Chapter 4 Review Questions (Key)
- Reproducibility refers to repeatability or consistency of a test
(or evaluation) from one use to the next. Validity is the ability to
accurate discriminate for a condition.
- Symptom = experienced by a patient; Sign = observation made by
examiner. Nausea is a symptom because it is experienced by patient, and
is not directly observable. Vomiting is an observable symptom. Having
a �red throat� as observed by an independent examiner is a sign, whereas
experiencing a sore throat is a symptom.
- Kappa
- Kappa measures percent agreement above chance. A good deal of
simple agreement is often merely by chance.
- False. A kappa of 0 indicates random agreement
- Definitions. Let D = disease status and T = test status. Thereby:
(a) TP = D+ and T+ (b) TN = D- and T- (c) FP = D- and T+ (d) FN = D+
and T-
- (a) Specificity = Pr(T-| D-)
(b) Sensitivity = Pr(T+|D+)
(c) PVP = Pr(D+|T+)
(d) PVN = Pr(D-|T-)
- SEN
- PVN
- Prevalence, SEN, and SPEC
- Because most people tested will be disease free and using even a
specific test in such a population results in many false positives (e.g.,
if you use a test that is 99% SPECific in a
million healthy people, you will derive 0.01
� 1,000,000 = 10,000 false positives).
- It will decrease the number of FNs,
increase SENsitivity, and decrease SPECicity.
- High sensitivity; you don't want to miss any cases.