Authors’ replyBMJ 2009; 339 doi: https://doi.org/10.1136/bmj.b2834 (Published 15 July 2009) Cite this as: BMJ 2009;339:b2834
All rapid responses
The author’s response , paper and the letters illustrate why the
process of diagnosis eludes critical evaluation . The indices of test
assessment used by the authors were the sensitivity and specificity.
These were designed to assess a test’s ability to screen asymptomatic
populations, a function that the TYM test is not designed to do .
However, tests can also be assessed for their performance as diagnostic
leads, as differentiators between differential diagnoses suggested by a
lead, as confirmers of diagnoses and as treatment indications .
The authors state the TYM test should only be done if a patient
presents with a cognitive problem and that if the score is low then a
cause needs to be found . It is therefore being proposed as a
diagnostic lead, which means that we need to know the differential
diagnosis of the various results it produces and the proportion of
patients with each of these differential diagnoses . The TYM is not
apparently designed to confirm a diagnosis or to suggest a treatment.
The TYM may be a helpful diagnostic differentiator however. If a
patient presents with an apparent cognitive problem and the TYM score is
47/50 or more then it might be a helpful differentiator between some of
the differential diagnoses of that cognitive problem. Does a high TYM
score occur rarely in all causes of dementia but commonly in some other
(non-dementing) causes of cognitive impairment? The ratio of such a pair
of likelihoods would allow the authors to assess the ability of a high TYM
score to act as a differentiator between such differential diagnoses of
cognitive impairment .
1. Brown J, Pengas G, Dawson K, Brown LA, Clatworthy P. TYM and
Alzheimer’s disease, Authors’ reply. BMJ 2009;339:b2834
2. Godlee F. Editor's Choice. Critical thinking BMJ 2009; 338:b1662.
3. Llewelyn D E H, Ang AH, Lewis K, Abdullah A. (2009) The Oxford
Handbook of Clinical Diagnosis, 2nd edition. Oxford University Press,
Competing interests: No competing interests