LONDON -- The overall death rate from SARS worldwide is about 15 percent -- double previous estimates -- according to the first in-depth global analysis by the World Health Organization.
Officials at the U.N. health agency said Thursday the calculation indicates the disease is more dangerous, particularly in the elderly, than experts had thought. More than half of SARS patients over 65 are dying.
In addition, a study to be published today suggests the SARS virus is surprisingly stable and not mutating significantly -- a finding that could indicate the virus may have been in humans for a while.
Until now, WHO had estimated the death rate to be between 6 percent and 10 percent. The new estimate was based on more complete and detailed data from China, Canada, Hong Kong, Singapore and Vietnam.
A study earlier this week estimated the death rate in Hong Kong to be about 20 percent overall, but about 50 percent in patients over 60.
"The global estimate we had previously was always clearly an underestimate," said Mike Ryan, who is coordinating SARS response at the World Health Organization.
"We always had a figure in our heads that was a range, but now we see that it's toward the upper end of that range. We didn't have a proper estimate and this is now much more accurate," he said.
Severe acute respiratory syndrome has made more than 7,000 people around the world ill and killed more than 500.
Until now, WHO has been making a crude calculation of the death rate by dividing the number of deaths by the number of SARS cases. Using that method, the death rate would be 7.1 percent, which underestimates the true death rate because it assumes everyone still hospitalized will survive.
Another method is to add up the number of people who have died and those who have recovered and divide the deaths into that total. However, that overestimates the death rate because deaths show up more quickly than recoveries and it ignores people who are hospitalized, Ryan said.
The new figure was reached by estimating what proportion of those hospitalized will recover and what proportion will die. It also takes into account the length of time patients have survived -- looking at the risk of dying in the first week of illness, the risk in the second week, and so on.
However, the figure is still only an estimate. The final death rate will be known only when the epidemic is over. Also, the estimate is for those diagnosed with SARS, not necessarily everyone infected with the SARS virus. Including such people would lower the death rate.
As suspected, the analysis indicated the chances of dying from SARS greatly depend on age and, linked to that, on the existence of other illnesses, Ryan added.
The death rate is below 1 percent for people aged 24 or younger, rising to 6 percent for those 25 to 44, 15 percent in those aged 45 to 64 and more than 50 percent for those over 65, WHO estimates.
"You really have to take into account the age groups because that will determine the outcome," Ryan said. "Some countries, like Canada, appear to have a high case fatality. In fact, that's more determined by the fact that their patient group was older.
"In Vietnam, their fatality rate was 8 percent. That's probably determined less by the quality of care but more by the fact that the majority of those cases were in young, healthy adults," he said.
By contrast, WHO said the death rate was 14 percent in Singapore and 15 percent in Hong Kong. The agency did not give a rate for China or Canada.
WHO experts said there is no evidence the SARS virus is becoming more lethal.
In a new study published Friday, meanwhile, scientists found few mutations among 14 SARS virus samples, indicating the genetic makeup has remained remarkably stable.
In the first major analysis of mutations of the SARS virus, posted on The Lancet medical journal's Web site, scientists from Singapore's Genome Institute identified genetic differences that reveal two geographically distinct strains of the virus -- signatures that could help in tracing where an individual's SARS strain originated.
The apparent stability of the virus is a double-edged sword, experts say. It does not seem to be mutating into a more virulent form, as feared. But it is also not becoming more benign.
"This virus is seeming quite constant as it is passing through people. Although there are some changes, there are not a whole lot," said Earl Brown, professor of virology at the University of Ottawa in Canada, who was not involved with the study.
If the virus remains stable, any vaccine created is likely to stay useful, Brown said. He said the findings also suggest the SARS virus may have been in humans longer than thought.
"Generally you'd see a virus mutate more when it first comes into a new situation," Brown said. "I'd be concerned that maybe it's been in humans longer than we think and it's already made its little changes and has settled down to a happy lifestyle."
Recent fears that about 12 people in Hong Kong had relapsed after being sent home from the hospital have turned out to be unfounded, said Dr. David Heyman, the agency's communicable diseases chief.
"SARS has not reoccurred in any of these patients. These patients have been readmitted to hospital, but with diseases other than SARS," he said.
The World Health Organization said it was still optimistic it could prevent SARS from taking root in communities, even in China, where the number of reported cases continues to rise.