100,000 live births.
56
Define/compare infectivity, pathogenicity, and virulence.
Infectivity: the ability to invade a patient = #infected / #susceptible (at risk)
Pathogenicity: the ability to cause a clinical disease. (# with clinical disease / # infected)
Virulence (synonymous with Case-Fatality Rate): the ability to cause death. (# of deaths / # with infectious disease.)
57
For the 2x2 table of measures of association, what is the category for the rows and the category for the columns?
Rows: Exposure/Rx Yes or No.
Columns: Disease/Outcome Yes or No.
58
To find an absolute difference you ____ the numbers. To find a relative difference you ____ the numbers.
There can be relative ____ and relative ____.
subtract.
divide.
frequencies. proportions. One tells you the raw number difference. The proportion tells you the proportion difference.
59
Between the two differences (absolute and relative) which one tends to be smaller?
Absolute tends to be smaller.
60
How do you calculate risk, aka ____?
What does the result tell you?
Incidence Risk.
Simply divide the probability of the outcome in the group (either exposed or not) by the total amount of people in the group.
The result tells you the risk of the outcome in each group (either exposed or non-exposed)
61
How do you calculate Absolute Risk Reduction (ARR) aka ____?
aka "Attributable Risk"
Simple difference between the risk difference of the outcome attributable to exposure.
Risk of outcome in Rx is 40.9%
Risk of outcome in placebo is 53.6%.
ARR = 53.6-40.9 = 12.7%
62
How do you calculate Relative Risk Reduction (RRR)?
(ARR) / Risk(unexposed).
Example from slides: 12.7% / 53.6% = 23.7%
63
How do you calculate Number Needed to Treat (NNT)?
NNT = 1 / ARR (in decimal format).
Basically, you take the difference in risks between two groups, and place that difference in decimal-form in the denominator below 1.0.
64
How do you calculate Risk Ratio (RR) aka ____?
Relative Risk (RR).
Risk of outcome (exposed) / Risk of outcome (unexposed).
65
When interpreting ratios for any RR, OR, or HR, what 3 components are needed in the sentence?
1. Groups compared.
2. Direction of words
3. Magnitude
66
____, aka Hazard Ratio, is the frequency of an outcome occurring vs. not occurring.
How do you calculate it? What is a shortcut way to calculate it?
Odds Ratio (OR).
= Odds of exposure (in diseased) / Odds of exposure (in non-diseased)
Shortcut = [(A x D) / (B x C)]
67
Counterfactual Theory (Opposite) requires assumption of ____, which is comparability with respect to all other determinants of outcome.
When this is lacking, we must assume that there is a ____ bias
exchangeability.
confounding.
68
What 3 aspects of the study do researchers evaluate before declaring a real, true association between exposure and outcome?
1. Confounding or Effect Modification
2. Bias
3. Statistical Significance.
69
What is a confounding variable?
A 3rd variable that distorts an association (RR/OR/HR) between the exposure and the outcome.
The third variable can be mistaken for the effect of the exposure, or it can mix with the exposure and/or outcome to distort the association.
70
To be a confounder, a 3rd variable must be:
1. Independently associated with the exposure.
2. Independently associated with the outcome.
3. Not directly in the causal-pathway linking exposure to outcome.
(Direct Acyclic Graph)
71
What are the three steps to test for confounding?
1. Calculate the Crude (Unadjusted) measure of association (OR/RR) between exposure and outcome.
2. Calculate OR/RR between exposure and outcome for each individual strata of the 3rd variable. Create weighted-average (if near-equal). This is "Adjusted."
3. Compare the Crude vs. Adjusted. If >15% difference, then confounding IS present.
72
What are the 2 main impacts of confounders?
1. Magnitude of association.
| 2. Direction of association
73
Confounding can be eliminated, which is what we want to do, in two different stages, with each stage having multiple ways to control. What are the two stages?
1. Study Design Stage
| 2. Analysis of Data Stage
74
What are the three ways to control for confounding in the Study Design Stage?
Randomization, restriction, matching.
75
What are the two ways to control for confounding in the Analysis of Data Stage?
Stratification and multivariate statistical analysis (regression analyses).
76
Why do we want to control for confounders?
In order to get a more precise, accurate, truer-estimate of the measure of association between Exposure and Outcome.
77
What 5 ways can we control for confounding and name one strength and weakness of each
Randomization:
S = can make groups equal.
W = Need large sample size.
Restriction: (Confounders used to exclude people)
S = straight forward, convenient, inexpensive.
W = may limit subjects available and reduce sample size.
