Statistics Flashcards
Probability that a test will correctly identify a True Positive when used on a population with the disease
Sensitivity
Sensitivity formula
TP / (TP + FN)
Probability that a test will correctly identify a True Negative when used on a population without disease
Specificity
Specificity formula
TN / (TN + FP)
Probability that someone who tests positive actually has the disease
Positive predictive value
Positive predictive value formula
TP / (TP + FP)
Probability that someone who tests negative is actually disease free
Negative predictive value
Negative predictive value formula
TN / (TN + FN)
PPV/NPV relationship to prevalence
PPV is directly correlated with prevalence
NPV is indirectly correlated with prevalence
Results obtained due to chance alone
Null hypothesis
The chance that the null hypothesis was rejected in error
P value
(Null was correct and the results were due to chance)
Type 1 error
Rejected the null hypothesis in error
Same as p value
False positive
Alpha
Type II error
Accepted null hypothesis in error
False negative
Beta
Number Needed to Treat (NNT)
Number of patients that would need to be treated to prevent one additional adverse effect
1 / ARR
Absolute Risk Reduction (ARR)
Rate in untreated group - rate in treated group
OR
rate in treatment A - rate in treatment B
Positive likelihood ratio
Sensitivity / (1-specificity)
Negative likelihood ratio
1 - Sensitivity / specificity
LR > 1
Increases probability that the condition is present
LR < 1
Makes condition less likely
Absolute Risk
+ disease patients / total patients exposed
Relative Risk
Exposed patients (disease/total) / unexposed patients (disease/total)
Odds Ratio
Cases (exposed/unexposed) / Controls (exposed/unexposed)
Cohort studies
Compare group with exposure to group without (control)
Case-control studies
Compare group with disease (case) with group without disease (control)
Analyze for exposure to risk factors
Cross-sectional studies
Observational
One point in time