Biostat Flashcards
What is a case control study?
What statistical measure does it use?
- Compares group with dz and group without dz….looks for exposure or risk factor (retrospective)
- Odds ratio
What’s a cohort study?
What statistical measure does it use?
- Group with known exposure or rf vs group without exposure…..does exposed group have higher chance of disease?
- Relative risk
Cross sectional study?
Assesses FREQUENCY of disease at a particular point in time. Asks “what is happening?”
-measures disease prevalence
Calculation for sensitivity:
TP / (TP +FN)
Calculation for specificity:
TN / (TN + FP)
PPV calculation:
TP / (TP + FP)
NPV calculation:
TN / (TN + FN)
Odds ratio is used in ____ studies. The formula is ____
- Case control
- AD / BC
Relative risk is used in ____ studies. Calculation is ____
- Cohort studies
- ( A / (A+B) ) / ( C / (C+D) )
*risk of dev. disease in exposed group vs risk in unexposed group
How to calculate Absolute risk reduction:
Risk in non-treated - risk in treated
Calculation for NNT?
1 / ARR
Calculation for # needed to harm?
1 / attributable risk
Raising the cut point (increasing inclusion criteria) of a screening test increases _____ and decreases _____
- Increases: Specificity
- Decreases sensitivity
What is susceptibility bias?
A subgroup of selection bias…Happens when the tx selection for a pt depends on the severity of their condition
What is observer bias?
Observer is influenced by prior knowledge or details of the study
What is matching?
what studies is it used in?
What does it help control?
- Matching participants in regards to certain confounding variables
- Case control studies
- Counfounding
what study is best to determine incidence?
Cohort study
Loss to folowup can create what kind of bias?
Selection bias
WHat is respondent bias?
When outcome is obtained by the pt’s response rather than objective diagnostic methods
WHat is the attributable risk percent (ARP) ?
-calculation?
- Etiologic fraction. Measure of the impact of a RF being studied
- ARP = (risk in exposed - risk in unexposed) / risk in exposed
OR
ARP = (RR -1 ) / RR
Formula for +Likelihood ratio?
Proportion of pts with disease who test positive / pts without disease who also test positive
LR +=Sensitivity / 1 - specificity
LR- = 1 - Sen / Spec
How to calculate pre-test odds?
pre-test probability / (1-pre-test probability)
AKA = prevalence / 1 - prevalence
How to calculate post-test odds?
LR x Pre-test odds
How to test Post-test probability?
post-test odds / (post test odds+1)