Matching: (Equally distributes confounding among groups)
S = Can improve analytic efficiency.
W = Difficult to accomplish.
Stratification:
S = Enhances understanding of data.
W = Can't utilize when multiple confounders are present.
Multivariate Analysis: (Use stats to factor out the effects of the confounding variable)
S = Can control multiple confounding variables.
W = Can be difficult/time consuming.
78
What is effect modification, aka ____?
How is it different from confounding?
aka Interaction.
A 3rd variable that modifies the magnitude of effect of a true association by VARYING IT WITHIN DIFFERENT STRATA.
Unlike confounding, effect modification should be described and reported at each level of the variable, rather than controlled or adjusted for.
79
When is effect modification present?
When among the different strata of the study, there is a greater than 15% OR difference between the lowest and highest strata.
80
What are 3 aspects of internal validity researchers must evaluate in order to declare a real, true association?
1. Check for confounding effect modification.
2. Check for bias.
3. Check for statistical significance.
81
Define bias.
Systematic, non-random, error in study design or conduct leading to erroneous results.
82
Once it occurs in the study, which one can be fixed, or corrected for, at the end of the study? Confounding effects or bias?
Only confounding effects can be corrected for. Bias cannot be corrected for.
83
What are the 3 elements of bias impact?
1. Source/type.
2. Magnitude/strength (bias can account for a weak association but not likely to account entirely for a very strong RR/OR)
3. Direction. (Above or below 1.0)
84
What are the 2 main categories of bias?
1. Selection-related: when the researcher creates a systematic difference because of the way the subjects were selected.
2. Measurement-related (information/observation): error in information collection of measurement/observation.
85
Types of selection bias?
Healthy-worker bias: only selecting those who are at work, you are missing the sick people aren't at work.
Self-selection/participant (responder) bias: those that participate may be different than those who do not.
86
Types of subject-related bias?
Recall (reporting) bias: memory/recall differences between subjects.
Hawthorne Effect (Observation Effect): people modify their behavior because they are part of a study and know they are being observed.
Contamination bias: members of control group receive a similar treatment, maybe from another doctor.
Compliance/Adherence bias: groups have different compliance/adherence with study.
Lost to Follow-up bias: people in one group drop out more than the other, or there are other differences between those that stay and drop out.
87
Types of Observer-related bias?
Interviewer bias: systematic (conscious or unconscious) differences in the way the interviewer solicits, records, or interprets.
Diagnosis/Surveillance (Expectation) bias: Hawthorne-like effect yet from the researcher's perspective. Different types of evaluation based on expectations.
88
Type of screening-related bias?
Lead-time bias: Some people who are treated earlier live longer and die later, so they may live longer than a group they are being compared to.
89
Misclassification bias is what? Describe the two types.
Error in classifying either disease or exposure status, or both.
1. Non-differential: error in both groups equally (not different). The misclassification of disease or exposure is UNRELATED to the other, therefore both groups move closer to 1.0
2. Differential: error in one group differently than the other. The misclassification of disease or exposure is RELATED to the other. Can move the OR/RR away from or towards 1.0.
90
List ways to control for bias.
Select the most precise, accurate, medically-appropriate measures of assessment and evaluation.
Blinding/masking.
Randomize observers/interviewers for data collection.
91
Define cause.
A precursor event, condition, or characteristic REQUIRED for the occurrence of the disease or outcome. association =/= cause.
92
What are the 3 types of associations and how can they manifest?
1. Artifactual (false) associations. Can arise from bias or confounding.
2. Non-causal associations. Two ways: disease may cause the exposure (RA leading to physical inactivity) or the disease and exposure are both associated with a third factor.
3. causal associations. (true causal where exposure --> outcome)
93
Describe the three types of causal relationships?
Sufficient cause: If cause precedes disease, disease will ALWAYS occur. Can have component causes (risk factors) that can collectively induce disease.
Necessary Cause: Cause must be present for disease to occur, but cause can be present with disease (TB for example)
Component Cause (aka Risk Factor; RF): factor that increases the probability of a particular disease.
94
Describe the two types of interactions in causal research.
Synergism: factors work together. combined measure of effect is greater than the sum of the individual effects.
Parallelism: factors work in parallel. 2 factors have a greater measure of effect than if either is present.
95
Name the 5 Hill's Criteria focused for this exam.
Strength, consistency, temporality, biologic gradient, plausibility.
96
The Women's Health Initiative (WHI) study, that disproved the positive effects of menopausal hormone therapy after they performed a double-blind, randomized trial, was an example of an exception to this Hill guideline.
Consistency.
