The research is revealed in an early-release article for the July edition of Emerging Infectious Diseases, the open-access and peer-reviewed journal published by the Centres for Disease Control and Prevention of the United States.

Family clusters of infections have helped the spread of a pandemic in which the world’s total cases are approaching 2 million and the death toll has passed 120,000.

The first patient of the 10 cases studied in Guangzhou had on January 23 returned from Wuhan, where the coronavirus was first reported in December.

The person had lunch with three family members the next day at a windowless restaurant with an air conditioner on each floor.

Two other families sat at neighbouring tables, with about one metre between each and an overlap in dining time of about an hour, the report said.

The first patient had a fever and cough later that day and went to hospital. Within two weeks, four further members of their family, three members of the second family and two of the third family had become ill with Covid-19, the disease caused by the coronavirus.

After detailed investigation, it was found that the only known source of exposure for the second and third families was the first patient in the restaurant.

“From our examination of the potential routes of transmission, we concluded that the most likely cause of this outbreak was droplet transmission,” the report said.

“We conclude that in this outbreak, droplet transmission was prompted by air-conditioned ventilation. The key factor for infection was the direction of the airflow.”

It said another 73 customers who dined on the same floor were identified as having close contact with the first patient, but were found to have no symptoms of Covid-19 during a 14-day quarantine and provided throat swab samples that tested negative. No staff working at the restaurant were found to be infected.

Six smear samples from the air conditioner’s outlet and inlet all tested nucleotide negative, it said.

“This finding is less consistent with aerosol transmission,” the report said. “However, aerosols would tend to follow the airflow, and the lower concentrations of aerosols at greater distances might have been insufficient to cause infection in other parts of the restaurant.

Coronavirus can travel twice as far as official ‘safe distance’, study says

“To prevent the spread of the virus in restaurants, we recommend increasing the distance between tables and improving ventilation.”

The Guangzhou team stated that the research had limitations because they did not conduct an experimental study simulating the airborne transmission route, nor perform antibody studies of swab sample-negative, asymptomatic family members and other diners to estimate the risk of infection from them.