For the historical interest (and because echoes of the hysteria are still heard in the context of “bird flu”) I have posted here some analysis from the time of the mass hysteria known as SARS. I have updated the links where they have changed and I can still find the relevant document. Unfortunately two documents no longer appear to be online, as noted in the text. Any help locating them would be gratefully received.
*** PLEASE NOTE ***
On 23 June 2003 the WHO declared that Hong Kong was no longer an area with recent local transmission of SARS. The outbreak is officially over, although a few patients remain in hospital. My thanks to various people who have emailed me with comments. I hope this site has helped to allay some people’s fears.
THE SARS OUTBREAK IN HONG KONG – DATA & RISK ANALYSIS
Click for a graphical summary
Click here for the underlying Excel spreadsheet.
Summary @ 23 June 2003
||Latest day’s change
|Of which: age 65-79
|0-64 chronically ill / late treatment
|Intensive Care (snapshot)
|Still in hospital (not ICU)
* by which I mean people who were under 65, not chronically ill, and sought treatment early
For a full list of the age, sex, and previous medical status of all the people dying so far click here.
The “long run” mortality rate of the cases admitted to hospital using the figure for the cases “concluded” one way or the other is 296/(296+1411) = 17.3%. However, calculating the overall mortality rate of SARS also requires us to know how many people in total have been exposed to the causative factors (virus and/or other things); no data is currently available on this except some small sample tests in North America which show that around 15% of the general population test positive for the virus.
There is a very helpful (but quite technical) information paper at the Chinese University website with data from Prince of Wales Hospital. It shows that approximately 20-25% of SARS patients had symptoms severe enough to require Intensive Care admission and 10-15% require intubation and ventilation (i.e. they cannot breath unaided).
Pneumonia in Hong Kong
In Hong Kong the “normal” annual number of cases of pneumonia which require hospitalisation is in the range 15,000 – 25,000. 50% of these pneumonias have an unknown cause. (Source: Hong Kong Medical Association [Update March 2007 – unfortunately this link is broken and I have not been able to find a copy of this report online.]) A much higher number of cases of pneumonia (perhaps 100,000-150,000) are not severe enough to warrant hospitalisation, but I have not found hard data on this.
The normal number of deaths due to pneumonia in Hong Kong is in the range 2,000-3,000 per year.
(Source: HK Government Health Department Annual Report)
The vast majority of these deaths from pneumonia occur amongst the elderly. For example, in year 2000 there were 3,041 deaths from pneumonia, of which 79 were in people age 44 or lower.
(Source: HK Government Health Department Annual Report ibid)
1. How easy is it to catch?
From the data on the Amoy Gardens Block E case it appears that about 1 in 4 of the people in that building have shown some symptoms. This makes it similar to common colds or influenza, but much easier to catch than “typical” pneumonia.
It seems that very few people who come in casual contact with virus carriers develop symptoms of pneumonia. Even amongst health workers attending SARS patients (albeit with a high degree of protection) only about 6% show any symptoms of SARS at all.
2. Who dies with it?
On 27th May the Health Department published a bulletin [Note: March 2007, the link is now to the Chinese version of this – I can no longer find the english version online.] including the incidence by age and sex of the cases and deaths up to 16th May. This emphasises the, not unexpected, higher mortality rates with increasing age. However, the age profile is skewed more towards younger people than the profile for “typical” pneumonia.
Men are significantly more likely to die if they are diagnosed with SARS than women, constituting about 45% of the cases, but 60% of the deaths.
Curiously, despite constituting about 22% of the cases of SARS, only six deaths have occurred amongst public hospital medical workers, which, by the “concluded case” calculation method, suggests that medical workers have a mortality rate of about 1.6% and everyone else has a mortality rate of just over 20%. Moreover, the Amoy Gardens patients have a mortality rate of about 10% and those patients who are neither medical nor Amoy Gardens have a mortality rate of about 33%.
3. How high is the risk relative to others?
Since SARS was first publicised (on 11 March) – i.e. in 104 days – there have been 296 deaths classified as SARS, amongst over 1000 total deaths from all types of “atypical pneumonia”. (Source: Presentation at FCC by Vivian Wong, 12 May. Extrapolated from end April. Not available online.) In other words, throughout this period, twice as many people have died of other types of atypical pneumonia as from SARS.
