Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

Summer Review Cards > Epi > Flashcards

Epi Flashcards

1
Q

Epidemiology

A

A public Health discipline basic science which studies the distribution and determinants of disease in populations to control disease and illness and promote health

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Absolute difference

A

Absolute value of X-y (always subtract). Ex: # of live births - total deaths

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Attack rate/ incidence proportion

A

Proportion of the population that develops illness during an outbreak. # of new cases of disease/ # of people at risk/in pop

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Secondary attack rate

A

Measure of the frequency of new cases of a disease among contacts of known patients

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Relative difference

A

Value of X/y (always divide) (ratio)

Ex: # of deaths from cancer / total # of deaths

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Risk

A

Probability that an event will occur (that a person will be affected by, or die from, an illness, injury or other health condition w/I a specified time/age span)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Incidence

A 
New cases of disease (INclusion of new diseases). 
# of new cases of disease/# of persons at risk for the disease 
(Not precise for dynamic populations)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Prevalence

A

Everyone who has gotten the disease PREViously, including new cases. # of exsisting cases of disease / # of persons in population

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Period prevalence

A

Prevalence over a given period of time

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Point prevalence

A

Prevalence at a give point in time

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Risk Ratio (RR)

A

Customarily used in studies where subjects are allocated based on exposure and evaluated for disease;
Risk in exposed / risk in non-exposed
= 1 - no difference
> 1 - increased ratio (higher, greater, more, etc.)
>/= 2 - use statement of “times control” (6.18 TIMES greater)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Risk/Incidence Ratio (IR)

A

Probability of outcome in exposed and non-exposed (proportion)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Number Needed to Treat (NNT)

A 
1 / absolute risk reduction (1/ARR)
Number of (whole) patients needed to be treated to experience the outcome. If outcome is beneficial, want NNT to be small. If it is harmful want NNT to be big.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Absolute Risk Reduction (ARR)

A

Simple “absolute” difference (subtraction) in risks. Ex: risk1 - risk2

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Odds and Odds Ratio (OR)

A

Customarily used in studies where subjects are allocated based on disease presence and evaluated for exposure (case-controls).
Occurance of an event happening / not happening

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Selection Related Bias

A

Any aspect in the way the researcher selects/acquires study subjects which creates a systematic difference in the composition b/w groups

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Self-selection/Participant (responder) bias

A

Those that wish to participate (volunteer) may be different to those that don’t

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Recall (reporting) bias

A

A differential level of accuracy/detail in provided info b/w study groups. Exposed/diseased may have greater sensitivity for recalling their history or amplify their responses

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

What are Hill’s Criteria?

A
  1. Strength: size of association
  2. Consistency: repeated observations
  3. Temporality: cause precedes effect/outcome Ex: don’t quit smoking b/c of cancer
  4. Biological gradient: observation of a gradient risk (dose-response) associated w/ degree of exposure. Ex: 4 pack/day worse
    than 1?
  5. Plausibility: biological feasibility
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Observational Studies

A

Study designs considered “natural.” Researcher “observe” subject elements occuring naturally or selected by individual (natural/freely). Useful for: unethical study designs using forced interventions. Most observational study designs aren’t able to prove causation. **There is NO researcher-forced group allocation.
Examples: Cases, cross-sectional, case-control, cohort

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

Interventional Studies

A

Study designs considered “experiemental.” Investigator selects interventions (exposure). **There IS researcher-forced group allocation. Randomization process.

Examples: Preclinical, Phase 1, 2, 3, and 4 studies

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

Simple Study Design

A

Divides (randomizes) (one time) subjects exclusively into >/= 2 groups. commonly used to test a single hypothesis (question) at a time.
Study Population
Group A. Group B
Outcome Outcome

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

Factorial Study Design

A

Divides subjects into >/= 2 groups and then further additionally subdivides (randomizes) each of the group into >/=2 sub groups. (2x2 or 3x3x2). Seen more at the phase 3/4 level. Used to test multiple hypotheses at the same time, which increases the sample size requirement - total starting population has to be bigger b/c they will be broken down into multiple groups.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

Parallel Study Design

A

Groups simultaneously and exclusively manages. NO SWITCHING of intervention groups after initial randomization. All simple and factorial designs are also parallel. When randomized, study subjects stay in that group throughout the study.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

Cross-over study design

A

Aka. Self-control. (Forced switching; controls for the differences in individuals b/c they will take drug A and B). Groups serve as their own control by crossing over. Allows for a smaller total “N” (sample size). Each patient contributes additional data.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
26
Q

Simple Randomization

A

Equal probability for allocation w/I 1 of the study groups

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
27
Q

Blocked Randomization

A

Ensures balance w/I each intervention group; when researchers want to assure that all groups are equal in size. Done by picking blocks of people; Ex: blocks of 4, every 4th person is placed in a set group.

