Biostatistics, Ethics, Patient Safety

Question 1

What is the “rare disease assumption”?

Answer 1

Disease is rare (low prevalence)→disease incidence typically low►OR generally approximates to the RR

Question 2

What is a Factorial design study?

Answer 2

Involves 2 or more experimental interventions with additional study of 2 or more variables

Ex: three treatment arms (metoprolol, amlodipine, ramipril) and outcomes: 2 BP endpoints (102-107 or <92)

Question 3

What is the standard error of the mean?

Answer 3

Standard Deviation/Square root of the sample (n)

*Measures how tightly group a set of data is►more sample→grouping is narrower→more precise

Question 4

Which bias do you look forward to controlling doing matching?

Answer 4

Confounding➡look for similar distributions in both groups

*Matching based on potential confounding variables (age, race, etc). Usually used in case-control studies

Question 5

What may suggest an association (OR) decrease after adjustment for some factors even if it remains statistically significant?

Answer 5

The association is not fully explained by the exposition, and may be explained by external factors not adjusted in the study - Additional factors may contribute to the association

Question 6

Most important determinant of the specificity of a test

Answer 6

Proportion of false positive

*Specificity= 1 - False positive rate➡specific tests have low false positive rate