In that same period up to June 23, around 550 people in HK would have been expected to die from accidental injury or poisoning.
b) Serious Illness
The data on the Chinese University site data above suggests that the cumulative number of people with SARS who have spent some time in ICU is around 400.
In year 2000 there were the following number of cases (not deaths) of these serious diseases in HK:
Viral hepatitis: 683
Bacillary dysentery: 310
(Source for all illness/mortality data: HK Government Health Department Annual Report 2000/2001)
4. Is this really something new?
At a Panel Lunch at the Foreign Correspondents’ Club in HK on 12 May, Vivian Wong, Director at the Hospital Authority presented data which showed that the total number of deaths from Atypical Pneumonias of all causes in the period Jan-April 2003 was less than 5% higher than the number of deaths in the same period in 2002. The number of reported cases of Atypical Pneumonia increased by about 40%, however, Ms Wong stated that a significant proportion of the cases being reported this year were suffering from symptoms that would not normally be sufficiently serious for them to be hospitalised or even to consult a doctor.
Reading from a printout of a graph, the total number of deaths from Atypical Pneumonia in Jan-Apr 2002 was around 1220. In Jan-Apr 2003 the number was around 1280, of which 157 were classified as SARS. In other words, SARS has not significantly increased the overall death rate from atypical pneumonias, and still only constitutes about 12% of the deaths in HK from them so far in 2003 (although SARS deaths constitute about one third of atypical pneumonia deaths since the start of SARS dignoses in mid-March). Pneumonia deaths fluctuate significantly by year, and an increase of 5% is well within the normal fluctuation.
Discussion, Tentative Conclusions & Open Questions
(This section contains personal opinions and hypotheses)
It seems to me that there is clear evidence that if you search for SARS using the current heuristic “tests” then you will find it in quite significant numbers. It seems reasonable that the medical staff are being examined for SARS most closely, and a fairly large number of them (386 @ 13 June) have been diagnosed with it. However, a press release on 30 April stated that 5800 HK medical staff (doctors plus nurses) were then engaged in treating SARS patients. So even with very intimate contact with other sufferers (albeit with heavy protection) only about 6% of medical staff have contracted any symptoms of SARS.
In the Amoy Gardens case there were efforts to diagnose all residents, but I presume these were not as rigourous or regular as those for medical staff. This results in a higher number of identified cases for Amoy Gardens, and hence a much lower mortality rate than non-medical, non-Amoy cases.
This third category (non-medical, non-Amoy) is essentially self-diagnosing. The case will only enter the statistics if the patient chooses to visit a medical facility. With less severe cases this will result it a much lower notification rate, particularly amongst men (if experience of other viruses is followed).
When considering the significantly higher ratio of death to diagnosis amongst men, it seems that there could be two reasons for this: either men are simply more likely to die from SARS (possibly due to many more being smokers) or far more women than men with mild cases of SARS choose to seek medical help.
For me, the key data is that on the medical staff. We can assume this is a thoroughly observed and tested group. Out of 5800 public hospital staff, only 6% have shown symptoms of the disease and only 1.6% of that 6% (i.e. six out of 5800 people) has died with the disease. Clearly, data on how many medical staff have required ICU treatment would be valuable.
For the more general population the data presented at the FCC on 12 May is critical – it shows that there is no significant increase in the number of deaths due to Atypical Pneumonia in Hong Kong. More cases have been identified because of the focus on it, but many of these are not severe enough to warrant hospital treatment except as a means of quarantine.
It is also interesting to note that the best available test for the CoronaVirus allegedly responsible for SARS showns positive in only 40% of patients labelled as “SARS” and also shows positive in around 15% of the general population without SARS symptoms. (Source: Head of Toronto Microbiology, widely quoted in the press)
Professor Yuen Kwok-Yung of HKU was reported in the South China Morning Post on 9 May: “He said most Sars patients died because of the serious side-effects of the steroids. High doses of steroids suppress
patients’ immunity, making them more susceptible to infection by other bacteria.” It would be very interesting to study the mortality rate of a random group of people matching the demographics and health condition of the SARS patients if treated with the regime of drugs which has been imposed on them in Hong Kong.
In a press conference on 10 May various senior medical figures talked of a revised treatment regime which is now much lighter on steroids. Read the Press Release here