28
Q

Stratified Randomization

A

Ensures balance w/ known confounding variables. Ex: gender, age, disease, severity/duration, comorbidities. Can also preselect levels to be balanced w/I each interfering factor (confounder). Ex: if the 1st person randomly placed into group A is a woman, at some point a woman will need to be placed into group B. More hands on approach.

29
Q

Intent to Treat

A

Keep dropouts/lost to follow ups in the study (most conservative decision). Ex: use the last known assessment/observation for all subsequent assessments –> pt. wont’ get better/won’t get worse. Convert all subsequent (yet missed) assessments to a null effect (no benefit) –> will use baseline; doesn’t show data of improvement or getting worse.
Results in: preserving the randomization process; preserves the baseline characteristics and group balance at baseline controls for known and unknown confounders; maintains statistical power; lose ability to see what effec tight be possible in best case scenario.

30
Q

Per-Protocol / Efficacy analysis

A

Compliance must be pre-defined (usually set at 80-90% compliance). any dropouts/ lost to follow ups will be ignored. Ex: if you can’t complete 80% of the study, then your data will be ignored. Doesn’t tell us what effects we would get from people who regularly only take 60% of their medication. Reduces the generalizability; the effect is commonly over estimated.

31
Q

Describe the purpose and methodology utilized in the conductance of a case-control study and when it might be used in place of other study designs

A

Case-control: observational, analytical studies allowing researchers to be a passive observer of natural events occuring in individuals w/ the disease/condition of interest (case) who are compared w/ people who don’t have the condition of interest (control). Control group supplies info about the expected baseline risk-factor profile in the population form which the cases are drawn.

32
Q

In Case-Control studies, group assignments are based on what?

A

DISEASE STATUS - useful when studying a rare disease or outbreak.

Population –> diseased/outcome yes or no? –> exposed or non exposed?

33
Q

What do case-control studies commonly generate?

A

OR as a measure of association

34
Q

Describe the purpose and methodology (design) utilized in the conductance of a cohort study and when this study design might be used in place of others

A

Cohort studies - observational, analytical studies allowing researchers to be a passive observer of natural events occuring in naturally exposed/unexposed (comparison) groups. Group allocation is based on EXPOSURE status OR group membership (something in common). Useful when studying a rare exposure. Aka incidence studies, follow up studies, longitudinal studies.

35
Q

Cohort studies commonly generate what?

A

Risk ratio (RR) as a measure of association

36
Q

Describe the purpose and methodology used in the conductance of cross-sectional studies and when it might be used in place of other designs

A

Cross-sectional studies - observational, descriptive/analytical studies that examine relationships of health/disease to other variables of interest at the same time. Aka prevalence study. Entire population or a subset is selected for study; called cross-sectional b/c information gathered represents what is occuring at a point in time across a large pop (a “snap shot” in time). If a study has the word “national” it’s likely to be cross-sectional. Focuses simultaneously on disease and population characteristics, including exposures, health status, healthcare utilization etc. seeks ASSOCIATIONS no causation. Ex: prevalence of diabetes in KC. Most are surveys or databases.

37
Q

Sensitivity

A

How well a test can detect presence of a disease when in fact the disease is present. Positivity of test in diseased. Proportion of time that a TEST is positive in a patient that DOES have disease.

TP / (TP+FN) x 100%
TP / (all diseased) x 100%

38
Q

Specificity

A

How well a TEST can detect absence of disease when in fact the disease is absent. (Accuracy in people who do not have disease). Negativity of test in healthy. Proportion of time that a TEST is negative in a patient that DOES NOT have disease.

TN / (TN+FP) x 100%
TN / (all NOT diseased) x 100%

39
Q

If a highly sensitive test has a low false negative rate, what does that represent and what box does it belong in?

A

sensitivity; box C

40
Q

If a highly specific test has a low false positive rate, what does that signify and what box does it belong in?

A

Specificity; box B

41
Q

Positive Predictive value (PPV)

A

How accurately a POSITIVE test predicts the presence of disease. Percentage of TPs in patients w/ a POSITIVE test (correct prediction).

TP / (TP+FP) x 100%
TP / (all positive tests) x 100%

Ex: if a test has a 50% PPV, what % of the 5000 patients who I screen and have a positive will actually be reported as a false positive?
= 50%

42
Q

Negative Predictive Value (NPV)

A

How accurately a NEGATIVE test predicts the absence of disease. % of TNs in patients w/ a negative test (correct prediction).

