TORONTO -- Authorities in Saudi Arabia are investigating whether there is a new cluster of MERS coronavirus cases linked to a hospital in the eastern part of the country, a separate incident from a previously reported large hospital outbreak there.

A case reported Wednesday and five reported Tuesday may be linked to a hospital, said the country's deputy minister of health, Dr. Ziad Memish. He would not reveal the name of the town or the hospital.

Memish suggested there may have been some person-to-person spread among the cases. Three of these patients have died.

"They're from the same town and from the same hospital," Memish said in an interview with The Canadian Press, adding the investigation into how the people became infected is ongoing.

"We're still working on it. But none of the patients from yesterday had any relation to al-Ahsa," he said, referring to the large outbreak linked to the al Moosa hospital in the town of al-Hofuf that began in April but only came to light in early May.

"They actually live outside of al-Ahsa and they have no family contact in al-Ahsa. And we think that there was a community introduction and there was transmission from that community introduction."

Two of the patients shared a hospital room, he said, adding that it's not clear at this point whether the other patients are linked, and if they are, how.

"These patients came in and out a few times to the facility," Memish said. "So trying to find out who are the RTs (respiratory technicians) who looked after them, who are the nurses, who are the physicians, who took blood from them, is it the same person -- it's a very difficult task."

"But the team is on it 24-7."

Hospitals played a critical role in amplifying the SARS outbreak in 2003, with unidentified cases infecting other patients, family members and hospital staff. The World Health Organization has been keeping a close eye on any suggested hospital outbreaks, fearful that the pattern might be repeated with MERS. SARS and MERS are from the same viral family, coronaviruses.

The MERS virus -- the newly coined acronym is short for Middle Eastern respiratory syndrome -- first came to the world's attention last September, though retrospective testing dates the earliest known cases to a hospital outbreak in Jordan in April 2012.

To date there have been 50 confirmed cases, 27 of them fatal. Saudi Arabia has reported 38 of those cases, and 18 of the deaths.

Cases have also been reported from the United Arab Emirates (UAE), and Qatar. Infections have also been detected in Germany, France, Britain and Tunisia, though in all those instances the cases originated from countries in the Arabian Peninsula. In Britain, France and Tunisia there was limited person-to-person spread from the case that introduced the virus from abroad.

Doctors who treated France's two patients warned Wednesday that the incubation period of the new virus may be longer than previously appreciated. The WHO currently estimates the incubation period -- the time from exposure to illness -- at 10 days, though its scientists readily admit that given the paucity of data on cases, that's a guesstimate based on the incubation period of the SARS virus and other viruses.

The French doctors reported in the journal The Lancet that a man in Northern France who caught the virus from a patient infected on a trip to the UAE may have gone 12 days before developing symptoms. The two shared a hospital room for three days in late April, before the first man was diagnosed with MERS. That man died Tuesday from his illness.

The authors suggested doctors who are on the lookout for cases should test people who become ill with MERS-like symptoms within 12 days of travel to countries where the virus is found, or within 12 days of contact with a known case.

Dr. Allison McGeer, a Toronto infectious diseases expert, said that fits with what was seen with SARS. Most people who came down with the virus developed symptoms within 10 days of exposure but some had a longer incubation period, she said.

Knowing the incubation period is critical to detecting cases and for trying to stop spread of the disease. During SARS, people who were exposed to known cases had to go into quarantine for 10 days to try to cut chains of transmission. At points in the Toronto outbreak thousands of people were holed up in their homes waiting to see if they would develop the disease.

"The amount of disruption that you cause to people's lives and the amount of work you have to do to identify people who may pose a risk or who may be getting sick is driven entirely by knowing what the incubation period is and when people are infectious," McGeer said.

Dr. Anthony Mounts, the WHO's point person for MERS, said the agency will probably look at whether it needs to change its surveillance recommendations to include a lengthier incubation period.

"I think we should consider it," he said from Geneva.

"The 10 days that we had in there was really based on expert opinion from experience with other viruses. So I think as more information comes in about this virus, we'll definitely want to revisit that."

Memish said the Saudi Arabian experience suggests the incubation period can be as long as 14.5 days.

The French study also suggested doctors trying to diagnose suspected MERS cases may need to get specimens, like sputum samples. Their second case tested negative when swabs were run along the inside of his nasal passages, a common way to test for respiratory viruses. But lung samples came back positive.

The WHO had already advised doctors that if they have a strong suspicion a patient has MERS but nasal swabs come back negative that they should get specimens from the lungs if they can. "The sensitivity out of the upper respiratory specimens looks like it's probably not as high as it is for lower," Mounts said.

In other MERS news, Memish and colleagues from the Saudi ministry of health and several institutions in London reported on a cluster of cases that happened in Saudi Arabia last fall. Four members of a large extended family became ill, and two died. Three of the four tested positive and the fourth, a teenaged survivor, is considered a probable case.

The report was published in the New England Journal of Medicine.

There may have been person-to-person spread in this cluster of cases, the authors said. In fact, there may have been three generations of spread -- from the initial case, a 70-year-old patriarch of this large family to two of his sons and a grandson. But because the source of the virus is still not known, they cannot rule out the possibility that all four contracted the virus from a source in nature.

The authors did note that the family members hadn't travelled before becoming ill, kept no pets and had no contact with animals before the illness struck the family. Interestingly 24 other members of this large household did not become ill.