Amit
I discovered the same at the 2nd Nordic wshop i EBHC last year in Oslo and
we wrote a letter which you may find via my EBM website or directly at
http://www.ama-assn.org/sci-pubs/journals/archive/jama/vol_278/no_10/letter_
3.htm
and the authors revised the table of LRs.
Often these LRs are based on mutliple logistic regression analyses, which
may be the explanation of why you couldn't do the simple maths....
Atle
At 23.10.98 12:37 -0500, amit kumar ghosh wrote:
>I have a question on the The Rational Clinical examination-series,
article: Does
>the patient have appendicitis? (JAMA 1996; 276:1589).
>Table 3. indicates the for Right lower quadrant pain the sensitivity was
81% and
>specificity was 53%. The likelihood ratio + is reported as 7.31-8.46.
>However my calcualtions indicate a LR + of 1.72. Any reason why I am of
mark by
>X 6 ?
>Also I stumbled upon 2 sites which does the LR , NNT math and many more.
>http://hiru.mcmaster.ca/ebm/userguid/2_tools.htm
>http://hiru.mcmaster.ca/ebm/userguid/3_tools.htm
>
>
>Amit Ghosh, MD
>Assistant Professor
>Internal Medicine
>Box 741
>University of Minnesota
>Minneapolis, MN 55455
>[log in to unmask]
>Fax: 612-624-3189
>
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|