EBCR III Flashcards
Intro to Study Designs
Relative strength of Evidence: Hierarchy of study
RCT>[Cohort>Case-Control> Cross-Sectional]>Case Series/Case Reports>Ideas, Opinions, and reviews
Cross-Sectional Study
Study Designs
no hypothesis test>Descriptive
Testing a hypothesis> Analytical >Experimental or Observational
How to evaluate causal inference (i.e. cause and effect)
- temporality
- Strength
3.Dose-response - Replication
- Biologic plausibility
- Alternate explanation
- Cessation of exposure
- Specificity
Cross-sectional Study
Identify the prevalence or characteristics of a condition in a group of individuals (prevalence study)
- Snapshot of the population at a certain point in time
-hypothesis generating
- define population
- gather data on exposure and disease
- Exposed; have disease, Exposed; Do NOT have disease, NOT Exposed; Have Disease, NOT exposed; Do NOT have disease
Wha
What are the four possible groups in a Cross-Sectional Study?
- Exposed; have disease
- Exposed; Do NOT have disease
- NOT Exposed; Have Disease
- NOT exposed; Do NOT have disease
How to obtain information
- define population
- gather data on exposure and disease
- Exposed; have disease, Exposed; Do NOT have disease, NOT Exposed; Have Disease, NOT exposed; Do NOT have disease
Advantages of Cross-Sectional Study
Easy Design
Data on exposure and outcome collected at one point in time
Studies are often accomplished by questionnaire, interview, or review of medical info in large database
Cohort Study Design
-identifies 2 groups of pts (cohort) - exposed and not exposed
-the 2 cohorts are observed over time to see which develops the outcome of interest
-can be perspective or retrospective
-can calculate the incidence of the disease
Case-Control Study
Disadvantages of Cross-Sectional Study
Case-Control Study
-A study which selects patients who have the outcome of interest (Cases) and pts w/o that outcome (controls), and looks back in time to identify characteristics that are inked to the outcome in case pts
-case control studies are retrospective
Disadvantages of Cross-Sectional Study
Disadvantages of Cross-Sectional Study
Case-Control Study
-A study which selects patients who have the outcome of interest (Cases) and pts w/o that outcome (controls), and looks back in time to identify characteristics that are inked to the outcome in case pts
-case control studies are retrospective
Cohort Study
-identifies 2 groups of pts (cohort) - exposed and not exposed
-2 cohorts are observed over time to see which develops the outcome of interest
-Can be prospective or retrospective
-can calculate the incidence of the disease
Bias in study designs
-Systemic error in study design or in the way study subjects are selected, measured, and analyzed that can lead to incorrect findings
Bias in study designs
-Systemic error in study design or in the way study subjects are selected, measured, and analyzed that can lead to incorrect findings
Disadvantages of Cross-Sectional Study
What can bias in study designs lead to?
-mistake estimate of an exposure’s effect on outcome
-over/underestimation of the efficacy of treatment
-error in outcome interpretation
Types of Bias
- selection bias
- information bias
- recall bias
- reporting bias
- publication bias
- interviewer bias
- misclassification bias
Selection bias
An error in the selection of or sampling of individuals for a clinical study
Information bias
An error in the recording of individual factors of a study
Selection bias
An error in the selection of or sampling of individuals for a clinical study
Information bias
An error in the recording of individual factors of a study
Disadvantages of Cross-Sectional Study
Recall bias
Better recall in cases vs controls
reporting bias
reluctance to report exposure based on attitudes, beliefs, perceptions
Publication bias
only results with positive outcomes are published
Interviewer bias
interviews not conducted informally
misclassification bias
controls misclassified as cases or with respect to exposure
reporting bias
reluctance to report exposure based on attitudes, beliefs, perceptions
misclassification bias
controls misclassified as cases or with respect to exposure
Disadvantages of Cross-Sectional Study
Publication bias
only results with positive outcomes are published
Disadvantages of Cross-Sectional Study
Publication bias
only results with positive outcomes are published
reporting bias
reluctance to report exposure based on attitudes, beliefs, perceptions
Case-Control Design
- Study which selects patients who have the outcome on interest (cases) and pts w/o that outcome (controls), and looks back in time to identify characteristics that are linked to the outcome in case pts
- Case controls studies are retrospective
Publication bias
only results with positive outcomes are published
What can bias in study designs lead to?
