Midterm Exam Flashcards
CADR & CIDER
Databases from insurance claims for large scale, big data trend searches
Controls
Should be the exact same as the cases and would have been included had they had the disease. All other covariates are the same
Types of Matching in Case Control
Frequency Matching
Based on the population distribution in cases
1:k
Must be acknowledged in the statistical test
Epidemiological Studies
Cross-Sectional
Cohort
Case Control
Exposure
Behavioral, environmental, or clinical factor that could influence the incidence of a disease or mortality
Incidence
Measure of the risk of developing some new condition within a specified time period
Prevalence
Proportion of the population with a given disease
Case Report
Simplest
Physician notes an event and then documents the course of the patient to develop of hypothesis
Case Series
Medical study tracks patients prospectively or retrospectively to observe the effects
Observational Studies
No causality
Have a large N number
Looks at association
Open to bias
Cross Sectional
Non-experimental
Data on exposure and disease taken at the same time
Bias: publication, survival, recall, selection
Statistics for Cross Sectional
OR, RR and logistic regression
If b>0 then it means that there is positive connection
if b=0 then there is no connection
if b<0 then there is a negative connection
Retrospective cohort study
Medical records of similar individuals except for a key factor is collected and then analyzed to see if there is a difference in the groups. RR and OR can be used
Good for rare diseases
Case Control
Disease status is known before the study and then the exposure status is collected retroactively to determine if there is an increased chance of it being in that group.
Use OR
Odds Ratio
The OR represents the odds that an outcome will occur given a particular exposure, compared to the odds of the outcome occurring in the absence of that exposure
OR=(ad)/(bc)
Prospective Cohort Study
Non-experimental A cohort of people is developed with differing exposure levels and then followed overtime Has the temporal relationship Incidence can be checked False correlations can be present
Aging Longitudinal Study
Starting point is birth
Starts with unaffected
Follow over time
Natural History
Follow a group of people overtime who have the disease or at risk for a disease
Gives information on how the disease manifests overtime
Time point for the start is normally diagnosis
Follow Up Study
Longitudinal study where individuals over time are monitored and use survival data
Systematic Error
error that is not determined by chance but is introduced by an inaccuracy inherent in the system
Longitudinal Statistic Methods
Mixed Linear Models
Generalized Estimating Equations
Random Coefficients Model
Clinical Trial
Can be blinded
There is an intervention that is being added and then the effect is measured
Experimental
Most definitive
Phases of Clinical Trial
Pre-Clinical, Phase I, Phase II, Phase III, Phase IV
Phase I
MTD is found as well as pharmokinetics
Done in healthy volunteers to determine the safety and the dosing of the drug
Done in small N
Used to find recommended dose
Phase II
Done to look at safety and efficacy in the target population
Looks at side effects and to determine if a bigger study should be done
Phase III
Superiority trial against the current standard of care
Done with much larger N
Phase IV
Post market surveillance to make sure the drug doesnt have additional issues
Types of Clinical Trials
Historical
Concurrent
Randomized
Historical CTs
Compared to controls who were not given the intervention but was information from previous studies
Concurrent CTs
Data from controls are collected at the same time as the intervention group
Randomized CTs
Assignment to the control and intervention are done at the same time
Randomization spreads baseline characteristics
CT classification
Screening Trials
Prevention Trials
Therapeutic Trials
Population
Well defined with inclusion criteria to specify who should be allowed in the study
Exclusion criteria of who is in the study but should be removed because it will bias the results or because they are at higher risk
Surrogate Endpoint
Measure of effect of a specific treatment that may correlate with a real clinical endpoint but does not necessarily have a guaranteed relationship.
Primary Outcome
Outcome of greatest importance