Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
26 Cards in this Set
- Front
- Back
|
What is Validity?
|
Accuracy; The ability of the test to distinguish b/w who has a disease & who does not (the truth)
|
|
|
What is reliability?
|
Precision; Repeatability. Gives consistent results when test is performed on the same individ more than once under the same conditions
|
|
|
What are the 2 components of validity?
|
Sensitivity & specificity
|
|
|
What is sensitivity?
|
The ability of the test to identify correctly those who HAVE the disease (rule in disease). Or the propability of testing positive if the disease is truly present
|
|
|
What is specificity?
|
The ability of the test to identify correctly those who DO NOT HAVE the disease (rule out disease). Or the probabilty of testing neg if the disease istruly absent
|
|
|
What is a "gold standard test"?
|
A definitive test used to determine if an individ has a disease
|
|
|
How do you calculate sensitivity?
|
TP/(TP+FN) or a/(a+c)
|
|
|
How do you calc specificity?
|
TN/(FP+TN) or d/(b+d)
|
|
|
What is an ideal test?
|
Doesn't exist. 100% sensitive; All individ w/ disease would be identified as + & No false neg. 100% specific; all individ w/o disease would be identified as neg & no false +
|
|
|
How do you minimize false +?
|
maximize specificity
|
|
|
How do minimize fale neg?
|
Maximize sensitivity
|
|
|
Are biologic variables dichotomous or continuous variables?
|
Continuous. Need cut pt to categorize persons appropriately
|
|
|
What are the 2 possible distributions of continuous variables?
|
Bimodal (1peak for normal & 1 peak for disease) & Unimodal (no clear cut point b/w normal and abnormal
|
|
|
What is themost common distribution of continuous variables?
|
Unimodal
|
|
|
What is the basis for choosing a cut point?
|
Use biologic info but often we don't know
|
|
|
What factors influence cut point?
|
Specificity & sensitivity
|
|
|
What is positive predicitve value (PPV)of a test?
|
The proportion of patients who test positive & actually HAVE the disease.
|
|
|
What is a Negative Predictive Value (NPV) of a test?
|
The proportion of patients who test neg & actually DO NOT HAVE the disease.
|
|
|
What are the charac of predictive values?
|
Depends on incidence, inc w/ disease prevalence in pop given same sensitivity/specificity, important in targeting high risk pop, results must be interpreted in context of prevalence
|
|
|
What is the importance of the relationship b/w PPV & Prevalence?
|
The higher the prevalence the higher the + predictive value, screening programs are most productive when in a high risk pop, results must be interpreted in context of prevalence
|
|
|
What factors affect reliability?
|
Intrasubject variation (diff in values of sam subject), Intraobserver variation, & Interobserver variation
|
|
|
What are the measures of reliability?
|
Overall percent agreement & Kappa Statistic
|
|
|
How can reliability be improved?
|
By standardization of prcedures & training
|
|
|
What question does the Kappa statistic answer?
|
To what extent does the agreement b/w 2 observers exceed the level of agreement that would result form chance alone?
|
|
|
What is the equation for kappa statistic?
|
(% observed agreement)-(% agreement expected by chance)/100%-(% agreement expected by chance)
|
|
|
What is the interpretation of kappa?
|
>0.75 Excellent, 0.40-0.75 Intermediate, <0.40 Poor
|
