The term “antibiotic resistance” doesn’t have the fear-inducing power of an ailment like cancer or AIDS, but it should. Experts say dealing with bacteria that doesn’t respond to drugs should be on the top of our to-do list — or they will kill more people than cancer by 2050. Both overprescribing and feeding livestock antibiotics have contributed to the growing problem, so now researchers are looking for ways to reduce their use to preserve those that are still useful and prevent a total public health emergency. One Chinese team has developed a test that could determine when antibiotics are actually necessary just by analyzing a patient’s breath.

Researchers based at Zhejiang University have initiated a pilot study with the aim of creating an efficient, painless, and accurate test for clinicians to determine the necessity of antibiotic treatment for a patient. If doctors were given this kind of certainty, they could avoid prescribing the medicines to patients who didn’t need them, slowing antibiotic resistance.

For their research, the team looked at pneumonia patients in the intensive care unit. These participants had a special type of the illness called ventilator-associated pneumonia, which occurs in those who are on breathing machines in the hospital, a population in which it is critically important to identify any life-threatening bacterial infections. However, common colonizations can sometimes be mistaken for more sinister bacteria, leading to overmedication.

“To confirm whether patients have a bacterial infection of the respiratory tract, doctors currently have to take a number of different samples (blood and septum) and even chest X-rays in the case of pneumonia,” said project coordinator Kejing Ying in a statement.

After analyzing biomarkers in the breath samples of 60 patients, the scientists found a link between bacterial pneumonia and the presence of certain volatile organic compounds (VOCs) produced by a pathogen called A. baumannii.

The study is in no way conclusive. “The challenge we face is that many VOCs are not unique to one pathogen,” Ying added.

However, more studies could provide a useful library of sicknesses doctors could detect on the breath. The authors wrote that “this pilot study mainly provides a proof of concept that the direct detection of exhaled A. baumannii-derived VOCs might be adopted for an early alert” of bacterial presence in ICU patients, but “further refinement and validation are required before clinical use.”

Source: Gao J, Zou Y, Wang Y, Wang F, Lang L, Wang P. Breath Analysis for Noninvasively Differentiating Acinetobacter baumannii Ventilator-associated Pneumonia from its Respiratory Tract Colonization of Ventilated Patients. Journal of Breath Research. 2016.