Biostatistics Q.......2 - psychmledr

[ArchivalUser]

tc10,
You are very welcome .
It is good to "see" you again

[ArchivalUser]

Bump

[ArchivalUser]

thanks mledr. brilliant question.
is there a shorter way of arriving at answer? creating table looks like it can take me over a minute if this were to appear in an exam

[ArchivalUser]

here is the formula
PPV=sensitivity*prev/sensitivity*prev + (100-specificity)*(100-prevalence)
NPV=specificity*(100-prevalence)/specificity*(100-prevalence) +(100-sensitivity)*prevalence

http://www.usmleforum.com/showthread.php?tid=753272-1

[ArchivalUser]

Try this Qs

Q1) In a poor, small village you use a new, cheap lab test X to rule out the possibility of TB. A pt from this village has persistent cough and comes to your temporary clinic to check for TB. His result is negative. Test X has a specificity of 80% and a sensitivity of 90% when tested in 100 subjects with TB and 100 subjects without TB in this village. What is the probability that this patient truly does not have TB?
A. 18%
B. 53%
C. 66%
D. 82%
E. 89%

Q2) In a poor, small village a cross-sectional study with a viral RNA shows that 20% of the people are HIV positive. You discover a new cheap diagnostic kit to test for HIV in this village. this test has a specificity of 80% and a sensitivity of 90%. This test was positive in one of the patients from the village. What is the best estimate of the likelihood that this patient has actually HIV?

[ArchivalUser]

Q1) E
Q2) 53%