Jump to content

Talk:Levine's sign

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

This

[edit]

http://www.med.uiuc.edu/m1/ihd/qafrost0405.htm mentions the history of the sign, but I could find nothing more on the good Dr Levine. JFW | T@lk 21:47, 17 May 2005 (UTC)[reply]

Sensitivity v Specificity

[edit]

The article states:

The sign has a sensitivity of 80% for myocardial infarction (MI), i.e. in patients with cardiac pain and a heart attack, Levine's sign was positive in 80%; however, it is not very specific, as the absence of the sign in patients with cardiac pain would falsely miss 51% (Edmonstone 1995).

To clarify, specificity means the degree to which the presence of the sign confirms the presence of the disease. If the sign is absent when the disease is present, it lessens the sensitivity, not the specificity. This paragraph contradicts itself (or is written in a confusing way). MoodyGroove 06:03, 5 February 2007 (UTC)MoodyGroove[reply]

Removed entire section

[edit]

It's unencyclopedic and reads like an article citique rather than an article about Levine's sign. In addition, it seems to be confused with regard to sensitivity and specificity. I'm placing a stub template on the article and putting it on my watch list. I will try to improve it as time allows. MoodyGroove 02:10, 3 May 2007 (UTC)MoodyGroove[reply]

The sign has a sensitivity of 80% for myocardial infarction (MI), i.e. in patients with cardiac pain and a heart attack, Levine's sign was positive in 80%; however, it is not very specific, as the absence of the sign in patients with cardiac pain would falsely miss 51% (Edmonstone 1995). Whether or not the sign is present in all cases of MI does not refer to specificity, but to sensitivity. Specificity refers to how often the sign is present in diseases other than MI (ex: chocking,...). If you go on to actually read this reference, you will find that it refers to a 1992 British study of 203 patients consecutively admitted to a CCU in whom it was recorded how they placed their hands on their chest to describe their chest pain, whether they had a clenched fist over the chest, or a flat hand over the chest, etc. The study then did a chart review a year later to find out how many of those patients actually had a cardiac cause for their pain. In this small study, 138 patients were considered to have a cardiac cause for chest pain, and 44 had an uncertain cause for chest pain (with possible coexisting gastroesophageal disease and cardiac ischemia). Only 19 patients with a cardiac cause for chest pain actually had a true Levine sign, with the clenched fist to the center of the chest, but 64 patients with a cardiac cause used a flat hand placed to the center of the chest. Of the 21 patients with non-cardiac chest pain, 4 had a positive Levine sign with a clenched fist, and 4 patients with uncertain cause also had a positive Levine sign. Thus, a true positive Levine was recorded in 19 patients with cardiac chest pain but also in 8 patients who may not have had true cardiac chest pain. If you were going to criticize this study, some criticisms might be that study didn't really distinguish the patients by diagnosis very well, and the numbers of patients weren't very high. The only statistically significant result in the study is that 110 patients used their hands somehow in reference to their chest (either with levine's sign, hands flat, or hands drawn laterally from the chest), which was statistically significant when compared to the fact that only 33 with either unknown pain or non cardiac pain did so. Thus, it is inappropriate to say from this study that the true Levine sign with a clenched fist has a sensitivity of 80%, because the study included some patients with their hands on their chests who didn't really have a true Levine sign.

illustration?

[edit]

you know what this needs? an illustration. I can think of a dozen different motions that'd fit the description of 'fist clenched against chest' - is Levine's sign left purposely vague, or are there just no graphics representing it? MattLohkamp 09:10, 30 December 2008 (UTC)