-mistake estimate of an exposure’s effect on outcome
-over/underestimation of the efficacy of treatment
-error in outcome interpretation
reporting bias
reluctance to report exposure based on attitudes, beliefs, perceptions
Disadvantages of Cross-Sectional Study
How to control for confounding
- randomization
- Restriction
- Matching
- Regression methods
- Stratification
Confounding
Occurs when factors (confounders) are present that are correlated to the independent variable and affect the dependent variable
–alters the ability to determine the true effect of the independent variable. on the dependent variable (outcome). It may hide OR exaggerate a trie association
Case-Control Design
- Study which selects patients who have the outcome on interest (cases) and pts w/o that outcome (controls), and looks back in time to identify characteristics that are linked to the outcome in case pts
- Case controls studies are retrospective
Confounding
Occurs when factors (confounders) are present that are correlated to the independent variable and affect the dependent variable
–alters the ability to determine the true effect of the independent variable. on the dependent variable (outcome). It may hide OR exaggerate a trie association
Publication bias
only results with positive outcomes are published
Disadvantages of Cross-Sectional Study
reporting bias
reluctance to report exposure based on attitudes, beliefs, perceptions
How to control for confounding
- randomization
- Restriction
- Matching
- Regression methods
- Stratification
What can bias in study designs lead to?
-mistake estimate of an exposure’s effect on outcome
-over/underestimation of the efficacy of treatment
-error in outcome interpretation
Case-Control Design
- Study which selects patients who have the outcome on interest (cases) and pts w/o that outcome (controls), and looks back in time to identify characteristics that are linked to the outcome in case pts
- Case controls studies are retrospective
Bias vs Confounding
bias - method used to select subjects, collect data and results in incorrect association
confounding - observed association is due to “lurking” variable that’s associated with risk factor (exposure) and outcome
randomization
works by balancing the factors that can confound results between cases and controls
restriction
excluding individuals with potential confounding characteristics
regression methods
adjusts for potential confounders through mathematical modeling (e.g., logistic regression)
matching
adjusts for factors by making like-to-like comparisons
Publication bias
only results with positive outcomes are published
regression methods
adjusts for potential confounders through mathematical modeling (e.g., logistic regression)
randomization
works by balancing the factors that can confound results between cases and controls
matching
adjusts for factors by making like-to-like comparisons
reporting bias
reluctance to report exposure based on attitudes, beliefs, perceptions
Disadvantages of Cross-Sectional Study
matching
adjusts for factors by making like-to-like comparisons
Confounding
Occurs when factors (confounders) are present that are correlated to the independent variable and affect the dependent variable
–alters the ability to determine the true effect of the independent variable. on the dependent variable (outcome). It may hide OR exaggerate a trie association
randomization
works by balancing the factors that can confound results between cases and controls
Disadvantages of Cross-Sectional Study
Publication bias
only results with positive outcomes are published
regression methods
adjusts for potential confounders through mathematical modeling (e.g., logistic regression)
Stratification
Divides the dataset into homogenous subgroups and no subset analyses
reporting bias
reluctance to report exposure based on attitudes, beliefs, perceptions
What can bias in study designs lead to?
-mistake estimate of an exposure’s effect on outcome
-over/underestimation of the efficacy of treatment
-error in outcome interpretation
Publication bias
only results with positive outcomes are published
Stratification
Divides the dataset into homogenous subgroups and no subset analyses
randomization
works by balancing the factors that can confound results between cases and controls
Disadvantages of Cross-Sectional Study
matching
adjusts for factors by making like-to-like comparisons
Confounding
Occurs when factors (confounders) are present that are correlated to the independent variable and affect the dependent variable
–alters the ability to determine the true effect of the independent variable. on the dependent variable (outcome). It may hide OR exaggerate a trie association
regression methods
adjusts for potential confounders through mathematical modeling (e.g., logistic regression)
What can bias in study designs lead to?
-mistake estimate of an exposure’s effect on outcome
-over/underestimation of the efficacy of treatment
-error in outcome interpretation
reporting bias
reluctance to report exposure based on attitudes, beliefs, perceptions
