Skip to main content
Premium Trial:

Request an Annual Quote

Thai Researchers Developing Test to Detect COVID-19 Through its Smell


NEW YORK – A team of researchers from Chulalongkorn University in Bangkok has drawn on mass spectrometry to develop a sweat-based SARS-CoV-2 test.

Inspired by a separate project at the university using COVID-19 "sniffer dogs" to smell sweat samples and detect asymptomatic cases of SARS-CoV-2, the Chulalongkorn team created an analyzer and test to do the same — minus the dog.

When someone is infected with SARS-CoV-2, the bacteria present on the body that feed on sweat react abnormally to the virus and produce distinct odors unique to the infection, according to Chadin Kulsing, an assistant professor in the department of chemistry at the university.

He and his team looked at the collected samples to figure out which specific compounds the dogs detected that indicated infection. Once they determined the compounds the dogs were smelling, the team created a device to analyze sweat samples taken from people's armpits and detect those compounds.

By using a portable chemistry analyzer generally used to measure toxic chemicals in the environment and adding a pump to the instrument to filter out unwanted compounds, Kulsing and his colleagues came up with the Portable Sweat Test for COVID-19 Detection.

The sweat sample, collected with a cotton swab, is put into a bottle that is then placed into a chamber in the device. The chamber is then heated to evaporate the sample and increase the concentration of compounds so they're more easily detectable, Kulsing said.

Once the sample has evaporated, the operator uses a suction hose to draw the gas through the filter to a photoionization detector that shoots the sample with light. If any compounds are detected, an electrical signal will be produced that indicates a positive test result, Kulsing said.

"Because you've already filtered out the things you don't want at the beginning, anything passing through [the photoionization detector] is a key compound," Kulsing said, adding the team is exploring adding a chemical solution or solvent to the sample to enhance the signal and concentration of compounds.

While there's no published data yet, out of 322 positive cases and 3,680 negative cases — as determined by PCR — the test was able to identify 96 percent of positive cases and 99 percent of negative cases, Kulsing said.

He added that the team wants to test at least 10,000 samples before publishing its data, a milestone he expects to reach in the next few weeks.

Part of the issue, he said, is finding positive samples to test. He's currently using the device to screen people in the laboratory, which amounts to about 500 cases per day, but most of those cases are negative.

There are already plans for a next-generation version of the device, which will be smaller and more portable, Kulsing said. It'll be a box with a small microwave and a chamber where the sample can be placed. There will also be a mechanism to push the gas from the sample into the filter automatically, rather than having it manually extracted.

Sample detection takes around 30 seconds with the manual device, which costs around 210,000 baht ($6,213). The automatic device would have the same turnaround time and would cost 250,000 baht but would cut down on sample preparation time.

Currently, sample collection from a person's armpit and sample prep takes about 15 minutes, Kulsing said. However, the time of day when the sample is collected plays a factor, he said. If someone has recently showered and has less pungent body odor, the sampling can take up to an hour, which is why the team has largely been taking samples after lunchtime or when people disembark from a plane.

Further down the road, the team plans to convert the instrument to an even more portable pen or wand device that would have the sample manually loaded but would return a result in three seconds, he said. That device would cost less than 10,000 baht, Kulsing added.

He noted that the team is actively looking for a partner to help develop the wand device. Kulsing said he is also looking for funding to distribute the device more widely across Bangkok.

The first version is currently being used for free across the city, but Kulsing said he's open to working with partners to commercialize the next-generation versions of the device.

Sweat testing has been used for diagnosing cystic fibrosis and diabetes, but most groups have focused on noninfectious diseases when using sweat as a sample type. Intelligent Fingerprinting, a UK-based diagnostic company, uses fingerprint sweat to detect the presence of drugs of abuse but hasn't tackled a sweat-based test for SARS-CoV-2.

A team at the University of Texas at Dallas, meantime, has developed a sweat-based test to track someone's cytokine response, and earlier this year they published a proof-of-concept paper in Bioengineering and Translational Medicine.

Voravee Hoven, a professor in the department of chemistry at Chulalongkorn who wasn't involved in the development of the test, said that the benefit is that the test is noninvasive and "can be done very quickly" without the need for biochemical agents. She added that sweat does not contain the actual virus, so it is safer for people performing the test.

Hoven noted that user perception of the test is particularly important, as well. "We have to admit that the testing approach is relatively new," she said. "It remains a challenge to gain real trust from the users."