(A) Comparison of EBCD results with clinical tests for COVID-19. Results are shown for 60 patients who used the EBCD to produce a sample after they were tested for COVID-19 using nasal swab samples and the Abbott ID NOW nucleic acid-based rapid diagnostic test (n = 42), or RT-qPCR (n = 18). 44 patients tested positive on the clinical test (blue dots) and 16 patients tested negative on the clinical test (red dots). EBCD results are plotted as the number of exhaled SARS-CoV-2 RNA copies over the 10-minute breathing session. (B) Dilution series. Dilution series were used to convert cycle threshold values to viral copies. (C) Identification of SARS-CoV-2 variant by clade for participant samples over time. Background colors represent SARS-CoV-2 variant clade percentage in Chicago between April 2020 and February 2022. Dots represent peak amount of viral RNA detected on breath for each COVID-19-positive participant. Participant number, putative clade and percentage likelihood are indicated next to each dot; +, clade identified by genome sequencing; vertical lines indicate day of symptom onset (n = 42) or date of diagnosis (n = 2; P15 and P43). (D) Peaks levels of viral RNA shedding on breath across participants, separated by variant and sorted by exhaled viral RNA copy number. Black horizontal lines indicate average within each variant type. Red, pre-Delta; Gray, Delta; Blue, Omicron.

COVID-19-positive participant demographics.

(A) Levels of exhaled SARS-CoV-2 RNA copies per minute over the course of infection. Each dot represents the daily average number of exhaled SARS-CoV-2 RNA copies exhaled per minute, for each participant. The overlaid black line represents the rolling average of exhaled viral RNA copies over days since symptom onset. The overlaid red line is the fitted S curve (R2 = 0.7657, adjusted R2 = 0.7567, RMSE: 0.3755). The green vertical dashed line indicates the inflection point of the S curve. (B) Viral RNA copy numbers binned over days since symptom onset. Each dot represents results from one sample. Mean and standard error are represented by red lines within each bin. * indicates P < 0.05. (C) The same data from panel A are plotted, with colored dots representing samples that were collected from vaccinated participants, and grey dots representing unvaccinated participants. (D) The same data from panel A are plotted with blue indicating participants infected with Omicron strains; red indicating infection with Delta strains; and grey indicating infections prior to Delta. All panels: blue horizontal line represents 35 CT equivalent.

Levels of SARS-CoV-2 RNA copies in exhaled breath for (A) vaccination status, (B) viral variants and (C) self-reported symptom severity. Each colored dot represents one sample, and the color scale represents days since symptom onset. Black bars denote average, error bars denote standard error, and blue horizontal line represents 35 CT equivalent. * indicates P < 0.05.

Exhaled Breath Collection Device Parts and Use. (A) Exhaled Breath Collection Device parts include a 50 CC catheter-tip syringe plunger and body, with tip trimmed to approximately 6 mm length (Monoject, Cardinal Health, top panel); a mouthpiece and one-way flutter valve (Vyaire, middle left); a cooling sleeve (Torex, middle right); and a cloth insulating cover (Surgilast, lower left). The assembled Exhaled Breath Collection Device is shown lower right; the device can be fitted with a mask cover over the narrow end by cutting a 10 cm square from a surgical mask and taping it over the end of the Exhaled Breath Collection Device (bottom) for use in clinical settings. (B) Exhaled Breath Collection Device in use. The user performs a breath condensate collection session by placing the mouthpiece in their mouth, breathing at a natural pace, and exhaling through the device at a natural resting pace. (C) Recovering a sample from the Exhaled Breath Collection Device. Once the breathing session is complete, the mouthpiece and cooling sleeve assemblies are removed, and the plunger is inserted in the wide end of the syringe body (top). The condensate sample is then recovered from the syringe body by pushing the plunger through the syringe body, forcing the liquid into a sample vial, which contains MTM (Primestore). Sample vials can be mailed for lab analysis (bottom).

Shear rate vs. shear stress plots of condensates (nasal, red, oral, blue), saliva (green) and water (black), suggesting minimal contamination from saliva. Error shades correspond to 95% confidence levels. Nasal and oral samples were indistinguishable from each other and from water, strongly suggesting that oral breath samples collected with EBCD are not contaminated by saliva, and rather represent a pure exhaled breath condensate sample.

Detail of viral clades, peak exhaled viral load, number of samples and days since symptom onset for each participant.

Comparison of different models fitted to the average time course of viral RNA exhaled per minute.

Estimated marginal mean of the number of SARS-CoV-2 RNA copies (log-transformed) for each of the variant.

P values of post-hoc Tukey test on the number of SARS-CoV-2 RNA copies across variants.

Estimated marginal mean of the number of SARS-CoV-2 RNA copies (log-transformed) for each of the symptom severity groups.

P values of post-hoc Tukey test on the number of SARS-CoV-2 RNA copies across self-reported symptom severity.

Estimated marginal mean of the number of SARS-CoV-2 RNA copies (log-transformed) for each of bin of days.

P values of post-hoc Tukey test on the number of SARS-CoV-2 RNA copies across days since symptom onset.