Page 94 - Radiology Book

ebm definitions continued
Low disease prevalence: ↓ PPV ↑ NPV High disease prevalence: ↑ PPV ↓ NPV ↑ Speci city: ↑ PPV
↑ Sensitivity: ↑ NPV
· Sensitivity and speci city remain the same regardless of prevalence.
· If a disease is rare in a screened population, even with high sensitivity and speci city, it will produce many more false positives than true positives.
· PPV: the proportion of infected patients that test positive for a disease.
· NPV: the proportion of noninfected patients that test negative for a disease.
· P value: the probability that the results of a study could have been obtained by chance alone. A p value <0.05 is considered statistically signi cant; one is willing to assume a 5% risk that a false positive error has occurred.
· Number needed to treat (NNT): the number of patients who would have to receive the treatment to prevent one additional bad outcome.
· Con dence interval: the range of values within which one can be con dent that the population value estimated by the study lies.
· Relative risk (RR): ratio of the exposed group risk to the the unexposed group risk. If equal, RR = 1. If exposure is harmful, RR >1. If exposure is protective, RR <1.
Screening tests: used to rule out a diagnosis; should have a high degree of sensitivity.
Con rmatory tests: used to rule in a diagnosis; should have a high degree of speci city.
Measure
Definition
Relative Risk (RR)
(event rate in intervention group)/ (event rate in control group)
RR reduction
1 - RR OR
(absolute risk reduction)/ (event rate in control group)
Absolute risk reduction
(event rate in intervention group) - (event rate in control group)
NNT
1/(absolute risk reduction)
Southfield16
EpidEmiology