statistics Flashcards
Decreasing the pre-test probability has what effect on positive predictive value?
Decreases PPV.
Decreasing the pre-test probability has what effect on negative predictive value?
Increased NPV (inversely proportional)
Equation for PPV
True positives/True positives + False positives
TP/TP + FP
Equation for NPV
True negatives/True negatives + False negatives
TN/TN + FN
Equation for sensitivity
True positives/True Positives + False negatives
TP/TP + FN
Equation for specificity
True negatives/true negatives + false positives
TN / TN+FP
What is the usual outcome measure in a cohort study?
Relative risk
What is the usual outcome measure in a case-control study?
odds ratio
Describe a type 1 error
What is it increased by?
rejecting the null hypothesis when it is true i.e. false positive
“alpha” error
increased if the number of end-points are increased. (i.e. likely that by chance one of these end points will be reached) and increading significance level
Describe a type 2 error
What is it decreased by:
accepting the null hypothesis when it is false i.e. false negative
“beta” error
Decreased risk by increasing power and sample size (increased sample size = increased power)
Define the power of a study
What is the equation?
Probability of correctly rejecting the null hypothesis when it is false, i.e. the probability of detecting a statistically significant difference
Power = 1 – the probability of a type II error
What is the equation for NNT?
NNT = 1 / Absolute Risk Reduction
What is the equation for absolute risk reduction?
If the outcome of the study is undesirable
- Control event rate - experimental event rate
i. e. (b/b+d) - (a/a+c)
If the outcome of the study is desirable
- EER - CER
i. e. (a/a+c) - (b/b+d)
What is the equation for relative risk?
Relative risk = EER / CER
When are odds ratios most commonly used?
Case control studies
What is the equation for relative risk reduction?
Experimental event rate - control event rate/control event rate
(EER – CER) / CER
What is the equation for likelihood ratio of a positive result?
sensitivity/(1 - specificity)
What is the equation for likelihood ratio of a negative result?
(1 - sensitivity)/specificity
Which remains constant across different populations with different risk of underling disease - absolute or relative risk reduction?
Relative risk reduction remains constant (ratio)
Absolute risk reduction varies with underlying risk of disease in population
What level of evidence do case studies provide?
LEVEL III
Ia- evidence from meta-analysis of randomised controlled trials
Ib- evidence from at least one randomised controlled trial
IIa- evidence from at least one well designed controlled trial which is not randomised.
IIb- evidence from at least one well designed experimental trial
III- evidence from case, correlation and comparative studies.
IV- evidence from a panel of experts.
What are grade B recommendations based on?
Evidence from non-RCT (IIa,IIb,III)
Grade A- based on evidence from at least once RCT (Ia,Ib)
Grade C- based on evidence from a panel of experts (IV).