Antibiotics are lifesaving drugs, but overuse is leading to one of the world's most pressing health threats: antibiotic resistance. Researchers are developing a tool to help physicians prescribe antibiotics to patients who really need them, and avoid giving them to individuals who don't.

Scientists identified 11 genetic markers in blood that accurately distinguished between viral and bacterial infections (antibiotics help us fight bacterial infections, but aren't effective and shouldn't be used to treat viruses). The finding, published in the journal Scientific Reports, is important because physicians don't have a good way to confirm bacterial infections like pneumonia and more-often-than-not default to an antibiotic.

"It's extremely difficult to interpret what's causing a respiratory tract infection, especially in very ill patients who come to the hospital with a high fever, cough, shortness of breath and other concerning symptoms," said lead study author. "My goal is to develop a tool that physicians can use to rule out a bacterial infection with enough certainty that they are comfortable, and their patients are comfortable, foregoing an antibiotic."

A group of 94 adults hospitalized with lower respiratory tract infections were recruited to participate in Falsey's study. The team gathered clinical data, took blood from each patient, and conducted a battery of microbiologic tests to determine which individuals had a bacterial infection (41 patients) and which had a non-bacterial or viral infection (53 patients).

Gene expression was measured using RNASeq and qPCR. Genes discriminatory for bacterial infection were identified using the Bonferroni-corrected Wilcoxon test. Researchers used complex 90 genetic and statistical analysis to pinpoint markers in the blood that correctly classified the patients with bacterial infections 80 to percent of the time.

"Our genes react differently to a virus than they do to bacteria," said another author. "Rather than trying to detect the specific organism that's making an individual sick, we're using genetic data to help us determine what's affecting the patient and when an antibiotic is appropriate or not."

Authors say that the main limitation of their study is the small sample size and that the genetic classifiers selected from the study population may not prove to be universal to all patients. A patent application has been filed for their method of diagnosing bacterial infection.

https://www.urmc.rochester.edu/news/story/5108/urmc-researchers-developing-new-tool-to-fight-antibiotic-resistance.aspx

https://www.nature.com/articles/s41598-017-06738-3