D\'s Epidemi & Ethics

Home

Get App

Create

Gold standard for testing therapies
Randomized Clinical Trial
- Double Blinded
- Randomized
Examines group exposed to given situation or factor. Prospective or retrospective study.

Cohort
Cohort study conclusions

Relative risk
Retrospective comparison of pts with disease with healthy controls, frequency of certain exposures in both groups is considered

Case Control
Case control conclusions

Odds ratio
Survey of large number of people at one time to assess exposure and disease prevalence

Cross sectional
Cross sectional conclusions
- Hypothesis for risk factors
- Disease prevalence
Report of characteristics of a disease by examining multiple cases

Case Series
Case Series conclusions

Hypothesis for risk factors
Pooling of multiple studies examining a given disease or exposure

Met-analysis
What study can examine rare exposures and study multiple effects of exposure

Cohort Study
Study that can examine rare dz or those with long course in short amount of time

Case control
Can study multiple typed of exposure and may examine small group size.

Case Control
Study that can be used to estimate disease prevalence following exposure

Cross Sectional
Larger study size that can resolve conflicts in literature.

Meta-Analysis
What study is susceptible to recall bias?

Case Control
Study Susceptible to selection bias?

Case control
Name the bias: nonrandom assignment of subjects to study groups

Selection/Enrollment Bias
Subjective interpretation of data by investigator deviates toward desired conclusions

Investigator
Name the bias: Screening test provides earlier diagnosis in studied group compared to controlled, but has no effect on time of survival.

Lead Time Bias
Name the bias: screening test detects several slowly progressive cases of a disease & misses rapidly progressive cases.

Lenth Bias
Name the bias: subjects may respond to subjective q's in a different way than normal bc awareness of study changes their perception of examined issue

Observational Bias
Studies showing a difference b/t groups more like to be published than studies that don't show a difference.

Publication bias
Name the bias: errors of memory within subject bc of prior confounding experiences

Recall Bias
Pts with a certain medical hx may be more likely to participate in a study related to their condition

Self selection bias
Consequence of selection bias

Results not applicable to general population
Consequence of investigator bias

Results of study incorrectly resemble proposed hypothesis
Consequence of lead time bias:

Time from diagnosis to outcone increases to cause a false appearance of increased time of survival.

Time from dz occurence to outcome actually remains tge same regardless of screening
Consequence of length bias

Effectiveness of screening test overstated
Consequence of observational bias

Effectiveness of therapy not accurately depicted in study group
Consequence of publication bias

Dats available for meta-analysis may not include studies that support the null hypothesis
Consequence of recall bias

Pt with negative experiences are more likely to recall negative details
Consequence of self selection

Subjects are not representative of the general population & introduce confounding variables
How do u avoid investigator and observational bias?

Double blind a study
Dz with  prevalence will e associated with

 Positive predictive value in a screening test
Dz with  prevalence will be associated with

 Negative predictive value in a screeninh test
Null hypothesis

States no association exists between exposure and disease (or treatment and response)
Type I error

Null hypothesis is rejected even though its true

(False Positive)
Type II error

Null hypothesis is not rejected even though it is false

(False Negative)
Risk of type 1 & 2 errors decrease by?

Increasing Sample Size

(increasing power)

Author

JCremer

29309

Card Set

D\'s Epidemi & Ethics

Description

Epidemi & Ethics

Updated

2010-08-08T17:47:12Z

Show Answers