News Release

Study finds exhaled breath could enhance detection, diagnosis of COVID-19 and variants

Research suggests volatile organic compounds in breath could mark distinction between COVID-19, variants and non-COVID illnesses

Peer-Reviewed Publication

Michigan Medicine - University of Michigan

The emergence of new COVID-19 variants has led to reduced accuracy across current rapid testing methods, but a recent University of Michigan study suggests that a patient’s breath might hold the key to a more precise diagnosis.

Investigators from the University of Michigan’s Max Harry Weil Institute for Critical Care Research and Innovation, including faculty and students from the College of Engineering and Michigan Medicine, used portable gas chromatography to examine breath samples collected during the pandemic’s Delta surge and its transition to Omicron (from April 2021 to May 2022.)

Their results, published February 28 in JAMA Network Open, showed that the GC technology could diagnose COVID-19 with a high level of accuracy. They also revealed that the volatile organic compounds contained in the breath of patients with Omicron differed from those in patients with Delta and earlier variants—molecular-level differences which, according to the team, could potentially be used to distinguish between COVID-19, its variants and non-COVID illnesses.

“Exhaled breath contains hundreds of VOCs, which the body produces in response to infection and inflammation,” said principal investigator and study author Xudong (Sherman) Fan, Ph.D., Richard A. Auhll Endowed Professor of Biomedical Engineering and associate director of the Weil Institute. “Early in the pandemic, we used GC technology to discover and define sets of VOCs for detecting COVID-19. However, we needed to gain a better understanding of how dynamically emerging variants impact this technology.”

Supported by a $2 million grant from the National Institutes of Health Rapid Acceleration of Diagnostics initiative’s Screening for COVID-19 by Electronic-Nose Technology program, the team conducted a diagnostic study of 167 adult patients in the Michigan Medicine ICUs and emergency department. They collected 205 breath samples from symptomatic and asymptomatic patients in 3 cohorts:

  • COVID-19 (2021): Patients with COVID-19 who were recruited before December 14, 2021 and were assumed to be infected by the Delta or earlier variants
  • COVID-19 (2022): Patients with COVID-19 who were recruited from January 2022 to the end of May 2022 and were assumed to be infected by the Omicron variant
  • Non-COVID-19 illness: Patients who were COVID-19 negative at the time of breath analysis, as well as patients who were previously COVID-19 positive but had recovered

Using a novel point-of-care GC device developed by Fan and the team, in combination with an advanced biomarker discovery algorithm and data analysis platform developed at the College of Engineering and the Weil Institute, the investigators defined four sets of VOCs that were able to distinguish between COVID-19 (2021) and non-COVID illness with a sensitivity of 92.7%, a specificity of 95.5% and an accuracy of 94.7%. However, when the team applied the same VOCs in a setting of presumed Omicron, sensitivity decreased drastically to 60.4%.

“We already knew clinically that different strains of SARS-CoV-2 can act quite differently,” said study co-author Robert Dickson, M.D., associate professor of Pulmonary and Critical Care Medicine and deputy director of the Weil Institute. “This decrease in performance supports our suspicion that their effects on lung biology are quite different as well.”

Based on their findings, the team hypothesized that it could be possible to use breath analysis to distinguish between COVID variants. They undertook additional biomarker searches and defined new VOCs to discern between Omicron and Delta, Omicron and non-COVID illness, and between patients with COVID-19 and non-COVID illness regardless of variants. The combined analysis resulted in the ability to detect COVID-19 infected patients (regardless of variant) from non-COVID patients with a sensitivity of 89.4%, a specificity of 91.0% and an accuracy of 90.2%.  This performance is close to that of RT-PCR tests (the gold standard) and better than many rapid antigen tests.

Co-author and co-principal investigator Kevin Ward, M.D., professor of Emergency Medicine and Biomedical Engineering and executive director of the Weil Institute said, “This work suggests that breath analysis using point-of-care GC may be a promising method for detecting COVID-19 and similar diseases that result in VOC production. However, as we are seeing with other detection and testing methods, the emergence of viral variants continues to pose challenges.” The team notes that further analysis will be needed to determine how to overcome these challenges and use breath analysis to improve the diagnosis and care of patients.

“The fact that we were able to diagnose COVID-19 in both symptomatic and asymptomatic participants is encouraging,” said Fan. “More studies on these particular VOCs, including their origins, may assist in the development of a better understanding of COVID-19 as well as the potential to develop new diagnostics.”

“Looking ahead, our diagnostic approach to COVID-19 and the lung injury it causes will need to be as dynamic as the virus itself,” said Dickson.

The portable GC technology has now been licensed from the University of Michigan by the company Blu Biotech for its commercialization. The company is looking to develop the technology for the diagnosis and monitoring of not only COVID-19 but other diseases such as sepsis, acute respiratory distress syndrome, inflammatory bowel disease and others that produce VOCs due to inflammation.


Paper cited: “Portable Breath-Based Volatile Organic Compound Monitoring for the Detection of COVID-19 During the Circulation of the SARS-CoV-2 Delta Variant and the Transition to the SARS-CoV-2 Omicron Variant,” JAMA Network Open. DOI: 10.1001/jamanetworkopen.2023.0982

Study authors include Ruchi Sharma, PhD (Weil Institute, Biomedical Engineering); Wenzhe Zang, PhD (Weil Institute, Biomedical Engineering); Ali Tabartehfarahani, PhD (Weil Institute, Biomedical Engineering); Andres Lam, MSc (Weil Institute, Biomedical Engineering); Xiaheng Huang, MSc (Weil Institute, Biomedical Engineering, Electrical Engineering & Computer Science); Anjali Devi Sivakumar, BTech (Weil Institute, Biomedical Engineering, Electrical Engineering & Computer Science); Chandrakalavathi Thota, PhD (Weil Institute, Biomedical Engineering); Shuo Yang, PhD (Weil Institute, Biomedical Engineering); Robert Dickson, MD (Weil Institute, Pulmonary & Critical Care Medicine); Michael Sjoding, MD (Weil Institute, Pulmonary & Critical Care Medicine); Erin Bisco, BA (Weil Institute, Emergency Medicine); Carmen Colmenero Mahmood, BSc (Weil Institute, Emergency Medicine); Kristen Machado Diaz, BSc (Weil Institute, Emergency Medicine); Nicholas Sautter, BSc (Weil Institute, Electrical Engineering & Computer Science); Sardar Ansari, PhD (Weil Institute, Emergency Medicine); Kevin Ward, MD (Weil Institute, Emergency Medicine, Biomedical Engineering); Xudong Fan, PhD (Weil Institute, Biomedical Engineering)

Disclosures: The University of Michigan is a partial owner and  Xudong Fan is an inventor of technology being used in this research project that is licensed to Nanova Environmental, Inc., a company in which Fan serves as an outside consultant unrelated to this project. Technology is also optioned to the Beijing Institute of Collaborative Innovation, a study sponsor.

Study researchers, Xudong Fan, Kevin Ward, Ruchi Sharma and Wenzhe Zang, are inventors of a technology being used in this project that is optioned to a company called BlueBiotech, Inc., which is partially owned byFan and Ward. The company was created to commercialize this technology. It is possible that the University of Michigan, Nanova Environmental, Inc., BICI, Blue Biotech, Inc., or the researchers may one day benefit financially from the results of the study.

Michael Sjoding and Sardar Ansari previously received grants from the NIH outside this project.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.