X-Message-Number: 21433 Date: Thu, 20 Mar 2003 09:08:27 -0700 Subject: Re: Some SARS numbers From: (Tim Freeman) Message #21432 From: Charles Platt <> >...I will be tracking the >number of announced cases during the next week or two in an effort to >determine whether SARS is still in the exponential growth phase. Good idea. Here are some numbers to start with: Mar 17, 2003 -- 167 cases per http://www.who.int/csr/table/en/index.html Mar 18, 2003 -- 219 cases per http://www.who.int/csr/sars/tablemarch18/en/ Mar 19, 2003 -- 264 cases per http://www.who.int/csr/don/2003_03_19/en/ Mar 20, 2003 -- 306 cases per http://www.who.int/csr/sarscountry/2003_03_20/en/ This is in exceess of 10% per day, which is pretty scary. However, at this point there's a lot of variance in what gets counted as SARS from one day to the next. The numbers are hard to interpret now, but after a few weeks this variance is unlikely to obscure exponential growth. >Based on very fragmentary evidence, some from CDC, some from WHO, and some >from an ICU doctor in the Price of Wales hospital in Hong Kong, I will >assume tentatively that the average incubation period for SARS is about 5 >days, and the average patient infects three new patients. In the US, >assuming we have 10 patients currently, the initial exponential growth >looks like this, assuming that a new patient does not become contagious >until the end of the 5-day incubation period (which may not in fact be >true): > >5 days 30 patients >10 days 90 patients >15 days 270 patients >20 days 800 approx patients >25 days 2400 approx patients >30 days 7200 approx patients >35 days 21600 approx patients >40 days 65000 approx patients >45 days 200000 approx patients According to http://www.nytimes.com/2003/03/20/international/asia/20CHIN.html?ex=1048741200&en=4813ab03a64e7ed7&ei=5062&partner=GOOGLE the Chinese apparently had their first case in November, 2002, if you're willing to believe that that's the same virus as SARS. That's 100 days, or 20 incubation periods by your model. Assuming they started with one case, by the simple model above that would be 3,486,784,401 cases by now, which can't possibly be true. Guangdong has 86 million people, according to: http://www.unescap.org/pop/database/chinadata/guangdong.htm We can repair the model in several possible ways: 1. Change the infectivity. Suppose Guangdong has 10000 secret cases now since the public cases are 306 and it's hard to keep a huge contagious disesase secret or contained. The 20th root of 10000 is about 1.6, so if these assumptions are true each infected person produces 1.6 additional cases per incubation period. 2. Maybe it is already past its exponential phase in Guangdong. 3. Maybe SARS is a different virus from what was observed in November in Guangdong. Hmm. The Chinese came close to successfully containing the infection in Guangdong, if it's the same infection. I hope they're eventually willing to tell people what they did. -- Tim Freeman Which is worse: ignorance or apathy? Who knows? Who cares? GPG public key fingerprint ECDF 46F8 3B80 BB9E 575D 7180 76DF FE00 34B1 5C78 Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=21433