TN / (TN + FN) x 100%
TN / (all negative tests) x 100%

43
Q

Diagnostic Accuracy (DA)

A

Proportion of time that a pt. is correctly identified as either having a disease or not having disease w/ a positive or negative test, respectively. (Best possible outcome)

(TP + TN) / (TP + FP + TN + FN) x 100%
(TP + TN) / (all patients) x 100%

44
Q

Regression

A

Outcome prediction/ association; **buzz word = PREDICTION

Ex: look at COB data from last 5 years and imply any characteristics/similarities to predict data

45
Q

Logistic Regression

A

NOMINAL regression test; regressions provide a measure of the relationship b/w variables by allowing the prediction about the dependent, or outcome variable (DV) knowing the value/category of independent variables (IVs) -

Outcome prediction / association (OR)

46
Q

Linear Regression

A

INTERVAL regression test

Outcome prediction/ association (OR)

47
Q

Correlation

A

Strength and direction (as 1 goes up the other goes down). Ex: is there an overall correlation b/w overall GPA and satisfaction in semester?

Provides a quantitative measure of the strength and direction of a relation b/w variables; values range from -1.0 - 1.0

48
Q

Spearman Kendall Correlation

A

ORDINAL correlation test

49
Q

Pearson Correlation

A

INTERVAL correlation test

If p>.05, just means that there is no LINEAR correlation; there may still be non-linear correlation present

50
Q

Chi-square test

A

NOMINAL data; 2 groups of independent data; compares group proportions and if they are different from that expected by change. No cell w/ expected count of

51
Q

Fisher’s exact test

A

Like chi square test but > or = 2 groups w/ EXPECTED cell count of

52
Q

Kaplan-Meier test

A

INTERVAL survival test; if desired comparison/assessment is EVENT-OCCURANCE or TIME TO EVENT, compares the proportion of, or time-to, event occurrences b/w groups

53
Q

Students Independent T Test

A

INTERVAL; 2 groups of INDEPENDENT data. Compares the means of all groups (along w/ inter and intra group variations) against a single DV

54
Q

ANOVA

A

INTERVAL data; > 3 groups of INDEPENDENT data. Compares the means of all groups (along w/ intra and inter group variations) against a single DV

55
Q

CONSORT

A

Consolidated standards of reporting trials. For INTERVENTIONAL studies. Randomized clinical trials. Extension documents: non-inferiority and equivalence trials, cluster trials, and design extensions for pragmatic trials.

56
Q

STROBE

A

Strengthening the reporting of observational studies in Epi. OBSERVATIONAL studies. (Cohort, case-control, cross sectional.).

57
Q

The data is NOMINAL, you want to compare frequencies/proportions, there are 2 groups, the groups or their assessments are independent, the cell has an expected frequency of

A

Fisher’s exact

58
Q

The data is NOMINAL, you want to compare frequencies/proportions, there are 2 groups, the groups or their assessments are independent, the cell does not have an expected frequency of

A

Chi-square

59
Q

The data is NOMINAL, you want to compare frequencies/proportions, there are 3 or more groups, the groups or their assessments are independent, the cell has an expected frequency of

A

Fisher’s Exact

60
Q

The data is NOMINAL, you want to compare frequencies/proportions, there are 3 or more groups, the groups or their assessments are independent, the cell does not have an expected frequency of

A

Chi-square

61
Q

The data is INTERVAL, you do want to compare means, the data is distributed normally, there are 2 groups, and the groups or their assessments are independent, what test you do you use?

A

Student T or ANOVA

62
Q

The data is INTERVAL, you do want to compare means, the data is distributed normally, there are 3 or more groups, and the groups or their assessments are independent, and there is not more than 1 DV, what test you do you use?

A

ANOVA

63
Q

The data is INTERVAL, you do not wish to compare means, you do wish to assess for association or predict group membership, what test do you use?

A

Linear regression

64
Q

The data is INTERVAL, you do not wish to compare means, you do not wish to assess for associations or predict group membership, you do wish to assess for a correlation, what test do you use?

A

Pearson Correlation

65
Q

The data is INTERVAL, you do not wish to compare means, you do not wish to assess for associations or predict group membership, you do not wish to assess for a correlation, you do wish to assess for time to event/survival, what test do you use?

A

Kaplan-Meier

66
Q

The data is ORDINAL, you do not wish to compare frequencies/proportions, you do not wish to assess for associations or predict group membership, you do wish to check for correlation, what test do you use?

A

Spearman correlation

67
Q

How do you determine if confounding is present?

A

If the difference b/w the crude and adjusted measures of association are differnt by 10-20%

Summer Review Cards (7 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions