CSIM 2.1 Diagnostics 1 Interpreting Tests Flashcards
100% Sensitivity
Test will find 100% cases of disease, but will also have false positives. If a test has good sensitivity, good to exclude presence of disease
100% Specificity
All those who haven’t got disease would get a negative result, but will be false negatives. If a test has good specificity, good to rule in presence of disease
Sensitivity
True positive rate = TP/(TP+FN)
Specificity
True negative rate = TN/ (TN+FP)
positive predictive value
TP/(TP+FP) % of people with positive results who had disease
Negative predictive value
TN/(TN+FN) % of people with negative results who didn’t have disease
Likelihood Ratio
likelihood of result meaning disease vs no disease. Independent of population composition. Best measure of test performance
Positive Likelihood ratio
TP rate/FP rate or Sensitivity/ 1-Specificity. Value of 1 means test useful. Criteria of 10 to be good test
Negative Likelihood Ration
FN rate/TN rate or 1- Sensitivity/ Specificity. Value of 1 means test useful. Criteria of 0.1 to be good test
ROC curves
ROC Slope: Likelihood ratio (LR)
ROC Area: Discriminatory power of test