How to read a paper: Papers that report diagnostic or screening tests
BMJ 1997; 315 doi: https://doi.org/10.1136/bmj.315.7107.540 (Published 30 August 1997) Cite this as: BMJ 1997;315:540All rapid responses
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 or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
Dear Sir, Madam,
I read Dr Greenhalgh's interesting paper about diagnostic tests with
great pleasure. However, I must note that the likelihood ratio is not
equal to the pre-test post-test ratio for the probability of having a
disease. As the pre-test probability of having a disease is equal to the
population prevalence, and the post-test probability (probability of
having a disease after testing positive) is equal to the positive
predictive value of a test, the pre-test post-test ratio is equal to the
positive predictive value of the test divided by the prevalence of the
disease in the population. The outcome of this division is mostly not
equal to the likelihood ratio.
Yours Sincerely
Daan Uitenbroek
Hilversum
The Netherlands
Competing interests: No competing interests
In this otherwise excellent contribution Trisha Greenhalgh has fallen
into a common pitfall. That is of generalising from a particular case and
drawing conclusios from the performance of one type of health delivery
system to that of another.
Contrary to the Dutch experience neonatal screening for congentital
hypothyroidism has a very high specificity in the UK at the cost of a
slightly decreased sensitivity.
The Dutch have a complex system of neonatal screening involving the
measurement of TSH, T4 and TBG. This is because they inisist on having a
system that will miss no cases i.e they are really carrying out population
based diagnosis rather than screening. In most of Europe including the UK
and throughout almost all of the rest of the screening world TSH only is
used for screening. This results in a very high sensitivity and
relatively few false positives at the cost of missing patients with
secondary hypothyroidism due to pituitary or CNS problems. These latter
cases usually present with other clinical problems and are less than 5% of
all cases of neonatal hypothyroidism
There is a problem with false positives in neonatal screening as
perceived by some who consider it a major factor in evaluating potential
new programmes. Currently (2001) it is an issue in the proposed screening
for MCAD and cystic fibrosis. It is important therefore that the low
numbers of false positives in the UK in current neonatal screening is
acknowledged and the problem not exagerated by those with the power to say
yes or no to new developments.
Competing interests: No competing interests
Using the likelihood ratio in diagnostic tests
Dear Sir
The correct approach to using the likelihood ratio in a diagnostic
test is to use the equation that
post test odds ratio = likelihood ration x pre test odds ratio
rather than the approach used in Greenhalgh’s paper (1).
This correction was incorporated into later editions of Greenhalgh’s
book (2). Unfortunately she never saw the need to correct her earlier
statement using bmj.com. It was certainly disappointing to see that the
BMJ referee did not spot a reference to a probability greater than 1 in
the paper (1), which indicated that something was wrong.
Yours faithfully
Alan C C Gibbs
References:
1. Greenhalgh T. Papers that report diagnostic of screening tests.
BMJ 1997; 315: 540-43.
2. Greenhalgh T. How to read a paper, 2nd ed. London: BMJ; 2001.
Competing interests:
None declared
Competing interests: No competing interests