Monthly Archives: June 2011

Here’s why that would be extremely useful : For example, we know that 8-9 out of 10 ear or throat infections are caused by viruses, and that antibiotics will not usually make the symptoms of these go away any quicker, while they may still have potential side effects, and add to the growing pool of antibiotic-resistant bacteria. But you want to be able to tell the 1 out of 10 that has a bacterial infection, and the same applies to chest infections. Most people who cough and have a fever just have some viral illness, and don’t need antibiotics, but some people do have bacterial infections, and especially in children there is a danger that these patients get really really sick.
At the moment we rely on blood tests like the C-reactive protein or the white cell count to give us an idea whether a bacterial infection could be present, but these markers are far from perfect, and their positive predictive value is rather average.

So a new study by researchers from Israel is suggesting a quick test based on certain signatures of circulating phagocytes, which are essentially a bunch of early response immune cells who eat invading bugs :

Polymorphonuclear leukocytes (PMNs) or phagocytes act as major players in the defense response of the host during an episode of infection, and thereby undergo functional changes that differ according to the infections. PMNs functional activity can be characterized by quantification and localization of respiratory burst production and assessed by chemiluminescent (CL) byproduct reaction. We have assessed the functional states of PMNs of patients with acute infections in a luminol-amplified whole blood system using the component CL approach. In this study, blood was drawn from 69 patients with fever (>38 °C), and diagnosed as mainly viral or bacterial infections in origin. Data mining algorithms (C4.5, Support Vector Machines (SVM) and Nave Bayes) were used to induce classification models to distinguish between clinical groups. The model with the best predictive accuracy was induced using C4.5 algorithm, resulting in 94.7% accuracy on the training set and 88.9% accuracy on the testing set. The method demonstrated a high predictive diagnostic value and may assist the clinician one day in the distinction between viral and bacterial infections and the choice of proper medication.

So in essence, those phagocytes light up differently when dealing with bacterial or viral intruders, and this test showed a pretty good predictive value for telling one from the other, far better than our current tests anyway. It’s got potential, me want !