Tuesday, April 04, 2006

Isabel and Dxplain tackle two more cases from the New England Journal of Medicine

(See my earlier post on this subject).

I managed to try the programs on two more NEJM cases before my Isabel free trial expired. Isabel outperformed Dxplain on this Clinical Problem Solving case of cholesterol atheroembolism. Presented with the case of a cachectic man with severe hypertension, renal failure, constitutional symptoms and atherosclerosis of the aorta both programs offered a long list of diagnoses but only Isabel included cholesterol atheroembolic disease. The atheroembolic syndrome is a great masquerader. I was a little disappointed in Dxplain. The presentation of an unexplained multisystem illness with renal failure and hypertension in a patient with atherosclerosis should always bring that diagnosis to mind.

The tables turned when I entered the CPC case of Brugada syndrome discussed elsewhere in this blog. Dxplain nailed it while Isabel struggled with numerous cardiovascular diagnoses, failing to mention the correct one. That concerns me. Given a couple of telltale ECG signs in a patient with cardiac arrest it should be a no brainer.

So Dxplain wins the best two out of three series. What can we conclude from this small sample? Perhaps only that both software programs are hit-and-miss. I’m more familiar with Dxplain, having toyed with it since the late 80s. It’s an unpretentious and longstanding work in progress of education and research, making no claims about saving lives. I have decided not to spend my medical group’s money on Isabel. I was a bit put off by its gaudy testimonials and hyped up claims.

No comments: