statistics Flashcards
Sensitivity
TP/(TP+FN)
Specificity
TN/(TN+FP)
PPV
TP/(FP+TP)
NPV
TN/(FN+TN)
Positive Likelihood ratio
LR+ = Sn/(1-Sp)
Likelihood that positive test expected in patient with disease compared with positive test result in patient without the disease
Negative Likelihood ratio
LR- = (1-Sn)/Sp
Pretest Odds
pretest odds= pretest prob/(1-pretest prob)
Post test odds
post test odds = pretest odds x LR
Post test probablility
Post test prob = post test odds/(1+post test odds)
Absolute risk AKA Event rate
events in group/total patients in group
Often calculated for experimental event rate EER and controlled event rate CER
Relative risk
RR= EER/CER
Absolute risk reduction
ARR= EER-CER
Relative risk reduction
RRR= (EER-CER)/CER
NNT
NNT= 1/ARR
NNH
NNH= 1/ARI where ARI is absolute risk increase which is EER-CER where the outcome is unfavourable
variance
variance = SD squared
standard error of the mean
SD/(square root n)
Small sample size increases the chance of what type of error with respect to the null hypothesis
Type II error
Type 1 error
incorrect rejection of the true null hypothesis
Type II error
Incorrect rejection of the false null hypothesis
hazard ratio
Sort of like relative risk. Used in survival analysis to reflect time survived to an event. useful if risk not constant with respect to time
Cohort study
prospective
cohort + exposure –>?disease
cohort + no exposure–>?disease
Case control
Retrospective
Disease–>?exposed
Control without disease –>?exposed
Phase 1 trial
pharmacokinetics
