Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles.
The authors comment the transfer of estimates of test accuracy(1). In
figure 3 they state that oddsratio, sensitivity, specificity and the
predictive values are all measures of test's discriminatory power. This
statement is true ,but a proof is not given.
The diagnostic oddsratio can be described as the quotient of the products
of nosological true and false numbers in a symmetrical four field table.
For the diagnostic oddsratio it is allowed to use row proportions only or
column proportions only. The oddsratio can be expressed as follows:
With these formula in mind it is possible to get an idea of
comparing the different estimates of test accuracy . (2) Only pairs of
the diagnosed proportions (Se, SP) or diseased proportion (PPV ,NPV) can
be interpreted because for every Se or PPV the complement can be
chosen.Then the oddsratio=1. In this case there is no association between
the result of the test and the nosology. With the design of a study not
only Se or PPV must be measured.
1. Les Irwig, professor a, Patrick Bossuyt, professor of clinical
epidemiology b, Paul Glasziou, professor of evidence based practice c,
Constantine Gatsonis, professor d, Jeroen Lijmer, clinical researcher b.
Designing studies to ensure that estimates of test accuracy are
transferable BMJ 2002;324:669-671 ( 16 March )
2. Lijmer JG, Mol BW, Heisterkamp S, Bonsel GJ, Prins MH, van der Meulen
JH, Bossuyt PM.Empirical evidence of design-related bias in studies of
JAMA. 1999 Sep 15;282(11):1061-6.